I Chronic Diseases in Canada n this issue

Health
Canada
Santé
Canada
Chronic Diseases
in Canada
Volume 22, No 1
2001
In this issue
1
Recent Trends in Fetal and Infant Outcomes Following Post-term
Pregnancies
Shi Wu Wen, K S Joseph, Michael S Kramer, Kitaw Demissie, Lawrence Oppenheimer,
Robert Liston and Alexander Allen for the Fetal and Infant Mortality Study Group,
Canadian Perinatal Surveillance System
6
The Duration of Major Depressive Episodes in the Canadian General
Population
Scott B Patten
12
Emigration Patterns of Cancer Cases in Alberta, Canada
Juanita Hatcher and Marilou Hervas
18
The Economic Burden of Mental Health Problems in Canada
Thomas Stephens and Natacha Joubert
24
The Storage of Household Long Guns: The Situation in Quebec
Michel Lavoie, Lise Cardinal, Antoine Chapdelaine and Danielle St-Laurent
Book Review
30
Design and Analysis of Cluster Randomization Trials in Health Research
Reviewed by Yang Mao
(continued on reverse)
Our mission is to help the people of Canada
maintain and improve their health.
Health Canada
(Contents continued)
31
New Resource
32
Calendar of Events
33
2000 Peer Reviewers
34
Indexes for Volume 21, 2000
Information for Authors (on inside back cover)
Published by authority of the Minister of Health
© Minister of Public Works and Government Services Canada 2001
ISSN 0228-8699
Aussi disponible en français sous le titre Maladies chroniques au Canada
Recent Trends in Fetal and Infant Outcomes Following
Post-term Pregnancies
Shi Wu Wen, K S Joseph, Michael S Kramer, Kitaw Demissie, Lawrence Oppenheimer,
Robert Liston and Alexander Allen for the Fetal and Infant Mortality Study Group,
Canadian Perinatal Surveillance System*
Abstract
All births and infant deaths in1985–87 and 1992–94 in Canada, except in Ontario and
Newfoundland, were analyzed to assess the potential impact of the recent increased use of
elective labour induction for post-term pregnancies. Probabilistic linkage was carried out
of infant death records (Canadian Mortality Database) and respective birth registrations
(Canadian Birth Database) for the periods 1985–87 and 1992–94. The combined fetal and
infant mortality declined by 20–30% between 1985–87 and 1992–94 at each gestational week
beginning at 37 weeks, with no increased reduction among post-term pregnancies. Asphyxiarelated fetal and infant deaths, the most likely cause of death being preventable by labour
induction for post-term pregnancies, did not decrease among post-term pregnancies. On the
contrary, a substantial decrease of asphyxia-related deaths was observed at 37 and 38 weeks
over the same periods of time. Because fetal and infant deaths are rare events and because
the number of pregnancies passing 42 weeks of gestation decreased dramatically during
1992–94, statistically unstable results may be inevitable in the comparison of mortality in this
group of pregnancies.
Key words: labour induction; mortality; post-term pregnancy
Introduction
Post-term pregnancies are pregnancies that reach at
least 42 weeks of gestation.1 Perinatal mortality and the
occurrence of various obstetric complications have been
found to be higher in post-term than in term pregnancies.2,3
Two general management approaches have been developed
to reduce the risk of these adverse outcomes: elective
labour induction when the pregnancy reaches 41 or
42 weeks of gestation, or expectant management with
frequent fetal monitoring and selective labour induction.
The two approaches remain controversial.4–9
Proponents of elective labour induction cite evidence
from randomized controlled trials showing that labour
induction is associated with reduced perinatal mortality.4–6
Concerns have been raised, however, that results from
tightly controlled trials may not be applicable in routine
practice and that widespread implementation of routine
induction for post-term pregnancies may lead to increased
rates of cesarean section and other obstetric interventions.7–9
Despite these controversies, the use of elective labour
induction for post-term pregnancies has increased
dramatically in Canada since the early 1990s, mostly at
41weeks of gestation.6 During this period of time, to our
knowledge, the only apparent difference in obstetric care
between post-term and term pregnancies in Canada was
a tendency to induce labour electively for the former,
but not for the latter.
Author References
Shi Wu Wen, The Bureau of Reproductive and Child Health, Centre for Healthy Human Development, Health Canada, Ottawa, Ontario; and
Department of Obstetrics and Gynecology and Clinical Epidemiology Program, University of Ottawa, Ottawa, Ontario
K S Joseph, Department of Pediatrics, Dalhousie University Faculty of Medicine, Halifax, Nova Scotia
Michael S Kramer, Departments of Pediatrics and of Epidemiology and Biostatistics, McGill University Faculty of Medicine, Montreal, Quebec
Kitaw Demissie, Department of Environmental and Community Health, Robert Wood Johnson School of Medicine, Piscataway, New Jersey
Lawrence Oppenheimer, Department of Obstetrics and Gynecology, Faculty of Medicine, University of Ottawa, Ottawa, Ontario
Robert Liston, Department of Obstetrics and Gynecology, University of British Columbia, Vancouver, British Columbia
Alexander Allen, Department of Pediatrics, Dalhousie University Faculty of Medicine, Halifax, Nova Scotia*
*Contributing members: Margaret Cyr (Statistics Canada), Martha Fair (Statistics Canada), Sylvie Marcoux (University of Laval), Brian McCarthy (CDC),
Doug McMillan (past member, University of Calgary), Arne Ohlsson (University of Toronto), Russell Wilkins (Statistics Canada)
Correspondence: Dr. Shi Wu Wen, The Bureau of Reproductive and Child Health, Centre for Healthy Human Development, Health Canada, Tunney’s
Pasture, AL 0701D, Ottawa, Ontario K1A 0L2; Fax: (613) 941-9927; E-mail: [email protected]
2001
1
A recent observational study suggests that the
increased use of elective labour induction for post-term
pregnancies has contributed to the decline in fetal deaths
among these pregnancies in Canada.6 However, this
study did not assess infant mortality or causes of fetal
death. We hypothesized that if routine, elective labour
induction in recent years has had an important impact on
the fetal and infant outcomes of post-term pregnancies in
the form of the reduction of fetal and infant mortality,
of asphyxia-related deaths compared to other causes of
death, and of perinatal compared to post-neonatal death,
it would be more evident in post-term pregnancies than
in term pregnancies. We carried out an epidemiologic
study using linked birth and infant death files to test
these hypotheses.
Methods
We used data for live births and stillbirths from
Statistics Canada’s Canadian Birth Database10 for the
years 1985–1994 and data for fetal and infant deaths
from the Canadian Mortality Database for the years
1985–1995. Fetal death is defined as stillbirth with birth
weight ≥ 500 g or gestational age ≥ 20 weeks; neonatal
death is defined as a live birth of an infant that died
before the 28th full day of life; post-neonatal death is
defined as a live birth of an infant that died between the
29th and the 364th full day of life; and infant death is
defined as a live birth of an infant that died before the
364th full day of life. A probabilistic linkage was carried
out using previously validated methods to link infant
death records with respective birth registrations.11,12
Uncertain linkages were resolved after a manual examination of the relevant birth and death registration
documents.
Ontario births were excluded from the analysis
because of documented problems with data quality.13
Newfoundland births were also excluded from the
analysis of time trends, because data from this province
were not available before 1991. Information in the linked
files of live births and infant deaths was subjected to
internal data quality checks, including procedures to
exclude duplicate records.
The linked birth and death information enabled the
creation of birth cohorts with follow-up information on
mortality in the first year after birth. Thus, infants born
in 1985 were followed through 1986 to calculate infant
mortality rates. Similarly, while the last birth cohort
constructed was of live births in 1994, follow-up for
infant death among newborns in this cohort extended
through 1995.
For the current study, only term and post-term births
(i.e., those of 37 or more weeks of gestation) were
included in the analysis. Because fetal and infant
mortality decreased continuously during the 10-year
period covered by the study and differences between
successive years were small, we combined the data for
1985–1987 births and 1992–1994 births to enhance the
statistical stability of the estimates. We calculated rates
2
Chronic Diseases in Canada
of fetal death, neonatal death, post-neonatal death, and
overall fetal and infant death rates in each gestational
week for the two study periods. Relative risks and 95%
confidence intervals were used to compare mortality
rates between the two study periods, with 1985–87
serving as the reference.
We further analyzed fetal and infant mortality caused
by asphyxia, on which labour induction probably has the
largest impact. Only one underlying cause of death is
recorded in Statistics Canada’s Canadian Mortality Database, which is coded using the International Classification
of Diseases, 9th Revision (ICD-9) classification system.
We used a classification system adopted by the International Collaborative Effort on Perinatal and Infant
Mortality14 to group codes with clinical conditions that
may directly or indirectly lead to asphyxia-related death.
These conditions include maternal death (ICD-9 761.6),
malpresentation before labour (ICD-9 761.7), placenta
previa or other placental abnormalities (ICD-9 762.0762.2), prolapsed cord or other unspecified conditions
of umbilical cord (ICD-9 762.4, 762.5, 762.6), breech
delivery and extraction (ICD-9 763.0), disorders relating
to long gestation and high birth weight (ICD-9 766),
birth trauma (ICD-9 767), intrauterine hypoxia and birth
asphyxia (ICD-9 768), meconium aspiration syndrome
(ICD-9 770.1), subarachnoid hemorrhage (ICD-9 772.2),
convulsions in newborn (ICD-9 779.0) and coma or
other abnormal cerebral signs (ICD-9 779.2).
We considered the potential impact of the reclassification of gestational age among post-term pregnancies,
caused by more frequent labour induction at 41 or more
weeks of gestation,6 on gestational age-specific mortality,
and then carried out a parallel analysis combining all
births at 41 or more weeks and compared the results with
the main analysis using finer groupings of gestational
age (37, 38, 39, 40, 41, 42 and ≥ 43 weeks of gestation).
Results
The number of births and deaths are presented in
Tables 1 and 2. Compared with 1985–87, births at 37,
38, 39 and 41 weeks of gestation increased as a proportion of total births in 1992–94, but decreased at 40 and
(especially) 42 and ≥43 weeks (Table 1)
The results of the comparison of fetal and infant
mortality rates between 1985–87 and 1992–94 are
presented in Table 3. The fetal death rate at 37 weeks
decreased by 28% from 1985–87 to 1992–94. The
reduction in fetal death rates increased sequentially from
40 weeks onwards, being 12%, 15%, 20% and 35%
among those at 40, 41, 42 and ≥ 43 weeks respectively.
On the other hand, the decrease in neonatal mortality was
greatest at 40 weeks (39% decrease in 1992–94 versus
1985–87; see Table 3). Compared with births at 40 weeks,
the decreases were smaller at 37, 38, 39 and 41 weeks
and there was no statistically significant reduction at
42 weeks (relative risk [RR] 0.94, 95% confidence
interval [CI] 0.64-1.37). At ≥ 43 weeks, there was even
a statistically nonsignificant increase. Post-neonatal
Vol 22, No 1
TABLE 1
Gestational age distribution of term and post-term births (combining stillbirths and live births),
Canada excluding Ontario and Newfoundland, 1985–87 and 1992–94
1985–87
Gestational age
(weeks)
Number
1992–94
Percent
Number
Percent
Percent change
33,143
4.76
40,444
5.69
+21.0
38
88,137
12.65
102,087
14.36
+14.8
39
136,154
19.54
153,215
21.56
+11.6
40
259,698
37.27
230,567
32.44
−12.0
41
95,505
13.71
104,147
14.65
+8.2
42
34,549
4.96
24,161
3.40
−30.5
≥43
2,878
0.41
956
0.13
−65.9
Total
650,064
37
655,577
TABLE 2
Number (rate per 1,000) of fetal death, neonatal death, post-neonatal death, total fetal and infant death,
and fetal and infant death for asphyxia-related conditions by gestational age in term and post-term
births, Canada excluding Ontario and Newfoundland, 1985–87 and 1992–94*
Gestational age
Fetal death
Neonatal death
Post-neonatal death
Fetal and infant death
Asphyxia-related fetal
and infant death
1985–87
1985–87
1992–94
1985–87
1992–94
1985–87
1992–94
1985–87
1992–94
37
232 (7.0)
205 (5.1)
134 (4.1)
139 (3.5)
129 (3.9)
121 (3.0)
495 (14.9)
1992–94
465 (11.5)
143 (4.3)
124 (3.1)
38
297 (3.4)
255 (2.5)
204 (2.3)
174 (1.7)
270 (3.1)
259 (2.5)
771 (8.8)
688 (6.7)
184 (2.1)
156 (1.5)
39
239 (1.8)
239 (1.6)
213 (1.6)
167 (1.1)
296 (2.2)
256 (1.7)
748 (5.5)
662 (4.3)
167 (1.3)
160 (1.0)
40
360 (1.4)
280 (1.2)
349 (1.4)
189 (0.8)
515 (2.0)
353 (1.5)
1,224 (4.7)
822 (3.6)
257 (1.0)
200 (0.9)
41
157 (1.6)
146 (1.4)
141 (1.5)
98 (0.9)
200 (2.1)
130 (1.3)
498 (5.2)
374 (3.6)
105 (1.1)
100 (1.0)
42
84 (2.4)
47 (2.0)
67 (1.9)
44 (1.8)
81 (2.4)
29 (1.2)
232 (6.7)
120 (5.0)
57 (1.7)
37 (1.5)
≥43
23 (8.0)
5 (5.2)
7 (2.5)
3 (3.2)
8 (2.8)
0 (0.0)
38 (13.2)
8 (8.4)
13 (4.5)
5 (5.2)
≥41
264 (2.0)
198 (1.5)
215 (1.6)
145 (1.1)
289 (2.2)
159 (1.2)
768 (5.8)
502 (3.9)
175 (1.3)
142 (1.1)
Total
1,392 (2.1)
1,177 (1.8)
1,115 (1.7)
814 (1.2)
1,499 (2.3)
1,148 (1.8)
4,006 (6.2)
3,139 (4.8)
926 (1.4)
782 (1.2)
*Fetal death is defined as stillbirth with birth weight ≥ 500 g or gestational age ≥ 20 weeks; neonatal death is defined as a live birth of an infant that died prior to the 28th full day of life; post neonatal
death is defined as a live birth of an infant that died between the 29th and the 364th full day of life; infant death is defined as a live birth of an infant that died prior to the 364th full day of life.
mortality declined at every week of gestation in 1992–94
compared to 1985–87; this decrease was most evident at
41 or more weeks of gestation.
The combined fetal and infant mortality rate had
decreased by 20%–30% for every week of gestation in
1992–94 compared to 1985–87. The reduction at 41 or
more weeks of gestation was slightly larger than the
reduction at 37–40 weeks, mainly because of a larger
reduction in post-neonatal mortality rates at these
gestations.
In general, fetal and infant mortality caused by
asphyxia decreased in 1992–94 compared to 1985–87.
However, the difference was statistically significant only
for 37 and 38 weeks of gestation. Results obtained from
analysis combining all births at 41 or more weeks
(second row from bottom of Table 3) were generally
2001
consistent with those using finer categories of post-term
pregnancies.
Discussion
Our study showed a moderate increase in the
proportion of births at 41 weeks of gestation and
substantial decreases in this proportion at 42 weeks and
≥ 43 weeks in 1992–94 compared to 1985–87. These
changes are probably partly attributable to better dating
of pregnancies with more frequent use of ultrasound
early in pregnancy,15,16 although more frequent labour
induction at ≥ 41 weeks has doubtless also played a role.
Although we have no direct data on labour induction, a
recent study showed a substantial rise in rates of labour
induction at 41 weeks from the early 1990s in the
majority of Canadian hospitals surveyed.6 The continued
increase in inductions among post-term pregnancies after
3
TABLE 3
Relative risk (95% confidence interval)* for fetal death (per 1,000 total births),
neonatal death (per 1,000 live births), post-neonatal death (per 1,000 survivors 28 days of age),
total fetal and infant death (per 1,000 total births), and fetal and infant death (per 1,000 total births)
for asphyxia-related conditions by gestational age in term and post-term births,
Canada excluding Ontario and Newfoundland
Fetal and infant
death
Asphyxia-related
fetal and infant
death
Gestational age
Fetal death
Neonatal death
Post-neonatal
death
37
0.72 (0.60, 0.87)
0.85 (0.67, 1.08)
0.77 (0.60, 0.98)
0.77 (0.67, 0.87)
0.71 (0.55, 0.91)
38
0.74 (0.63, 0.88)
0.74 (0.60, 0.90)
0.83 (0.70, 0.98)
0.77 (0.69, 0.85)
0.73 (0.59, 0.91)
39
0.89 (0.74, 1.06)
0.70 (0.57, 0.85)
0.77 (0.65, 0.91)
0.79 (0.71, 0.87)
0.85 (0.69, 1.06)
40
0.88 (0.75, 1.02)
0.61 (0.51, 0.73)
0.77 (0.67, 0.88)
0.75 (0.69, 0.83)
0.88 (0.73, 1.05)
41
0.85 (0.68, 1.07)
0.64 (0.49, 0.82)
0.60 (0.48, 0.74)
0.69 (0.60, 0.79)
0.87 (0.66, 1.15)
42
0.80 (0.56, 1.14)
0.94 (0.64, 1.37)
0.51 (0.33, 0.78)
0.74 (0.59, 0.92)
0.93 (0.61, 1.40)
≥ 43
0.65 (0.25, 1.72)
1.29 (0.33, 4.97)
0.00 (0.00, 2.03)
0.63 (0.30, 1.35)
1.16 (0.41, 3.24)
≥ 41
0.77 (0.64, 0.93)
0.69 (0.56, 0.85)
0.56 (0.47, 0.68)
0.67 (0.60, 0.75)
0.83 (0.73, 1.04)
Total
0.84 (0.78, 0.91)
0.72 (0.66, 0.79)
0.76 (0.70, 0.82)
0.78 (0.74, 0.81)
0.84 (0.76, 0.92)
*1992–94 versus 1985–87 rates; fetal death is defined as stillbirth with birth weight ≥ 500 g or gestational age ≥ 20 weeks; neonatal death is defined as a live birth of an infant that died prior to
the 28th full day of life; post neonatal death is defined as a live birth of an infant that died between the 29th and the 364th full day of life; infant death is defined as a live birth of an infant that died
prior to the 364th full day of life.
1991 has probably been fueled, at least in part, by
evidence from randomized controlled trials supporting
this practice4,5 and by the increasing availability of
intracervical and vaginal prostaglandin gels to assist with
cervical ripening.6
During the same period of time, we observed a
relatively larger (but statistically nonsignificant) reduction in fetal death among pregnancies at ≥ 41 weeks as
compared with those at 40 or 39 weeks. This finding is
consistent with a recent Canadian study.6
It is notable that the decrease in neonatal mortality
at 41 weeks was not greater than at 40 weeks, and that
no decrease occurred at 42 or ≥ 43 weeks; in fact at
≥ 43 weeks there was a statistically nonsignificant
increase (Table 3). In general, larger reductions in fetal
death have been accompanied by smaller reductions in
neonatal mortality. For example, the decrease in fetal
death was greater at 37 and 38 weeks than at 39 and
40 weeks, whereas the reverse was true for the decrease
in neonatal mortality. This pattern of smaller reductions
in neonatal mortality than in fetal mortality raises the
possibility that more frequent use of medical interventions such as labour induction may have merely postponed some deaths. An alternative explanation for this
phenomenon is that labour induction may be beneficial
in certain circumstances but harmful in others.
The reduction in post-neonatal mortality in 1992–94
versus 1985–87 was larger at ≥ 41 weeks than at 37 to
40 weeks (Table 3). This finding is in contrast to randomized controlled trials assessing the efficacy of labour
induction for post-term pregnancies that have focused on
perinatal mortality.4 It is possible that labour induction
may have a larger impact on reducing post-neonatal
4
Chronic Diseases in Canada
mortality than it does on perinatal mortality. If this is the
case, the assessment of the efficacy of labour induction
for post-term pregnancies should be expanded to include
the post-neonatal period.
We hypothesized that labour induction for post-term
pregnancies would have the largest impact on asphyxiarelated fetal and infant mortality and morbidity. The
reduction observed in fetal and infant mortality due to
asphyxia-related conditions at ≥ 41 weeks, however, was
not larger than at 40 weeks (Table 3). During the study
period, the clinical definition for certain asphyxia-related
conditions such as respiratory distress may have changed.
However, our earlier study based on hospital discharge
data found that the incidence of coded diagnoses such as
respiratory distress and meconium aspiration syndrome
was quite stable during the study period.17 Moreover,
since we used a broad and inclusive definition of
asphyxia-related deaths, such a shift in clinical definition
probably had a limited impact on our study results.
Our population-based results should reflect routine
practice better than controlled trials. We realize the
limitations inherent in any observational study based on
administrative databases. Such data are prone to a certain
degree of coding errors,18 which may be random or may
contain systematic biases. The observational study
design and the lack of information on induction in the
data render any inference about the relation between
labour induction and fetal and infant mortality necessarily
indirect. Moreover, because fetal and infant deaths are
rare events, and because the number of pregnancies
passing 42 weeks decreased dramatically during
1992–1994, statistically unstable results were inevitable
Vol 22, No 1
in the comparison of mortality in this group of pregnancies. Nonetheless, our findings are biologically plausible.
Trials and guidelines vary widely in terms of gestational age for induction. In the Cochrane systematic
review, trials demonstrated an effect of reducing perinatal
mortality only for induction conducted after 42 weeks of
gestation.4 The Society of Obstetricians and Gynecologists
of Canada initially recommended that women who reach
“41–42” weeks of gestation should be offered elective
induction,19 and in practice the tendency has been to
induce pregnancies closer to 41 rather than 42 weeks.6
The most difficult challenge in the management of postterm pregnancy may be related to determining the exact
time that a given pregnancy becomes “post-term.”9 It is
thus difficult to establish a rigid and arbitrary cut-off
point for induction.
Acknowledgments
We thank the Vital Statistics Registrars of the provinces and
territories who gave us access to their data files. This study
was conducted under the auspices of the Canadian Perinatal
Surveillance System.
References
1. Bakketeig L, Bergjo P. Post-term pregnancy: magnitude of
the problem. In: Chalmers I, Enkin M, Keirse MJNC,
Editors. Effective care in pregnancy and childbirth.
Oxford (UK): Oxford University Press; 1989. p. 765–75.
2. McClure Browne JC. Postmaturity. Am J Obstet Gynecol
1963;85:573–82.
3. Shime J, Gare DJ, Andrews J, Bertrand M, Salgado J,
Whillans G. Prolonged pregnancy: surveillance of the
fetus and the neonate and the course of labour and
delivery. Am J Obstet Gynecol 1984;148:547–52.
4. Cowley P. Interventions for preventing or improving the
outcome of delivery at or beyond term [Cochrane review].
In: The Cochrane Library; Issue 1, 1999. Oxford: Update
Software.
5. Hannah ME, Hannah WJ, Hellman J, Hewson S, Milner R,
Willan A and the Canadian Multicenter Post-term
Pregnancy Trial Group: Induction of labour as compared
with serial antenatal monitoring in post-term pregnancy.
N Engl J Med 1992;326:1587–92.
2001
6. Sue-A-Quan AK, Hannah ME, Cohen MM, Foster GA,
Liston RM. Effect of labour induction on rates of stillbirth
and cesarean section in post-term pregnancies. Can Med
Assoc J 1999;160:1145–9.
7. Luther ER. Post-dates and induction: Where are we now?
Reproductive Care Program of Nova Scotia Newsletter,
June 1996.
8. Keirse MJNC. Post-term pregnancy: New lessons from an
unresolved debate. Birth 1993;20:102–5.
9. Sanders N, Paterson C. Effect of gestational age on
obstetric performance: when is “term” over? Lancet
1991;338:1190–2.
10. Fair ME, Cyr M. The Canadian birth database: a new
research tool to study reproductive outcomes. Health Rep
1993;5:281–90.
11. Howe GR, Lindsay J. A generalized iterative record
linkage computer system for use in medical follow-up
studies. Computer Biomed Res 1981;14:327–40.
12. Smith ME, Silins J. Generalized iterative record linkage
system. In: Proceedings of the American Statistical
Association, Social Statistics Section, 1981:128–37.
13. Joseph KS, Kramer MS. Recent trends in infant mortality
rates and proportions of low-birth-weight live births in
Canada. Can Med Assoc J 1997;157:535–41.
14. Cole S, Hartford RB, Bergsjo P, McCarthy B.
International Collaborative Effort (ICE) on birth weight,
plurality, perinatal and infant mortality. Acta Obstet
Gynecol Scand 1989;68:113–7.
15. Kramer MS, McLean FH, Boyd ME, Usher RH. The
validity of gestational age estimation by menstrual dating
in term, preterm and post-term gestations. JAMA
1988;260:3306–8.
16. Goldenberg RL, Davis RO, Cutter GR, Hoffman HJ,
Brumfield CG, Foster JM. Prematurity, postdates, and
growth retardation: the influence of use of ultrasonography
on reported gestational age. Am J Obstet Gynecol
1989;160:462–70.
17. Wen SW, Liu S, Fowler D. Trends and variations in
neonatal length of in-hospital stay in Canada. Can J Public
Health 1998;89:115–9.
18. Huston P, Naylor CD. Health services research: Reporting
on studies using secondary data sources. Can Med Assoc J
1996;155:1697–702.
19. SOGC Committee Opinion. Management of post-term
pregnancy. April 1994:29. ■
5
The Duration of Major Depressive Episodes in the
Canadian General Population
Scott B Patten
Abstract
The National Population Health Survey (NPHS) has provided a wealth of new data
concerning major depression in the Canadian general population. The NPHS included a brief
predictor of major depression, and also two questions (only one of which was asked of each
subject) concerned with the duration of episodes in the preceding year. A striking finding was
that many of the episodes identified were very brief. In this paper the NPHS data were
examined from a different perspective in order to derive a complementary perspective on the
episode duration data. Data from the 1994/95 and 1996/97 cycles of the NPHS were used in
the analysis. The longitudinal data were used to generate approximations of age and genderspecific incidence for members of the population over the age of 12 years. An estimate of
prevalence was made from the 1996/97 cross-sectional file. A basic expression relating
prevalence to incidence and mean duration of illness was then applied within age and gender
categories. Taken together, the incidence and prevalence data from the NPHS suggest a
longer duration than was indicated by the NPHS interview duration item. A probable
explanation is that the NPHS duration question had an upper limit of 52 weeks, whereas some
episodes of major depression last longer than this. Particularly long episodes could have a
large impact on mean duration in the population. Nevertheless, these data confirm the
heterogenous nature of this condition; many people with the syndrome of major depression
may have quite brief episodes.
Key words: depressive disorder, epidemiology, prognosis, prevalence, incidence
Introduction
A variety of prevalence estimates for major depression
have been reported in various countries.1–7 Recent
estimates of the 12-month period prevalence of major
depression in the Canadian population derive from the
Canadian National Population Health Survey (NPHS) 8,9
and the Mental Health Supplement of the Ontario Health
Survey.10 Estimates from several additional Canadian
surveys have been published.11,12 Since the NPHS is a
longitudinal study, it is possible to approximate incidence
using the NPHS data,13 with the proviso that the followup interval for the longitudinal component of the NPHS
is two years and the measurement instrument employed
identifies episodes occurring only in the 12-month period
preceding the interview. The proportion of non-depressed
subjects in 1994/95 who were subsequently found to
have major depression in 1996/97 may overestimate the
annual incidence proportion, since it may include some
persons with an onset of major depression in the year
following the 1994/95 interview. This would occur if
these episodes did not resolve until two weeks or more
into the year preceding the 1996/97 interview. These
subjects would appear in the numerator of the incidence
expression, inflating the estimated incidence proportion.
The NPHS is an ongoing longitudinal community
study conducted by Statistics Canada. A national sample
consisting of more than 17,000 subjects was interviewed
in 1994/95, then re-interviewed in 1996/97 in the
longitudinal component. The 1996/97 data collection
cycle also included many “buy-ins” from specific
provinces, such that the cross-sectional data from the
1996/97 cycle included over 80,000 subjects (a minority
of whom were subjects being followed up two years after
the initial data collection). The NPHS interview incorporated a short form version of the Composite International
Diagnostic Interview (CIDI), called the Short Form for
Major Depression.14 This brief predictive instrument
assigns a probability of major depression using a set
of questions adapted from the CIDI by Kessler et al.14
Subjects reporting at least five of nine symptoms
Author References
Scott B Patten, Population Health Investigator, The Alberta Heritage Foundation for Medical Research; and Associate Professor, Department of
Community Health Sciences; and Department of Psychiatry, University of Calgary, 3330 Hospital Drive NW, Calgary, Alberta T2N 4N1;
Fax:. (403) 270-7307; E-mail: [email protected]
6
Chronic Diseases in Canada
Vol 22, No 1
constituting the “A” diagnostic criteria for major
depression in DSM-IV, when at least one of these
symptoms is depressed mood or loss of interest, are
assigned a predictive probability of 90%.
The CIDI short form includes two branched series of
questions. Neither branch is followed when a subject
does not report a two-week period of depressed mood or
a two-week period of loss of interest. Subjects reporting
the former symptom follow one branch and subjects
reporting the latter symptom (but not the former) follow
the other branch. Each branch includes questions about
other depressive symptoms requisite to the DSM-IV
definition of major depression. Each branch also contains
a question about the duration of depression, asking those
subjects whose responses indicated an episode of
depression to report the total number of weeks in the
preceding year that they felt “this way” (the question was
preceded by a summarizing statement referring to the
occurrence of an episode of depression in the preceding
12 months, and incorporating key phrases for reported
symptoms and their duration). Surprisingly brief
durations were reported; they are presented in Figure 1.
sible approaches to primary prevention. However, both
pharmacological and non-pharmacological approaches
to treatment generally take at least 4–8 weeks, and the
NPHS duration data appear to imply that the average
episode would resolve during that time.
The objective of this project was to estimate the mean
duration of major depressive episodes using an alternative approach: by integrating the available incidence and
prevalence data. This provides an alternative perspective
on the duration of these episodes.
Material and Methods
Average duration (weeks)
A retrospective cohort analysis was used to estimate
age- and gender-specific incidence proportions. Using
data collected in 1994/95, all subjects who were under
the age of 12 and all subjects with major depression
(according to the CIDI-SFMD) were excluded. The
remaining subjects were assessed in the 1996/97 data set
in order to estimate the proportion of these subjects who
had newly developed major depression. Standard errors
for these estimates were calculated using a bootstrap
method recommended by Statistics Canada. The bootstrap variance estimates were calculated
using an SAS16 macro developed by
FIGURE 1
Statistics Canada for this purpose. All
Number of weeks depressed in past 52 weeks*
estimates used sampling weights to
adjust for unequal selection probabilities
12
due to the complex survey sampling
strategies employed by Statistics Canada
Men
in this survey.
10
Women
Prevalence estimates were made from
the 1996/97 NPHS data, since the sample
8
size for this data collection cycle was
much greater than that of the 1994/95
6
survey. Again, the data were weighted,
this time using a different set of sampling
4
weights produced specifically for the
relevant data file. Bootstrap methods
2
were used to calculate confidence
intervals for the prevalence estimates.
0
In order to integrate the incidence data
16 à 17
20 à 24
35 à 44
55 à 64
75 +
into a broader epidemiological context,
12 à 14
18 à 19
25 à 34
45 à 54
65 à 74
an incidence-prevalence model was used.
This model represented the relationship
Age
between incidence and prevalence in an
assumed stable population (and in the
* Adapted from: Statistics Canada’s Internet site <http://www.statcan.ca/english/Pgdb/People/Health/health35.htm>.
absence of migration) as an equality
Date of extraction: November 1, 2000
between the inflow of new cases into a
prevalence pool and the outflow from the
Potentially, the brief mean duration of depressive
prevalence pool. The outflow includes all terminations of
episodes has important implications for public health.
disease, including recovery and death. This model was
In clinical settings, the major depressive syndrome is
described by Rothman and Greenland,17 as follows. In a
usually regarded as an indication for treatment. Public
stable population, the inflow of new cases into the pool of
health efforts, such as the American Depression
cases (prevalence pool) equals the outflow from the
Screening Day,15 tend to focus on encouraging people
prevalence pool:
with symptoms suggestive of these disorders to come
(Equation 1)
I ( N − P )∆t = (1 / D )P∆t
forward for treatment. Such secondary and tertiary level
prevention is an obvious strategy in the absence of fea-
2001
7
Where P represents the number of persons in the
population with a disorder, I represents the incidence,
N the total number of persons in the population, ∆t an
interval of time and D the mean duration of the disorder.
This expression may be simplified to:17
P
= ID
N −P
(Equation 2)
Here, the ratio on the left side of the equation is the
prevalence odds, a parameter that approximates
prevalence when a disorder is rare.
For the purpose of this analysis, point prevalence was
considered equivalent to one-month period prevalence.
One-month period prevalence is often considered equivalent to point prevalence in psychiatric epidemiology
because definitions of mental disorders include a requirement not only for the presence of signs and symptoms,
but also for the persistence of these signs and symptoms
over several weeks. For major depression, the symptoms
must persist over at least two weeks. To perform the
calculations using months as the time unit, it was necessary to denominate the incidence as a compatible personmonth rate by estimating a monthly incidence rate that
would be expected to result in the observed incidence
proportion over a one-year period. For the purpose of
generating this approximation, the “exponential formula”
was used:

 12
IP ≈ 1 − exp  − ∑ I k ∆t k 

 k =1
(Equation 3)
Here, IP is the approximation of the incidence
proportion from the longitudinal NPHS data and Ik is a
monthly incidence rate that would result in this incidence
proportion over a 12-month follow-up interval. The latter
estimate is the one suitable for substitution into Equation
2. If episode duration is measured in months, then the Ik
rate (which has the units months-1) will result in a
dimensionless prevalence odds.
The NPHS survey used a 12-month predictor of major
depression prevalence, whereas current prevalence is the
parameter of most relevance to the model under development. It was, therefore, necessary to estimate current
prevalence by combining the NPHS data with supplementary information from the literature. Fortunately,
the literature in this area was found to be strikingly
consistent. Kessler et al.,18 using data from 15–24 yearolds participating in the National Comorbidity Survey
(NCS), reported that the ratio of annual to current (30day) major depression was 12.4% to 5.8%, or
approximately 2:1, this ratio being similar in male (9.0%
to 3.8%) and female (16.1% to 8.0%) subjects. The overall
12-month period prevalence of major depression in the
NCS was 10.3%,19 compared to the 4.9% one-month
period prevalence,1 a ratio also approximating 2:1. The
ratio of annual to current cases in the NCS was similarly
8
Chronic Diseases in Canada
comparable among male (7.7% and 3.8%, respectively)
and female respondents (12.9% and 4.9%, respectively).
Since the NCS is an American study, the ratio of
annual to current major depression was also examined
using published data from an earlier survey conducted in
Edmonton. This study utilized methods resembling those
employed in the Epidemiological Catchment Area (ECA)
studies in the United States. Here, an annual prevalence
of 4.6% and a one-month prevalence of 2.3% were
reported, a 2:1 ratio.11 The ECA study itself reported a
one-year prevalence of (4.2%)20 and a one-month period
prevalence of 2.2%,4 also closely approximating 2:1.
Based on these very consistent findings, the ratio of
annual to current major depression in this analysis was
taken to be 2:1 for both men and women. Hence, prior to
estimation using equation 2, the annual prevalence from
the NPHS was multiplied by 0.5 to generate an approximation of point (30-day) prevalence.
Results
The 1994/95 NPHS had a sample size of 17,626. The
longitudinal data file included 15,670 subjects providing
follow-up data. The current analysis excluded subjects
who were under the age of 12 (n = 1908), subjects who
had major depression at baseline (n = 781) or who did
not provide valid data on the major depression predictor
either at the baseline interview, the follow-up interview
or both (n = 691). As such, the current analysis of
incidence was based on 12,290 subjects. The prevalence
estimates were based on 70,538 subjects over the age of
12 in the 1996/97 cross-sectional component of the
NPHS. The sample included 73,402 subjects within this
age group, but 2,864 (3.9%) who did not provide a valid
rating on the CIDI Short Form were excluded.
The age- and gender-specific incidence of major
depression, along with 95% confidence intervals, are
presented in Table 1. The observed pattern of incidence
generally resembled that of prevalence, as depicted in
Table 2. Among female subjects, incidence and prevalence
TABLE 1
Age and gender-specific one-year incidence
proportions and rates for major depression
Men
Women
Annual
incidence
proportion
95%
confidence
interval
Estimated
incidence
rate
(month-1)
Age 12–24
0.029
0.014–0.043
2.45e-03
Age 25–44
0.033
0.020–0.047
2.80e-03
Age 45–64
0.018
0.007–0.029
1.51e-03
Age ≥ 65
0.018
0.007–0.028
1.51e-03
Age 12–24
0.071
0.051–0.091
6.14e-03
Age 25–44
0.045
0.034–0.057
3.84e-03
Age 45–64
0.041
0.025–0.057
3.49e-03
Age ≥ 65
0.013
0.006–0.021
1.09e-03
Vol 22, No 1
TABLE 2
Age and gender-specific one-year prevalence
proportions and estimated current prevalence
rates for major depression
Men
Women
Estimated
current (30d)
prevalence
Estimate
incidence
rate
(month-1)
Estimated
mean
duration of
episodes*
(months)
Age 12–24
1.30%
2.45e-03
5.4
Age 25–44
1.75%
2.80e-03
6.4
1.30%
Age 45–64
1.30%
1.51e-03
8.7
1.7%
0.9–2.5M
0.85%
Age ≥ 65
0.85%
1.51e-03
5.7
6.7%
5.4–8.0
3.35%
Age 12–24
3.35%
6.14e-03
5.6
Age 25–44
6.8%
5.9–7.7
3.40%
Age 25–44
3.40%
3.84e-03
9.2
Age 45–64
5.0%
4.1–5.8
2.50%
Age 45–64
2.50%
3.49e-03
7.4
Age ≥ 65
1.6%
1.0–2.2
0.80%
Age ≥ 65
0.80%
1.09e-03
7.4
Annual
prevalence
proportion
(%)
95%
confidence
interval
Estimated
current
prevalence
proportion*
Age 12–24
2.6%
1.9–3.2
1.30%
Age 25–44
3.5%
2.8–4.2
1.75%
Age 45–64
2.6%
2.0–3.2
Age ≥ 65
Age 12–24
* 50% of the annual prevalence proportion; see text.
tended to be higher in younger women and to decline
with advancing age. In males, the incidence rates tended
to be slightly higher in middle-aged than in younger
subjects. In the subjects aged 75 years and older, the
incidence and prevalence increased slightly in both
genders.
The prevalence-incidence model embodied in
equations 1 and 2 relates incidence and the prevalence
proportion (approximately, prevalence) to the mean
duration of disease irrespective of whether the disease is
terminated by recovery or death. Completing these
calculations for each age and gender group yielded the
data presented in Table 3. There is no evidence of gender
differences in model-based estimates of disease duration,
nor is there a pronounced trend towards such differences
in association with age. However, the duration of the
episodes, as predicted by the incidence-prevalence model,
appears to be more brief in the youngest age group.
Discussion
The development of this model incorporated a variety
of assumptions. Some assumptions are involved in the
estimation of incidence rates using an exponential
equation. The use of this equation involves assumptions
that the population is closed, that there are no competing
risks (often, the equation is used to estimate mortality
rates) and that the number of events is small relative to
the number at risk.17 Another assumption was that the
12-month incidence proportion was being measured by
the CIDI Short Form. The CIDI Short Form questions
are designed to cover a 12-month period, but since the
population at risk was identified as those not having
major depression during the two previous years it is
possible that some of the episodes had their onset more
than one year prior to the interview. There may be
additional measurement concerns related to the CIDI
Short Form. This instrument does not include many
2001
TABLE 3
Estimated duration of
major depressive episodes
Men
Women
* Based on equation 2, the mean duration of episodes is calculated as: P / I*(1-P).
of the “clinical significance” probes that are contained
in the full CIDI and may, therefore, be less specific in
its measurement properties. Finally, the relationship
between annual and current prevalence had to rely on
data from the literature. Since none of these assumptions
can be definitely shown to hold true, the model presented
here should be regarded as a heuristic one. It provides a
description of the relationship between major depression
incidence and prevalence in Canada using the best
available data. Furthermore, for a recurrent condition
such as major depression, differences in the approach to
modelling may be fruitful. For example, the “lifetime
sick day proportion” 21 has been proposed as a means
of modelling the relationship between incidence and
prevalence in episodic conditions accounting both for
episode duration and number of episodes. However, such
models will also generally be subject to various approximations and assumptions.21
Despite these provisos, the model presented here
appears to provide a description of incidence-prevalence
relationships that are consistent with other available data.
The NPHS incidence rates are approximately consistent
with those reported elsewhere in the literature. In making
such comparisons, it should be emphasized that the NPHS
measures major depressive episodes, not disorders, so
that the incidence rates are higher than those studies
evaluating the first occurrence of depressive episodes
(these first episodes being considered the first onset of an
episodic depressive disorder in some studies). One recent
German study reported a 20-month incidence of major
depression in an adolescent sample,22 3.7% in male and
7.5% in female respondents. These authors estimated
that the 12-month incidence in their sample would have
been approximately 4.3%. Another prospective study of
high school-aged adolescents reported annual incidence
rates for major depression of 10.4% for female and 4.8%
for male subjects.23 One study of the incidence of major
9
depression in a very elderly (mean age 85) sample
reported an annual incidence of 1.4%.24 All of these
findings are very consistent with the data presented here.
The six-month follow-up of the Epidemiological
Catchment Area survey in New Haven reported a 4.3%
six-month incidence.25 As a six-month incidence estimate
(and from a study of older subjects), the New Haven
figure is higher than that of the German study, which is a
surprising result. The New Haven estimate may,
however, be an overestimate as a result of measurement
problems arising when the Diagnostic Interview
Schedule is used in follow-up studies.26 An advantage of
the data source for this project (the NPHS) over prior
studies is that it provides an estimate of incidence and
prevalence based on a comparable measure.
The predicted episode duration seems consistent with
previous reports. Some community studies evaluating
the duration of depressive episodes have reported
comparable durations of episodes. For example, the
mean duration of major depression in the Lewinsohn et
al. adolescent sample was 23.6 weeks.23 However, other
studies have reported more brief average episode
durations. For example, Rao et al.27 reported an average
episode duration of 10.3 weeks in a sample consisting of
17- and 18-year old women. Data from the Baltimore
ECA follow-up study found that the median episode
duration in that sample was only 12 weeks28 and
Kendler’s follow-up of a community sample of female
twin pairs reported a median time to recovery of only
eight weeks.29 These values for median duration are brief
relative to the model-based estimate of mean episode
duration from this study. This may relate to a right skew
in the distribution of episode durations. Community
studies have confirmed that a proportion of persons with
major depression experience very protracted episodes.
One analysis of ECA follow-up data found that 23.6%
of subjects with major depression at baseline remained
depressed after one year.30 A study following a series of
78 clinical patients with major depression (probably with
more severe and complicated disorders than those in the
community studies) found that only 34 (48.6%) had
recovered after one year. These findings suggest that
most episodes of major depression in the community are
brief (hence, having median durations less than three
months), but that a proportion of persons experience very
prolonged episodes, causing the mean duration to be
greater than the median. In Kendler et al.’s study, mean
episode duration was twice as long as median duration.29
Any study relying on a finite and relatively brief followup interval will truncate the observed duration of more
prolonged episodes.
Some of the differences between the duration of episodes reported by NPHS subjects and the mean durations
estimated indirectly using incidence and prevalence data
may simply reflect the concept of 12-month period
prevalence, as measured in the NPHS. A person with the
new onset of a protracted episode in the few weeks
preceding the NPHS interview would have their duration
of depression recorded only as the interval between onset
10
Chronic Diseases in Canada
and interview. This would also occur for a protracted
episode that resolved a few weeks into the year preceding
the NPHS interview. Finally, whereas some episodes of
this condition can last for years, the maximum value that
could be recorded using the questions employed in the
NPHS was 52 weeks. Hence, the distribution of episode
durations might have been substantially truncated.
Elements of consistency between these results and
published findings have been summarized in the
preceding paragraphs. It should be emphasized that the
validity of a result is not confirmed by its consistency
with other findings. Nevertheless, the development of
dynamic descriptions of major depression epidemiology
will need to be guided by principles of consistency with
the existing literature of epidemiological studies.
The NPHS data appear to suggest that a substantial
proportion of persons experiencing the major depressive
syndrome will have a relatively brief disturbance. In
itself, this has important public health implications. For
example, a large number of false positives are likely to
emerge from screening efforts if instruments like the
CIDI short form are used to detect people in need of
clinical intervention. However, these results serve to
emphasize that the duration data from the NPHS
interview do not necessarily reflect the average
experience of persons with major depression. This
appears to be a heterogenous condition, characterized
both by brief and protracted episodes.
References
1. Blazer DG, Kessler RC, McGonagle KA, Swartz MS. The
prevalence and distribution of major depression in a
national community sample: the National Comorbidity
Survey. Am J Psychiatry 1994;151:979–986.
2. Isometsa ET, Aro H. Depression in Finland: a computer
assisted telephone interview study. Acta Psychiatr Scand
1997;96:122–128.
3. Ohayon MM, Priest RG, Guilleminault C, Caulet M. The
prevalence of depressive disorders in the United Kingdom.
Biol Psychiatry 1999;45:300–307.
4. Regier DA, Boyd JH, Burke JD, Myers JK, Kramer M,
Robins LN et al. One-month prevalence of mental
disorders in the United States. Based on five epidemiological catchment area sites. Arch Gen Psychiatry
1988;45:977–986.
5. Szadoczky E, Papp Zs, Vitrai J, Rihmer Z, Furedi J. The
prevalence of major depressive and bipolar disorders in
Hungary. Results from a national epidemiological survey.
J Affect Disord 1998;50:153–162.
6. Weissman MM, Myers JK. Affective disorders in a US
urban community. Arch Gen Psychiatry 1978;
35:1304–1311.
7. Weissman MM, Bland RC, Canino GJ, Faravelli C,
Greenwald S, Hwu HG et al. Cross-national epidemiology
of major depression and bipolar disorder. JAMA
1996;276(4):293–299.
8. Beaudet MP. Depression. Health Reports 1996;7(4):11–24.
9. Stephens T, Dulberg C, Joubert N. Mental health of the
Canadian population: a comprehensive analysis. Chronic
Dis Can 1999;20:118–126.
Vol 22, No 1
10. Offord DR, Boyle MH, Campbell D, Goering P, Lin E,
Wong M et al. One year prevalence of psychiatric disorder
in Ontarians 15 to 64 years of age. Can J Psychiatry
1996;4:559–563.
11. Bland RC, Newman SC, Orn H. Period prevalence of
psychiatric disorders in Edmonton. Acta Psychiatr Scand
1988;Suppl 338:33–42.
12. De Marco RR. The epidemiology of major depression:
implications of occurrence, recurrence, and stress in a
Canadian community sample. Can J Psychiatry
2000;45:67–74.
13. Beaudet MP. Psychological health – depression. Health
Reports 1999;11:63–75.
14. Kessler RC, Andrews G, Mroczek D, Ustun B, Wittchen
HU. The World Health Organization Composite
International Diagnostic Interview Short-Form (CIDI-SF).
International Journal of Methods of Psychiatric Research
1998;7:181–195.
15. Jacobs DG. National depression screening day: educating
the public, reaching those in need of treatment, and
broadening professional understanding. Harvard Review
of Psychiatry 1995;3:156–159.
16. SAS/STAT User’s Guide. Cary, NC: The SAS Institute,
1994.
17. Rothman KJ, Greenland S. Measures of disease frequency.
In: Rothman KJ, Greenland S, editors. Modern
epidemiology. Philadelphia: Lippincott-Raven,
1998:29–64.
18. Kessler RC, Walters EE. Epidemiology of DSM-III-R
major depression and minor depression among adolescents
and young adults in the National Comorbidity Survey.
Depression and Anxiety 1998;7:3–14.
19. Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes
M, Eshleman S et al. Lifetime and 12-month prevalence of
DSM-III-R psychiatric disorders in the United States.
Results from the National Comorbidity Survey. Arch Gen
Psychiatry 1994;51:8–19.
20. Regier DA, Kaelber CT, Rae DS, Farmer ME, Knauper B,
Kessler RC et al. Limitations of diagnostic criteria and
2001
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
assessment instruments for mental disorders. Arch Gen
Psychiatry 1998;55:109–115.
Von Korff M, Parker RD. The dynamics of the prevalence
of chronic episodic disease. J Chron Dis 1980; 33:79–85.
Oldehinkel AJ, Wittchen H-U, Schuster P. Prevalence,
20-month incidence and outcome of unipolar depressive
disorders in a community sample of adolescents. Psychol
Med 1999;29:655–668.
Lewinsohn PM, Hops H, Roberts RE, Seeley JR, Andrews
JA. Adolescent psychopathology: I. Prevalence and
incidence of depression and other DSM-III-R disorders in
high school students. J Abn Psychol 1993; 102:133–144.
Forsell Y, Winblad B. Incidence of major depression in a
very elderly population. Int J Geriatr Psychiatry 1999;
14:368–272.
Bruce ML, Takeuchi DT, Leaf PJ. Poverty and psychiatric
status. Longitudinal evidence from the New Haven
Epidemiological Catchment Area Study. Arch Gen
Psychiatry 1991;48:470–474.
Newman SC, Bland RC. Incidence of mental disorders in
Edmonton: estimates of rates and methodological issues.
J Psychiatr Res 1998;332:273–282.
Rao U, Hammen C, Daley SE. Continuity of depression
during the transition to adulthood: a five-year longitudinal
study of young women. J Am Acad Child Adolesc
Psychiatry 1999;38:908–915.
Eaton WW, Anthony JC, Gallo J, Cai G, Tien A,
Romanoski A et al. Natural history of diagnostic interview
schedule/DSM-IV major depression. The Baltimore
epidemiological catchment area follow-up. Arch Gen
Psychiatry 1997;54:993–999.
Kendler KS, Walters EE, Kessler RC. The prediction of
length of major depressive episodes: results from an
epidemiological sample of female twins. Psychol Med
1997;27:107–117.
Sargeant JK, Bruce ML, Florio LP, Weissman MM.
Factors associated with 1-year outcome of major
depression in the community. Arch Gen Psychiatry
1990;47:519–526. ■
11
Emigration Patterns of Cancer Cases in Alberta, Canada
Juanita Hatcher and Marilou Hervas
Abstract
Cancer registries are a unique source of data for population-based analysis of survival of
cancer cases, but information on current vital status is essential. This paper describes a
method to determine the last known vital status of cases and the emigration pattern of cancer
cases diagnosed in Alberta. Data from the Alberta Cancer Registry (ACR) for the years
1985–1993 (83,446 cases) were linked to the Alberta Health Care Insurance Plan (AHCIP)
registration file to identify cases that had left the province and the date they emigrated.
Ninety-nine percent of the ACR cases linked correctly to the AHCIP registration file. Three
percent of cases had left Alberta by March 1998. For the first five years of follow-up between
0.6% and 0.8% of cases alive at the beginning of each year of follow-up left the province in
the succeeding year. Seven percent of those diagnosed under 45 years of age left the province
compared to less than 2% of those aged 65 and over. There was no difference in emigration
patterns between the sexes. The cancer sites with good prognosis tended to have the highest
proportion of emigrants.
Key words: cancer; emigration
Introduction
Data
The primary function of cancer registries is to identify
and register all incident cancer cases occurring within
their jurisdiction and to record information related to the
death of each cancer case.1 Cancer registries are thus a
unique source of data for analyzing the survival of the
population of cancer cases.2 As many cancer registries
do not actively follow up the incident cases, however,
the residency and current vital status of those not known
to be dead are often unknown, as is the impact of
emigration on survival.
In Alberta, Canada, the problems caused by lack of
active follow-up were overcome by linking the Alberta
Cancer Registry (ACR) to the Alberta Health Care
Insurance Plan (AHCIP) registration file maintained by
Alberta Health and Wellness (AHW). This enabled the
ACR to identify those cancer cases that had left the
province and the date that they left. This date would be
used as a censoring date in survival analysis, permitting
the determination of the extent of emigration of cancer
cases from Alberta and the potential impact on survival
analysis.
This paper describes the pattern of emigration of
cancer cases diagnosed in Alberta.
AHW maintains a historical database of all residents
of the province of Alberta who are/were registered in
AHCIP, which insures all required medical care available
in the provinces. The contract holder pays the premiums,
and benefits are provided for all members of the immediate family. All residents of Alberta, except serving
members of the Royal Canadian Mounted Police and the
Canadian Military, inmates of federal penitentiaries and
Status Indians, are eligible to register in AHCIP. The
federal government pays the AHCIP premiums for Status
Indians and other exceptions who are included in the
AHCIP registration file. Only 200–300 of the
approximately 3 million who are eligible to register
choose not to. The AHCIP registration file includes data
on more than 99% of the Alberta population.
The AHCIP registration data include a unique
personal identifier, the Personal Health Number (PHN)
and/or the AHCIP number, surname, initials, date of
birth, gender and postal code, as well as the date and
reason for any changes in coverage by AHCIP. Prior
to 1994, the individual identifier (AHCIP number)
consisted of an eight-digit base number to identify the
contract, and a three-digit individual identifier. If a
person changed the contract under which he or she was
covered, by, for example, leaving the family home to
Author References
Juanita Hatcher, Division of Epidemiology, Prevention and Screening, Alberta Cancer Board, Edmonton, Alberta
Marilou Hervas, Division of Epidemiology, Prevention and Screening, Alberta Cancer Board; and EPICORE Center, University of Alberta, Edmonton, Alberta
Correspondence: Dr. Juanita Hatcher, Division of Epidemiology, Prevention and Screening, Alberta Cancer Board, 11560 University Avenue,
Edmonton, Alberta T6G 1Z2; Fax: (780) 432-8645; E-mail: [email protected]
12
Chronic Diseases in Canada
Vol 22, No 1
marry, his or her AHCIP number would change. The
PHN, introduced in 1994, is unique to an individual and
remains the same for life. AHW has assigned a PHN to
people who were initially registered before 1994, and a
conversion file exists to link pre-1994 AHCIP numbers
to the new PHNs. AHW is usually informed of people
who leave the province by the medical plan of the
province to which the person has emigrated. The AHCIP
file is linked to the vital statistics file on a weekly basis
to identify those who have died in the province. The
estate of a dead person may also notify AHW when it
receives the bill for the premiums.
The ACR is mandated under the Alberta Cancer
Program Act to capture information on every incident of
invasive cancer diagnosed in Alberta.3 Pathology reports
and death information from the Alberta Registries vital
statistics file are the two main sources for identifying
incident cases of cancer in Alberta. The ACR includes
identifying information, (PHN, AHCIP number, current
surname, previous surnames, first names, date of birth,
gender and postal code); information on the incident
tumour; and for those cases known to have died, date and
cause of death. The Alberta Registries vital statistics file
also provides information on cases that have died in
Alberta. The ACR would not generally be informed of
the deaths in other provinces. There has been no followup information on cases that are not known to have died.
Methods
All invasive cancer cases diagnosed in Alberta
between January 1, 1985, and December 31, 1993,
inclusive were identified from the ACR. A file
containing PHN, AHCIP number, surname, initials,
gender and date of birth for each individual diagnosed
with cancer was submitted to AHW. This file included
records for residents and non-residents of Alberta. A
hierarchical deterministic linkage strategy was used to
link the ACR to the AHCIP registration file, linking on
the AHCIP number, the AHCIP base number and
surname, gender and month and year of birth (identical).
All links were checked using the other identifying
information common to the two files.
Discrepancies between the ACR and the AHCIP
records were investigated further by reviewing the
patient’s chart. Errors in the ACR were corrected, and
AHW was notified of suspected errors in its data set. The
agreement in the vital statistics of the linked cancer cases
in the two files was also checked. Where the ACR had
recorded that a case had died, but the AHCIP record
indicated the case was alive, the death information held
by the ACR was confirmed. In cases where AHCIP
indicated the case had died but the ACR had no record of
death, the case was deemed to be dead for subsequent
analysis and the death date taken as that recorded by
AHCIP. The ACR also had death information on some
cases that AHCIP deemed had left the province.
Cases for which the first invasive primary (excluding
non-melanoma skin cancer) occurred between 1985 and
2001
1993 and which were resident in Alberta at the time of
diagnosis were identified from the linked cohort. The
percentage of cases that had left the province as determined from AHCIP was examined for each cancer site.
Crude, age specific and age-standardized percentages
were estimated. The standard population used was the
total population of cancer cases used in the analysis. The
percentage emigrating each year up to five years past
diagnosis was estimated using the revised ACR vital
status definitions. Person-years at risk were calculated as
the time between diagnosis and date of last follow-up,
death, emigration or for those still alive and resident in
Alberta, as of March 31, 1998. All follow-up periods
were terminated at five years to standardize the potential
length of follow-up. The reduction in person-years at
risk produced by censoring emigrants at the time of
emigration rather than as of March 31, 1998, was
investigated.
Results
Of the 83,446 cases that were resident in Alberta for
at least one diagnosis between 1985 and 1993, 82,466
cases (98.8%) linked correctly to the AHCIP registration
file. There was no difference in linkage rate between
those who were registered as dead (98.8%) and those
who were not registered as dead (98.8%).
The following results are presented for the 82,466
cases who were residents in Alberta at the time of
diagnosis between 1985 to 1993, and who linked
successfully with the AHCIP registration file (Table 1).
Of the 45,925 cases that were registered as dead on the
ACR, 97.3% were also registered as dead on the AHCIP
registration file, 1.8% were registered as still alive and
0.9% were registered as having left the province. Of the
36,541 cases that were not registered as dead on the
ACR, 1.3% were registered as dead on the AHCIP
registration file and 5.6% had left the province. Of the
44,689 cases who were confirmed as dead on both the
ACR and the AHCIP registration file, 574 (1.3%) of the
dates of death do not agree between the two files. In 95%
of these cases the AHCIP registration death date is later
than that of the ACR. The discrepancies can be quite
large: only 31% have differences of less than one year.
The following results are presented for the 76,164
cases of a first primary invasive cancer (excluding nonmelanoma skin cancer) diagnosed between 1985 and
TABLE 1
Agreement between linked records for vital
status on ACR and AHCIP files
ACR
vital
status
Dead
AH vital status
Dead
44,689 (97.3%)
Alive
Left Alberta
Total
836 (1.8%)
400 (0.9%)
Alive
478 (1.3%) 34,012 (93.1%)
2,051 (5.6%)
45,925 (55.7%)
36,541 (44.3%)
Total
45,167 (54.8%) 34,848 (42.3%)
2,451 (3.0%)
82,466
13
TABLE 2
Vital status of cancer cases diagnosed by cancer site in Alberta 1985–1993 as determined
from 1998 AHCIP files
Dead
(%)
[adja %]
Alive in
AB
(%)
[adja%]
Prostate
4,322
46.4
40.5
4,816
51.7
57.6
Female breast
3,383
31.3
34.1
7,007
64.8
Lung
8,634
87.9
85.7
1,029
Colorectal
4,999
57.0
53.9
3,531
428
19.1
29.7
NHLb
1,374
55.0
57.6
Leukemia
1,392
58.5
528
23.7
1,191
Kidney
Testis
Melanoma skin
Uterus
Bladder
Cervix uteri
(%)
[adja %]
177
1.9
1.9
9,315
62.4
418
3.9
3.5
10,808
10.5
12.4
161
1.6
1.9
9,824
40.3
42.9
235
2.7
3.2
8,765
1,668
74.5
66.2
142
6.3
4.1
2,238
1,039
41.6
39.3
87
3.5
3.1
2,500
61.3
911
38.3
35.9
78
3.3
2.8
2,381
26.3
1,627
73.1
70.4
72
3.2
3.3
2,227
44.6
40.3
1,368
51.3
54.8
109
4.1
4.9
2,668
940
48.1
50.4
929
47.5
45.4
86
4.4
4.1
1,955
34
5.1
40.8
571
85.9
56.0
60
9.0
3.2
665
1,246
Left AB
Total
394
31.6
48.2
771
61.9
48.0
81
6.5
3.7
1,968
95.3
92.9
79
3.8
6.2
19
0.9
1.0
2,066
Other
11,161
63.8
67.0
5,812
33.2
30.4
533
3.0
2.7
17,506
Total
40,748
54.9
31,158
42.0
2,258
3.0
Pancreas
a
b
74,164
Age adjusted to the total cohort age distribution
Non-Hodgkin’s Lymphoma
1993 while resident in Alberta. On average, 3.0% of the
cases had left Alberta by 1998. (Table 2) This is
somewhat lower than the age-standardized five-year
emigration rate from Alberta between 1991 and 1996 of
3.5%.4 For the cohort of cases diagnosed between 1985
and 1993, and followed until March 1998, the cancer
sites with the highest proportion of emigrants were
testicular cancer (9.0%), cervical cancer (6.5%) and
melanoma skin cancer (6.3%). The cancer sites with the
lowest proportion of emigrants were prostate cancer
(1.9%), lung cancer (1.6%) and pancreatic cancer (0.9%)
(Table 2). Age standardization removes these differences
except for those cancers with short survival rates, lung
and pancreatic cancer, and for prostate cancer. In general,
younger cancer cases were more likely to emigrate than
older ones, but there is little difference in emigration
rates among cancer sites. (Table 3).
For the first five years of follow-up, between 0.6%
and 0.8% of all cases alive at the beginning of each year
of follow-up left the province during the succeeding
year. (Table 4). This pattern continues after five years of
follow-up. There are no marked differences in the
patterns of emigration among the cancer sites. There are
no differences in emigration patterns between the sexes.
The person-years at risk for the first five years of
follow-up is reduced on overall by 1.7% of the total
person-years at risk if all persons not known to be dead
are assumed to be alive. The largest effect (3.5%
reduction) is seen for cancer of the testis, and the
smallest (0.8% reduction) for cancer of the prostate.
14
Chronic Diseases in Canada
Discussion
Survival analysis requires the follow-up of all incident
cancer cases included in the analysis so that their current
vital status is known. The ability of cancer registries to
follow up cases is dependent on the health care system in
which the registry operates, and the registration practice
of each registry.5 The main method of determining the
death information for each cancer case is to link to the
death certificate data for the jurisdiction of the registry.
The ability to determine whether a case has left the
jurisdiction of the registry and thus is lost to follow-up
varies among registries. For the Scandinavian registries,
complete follow-up is possible due to the existence of a
unique lifetime identifier for each citizen. Thus date of
emigration and date of death are known for all cancer
cases.5 In Saarland, Germany, follow-up is largely
passive, due to restrictive legislation. It is possible to
determine the death information for cases that die in
Saarland, but not whether a case has emigrated.5 In
Canada, each of the provinces and territories operates its
own registry and submits data to the national Canadian
Cancer Registry. Most of these registries link their data
to provincial vital statistics death information to determine the death information for those cases that die
within the relevant province or territory. The Canadian
Cancer Registry is linked to the National Death file for
determining deaths among all cases that occurred
anywhere in Canada. At the time of this study, the
national death linkages have not been completed for the
more recent years.
Vol 22, No 1
TABLE 3
Distribution by age and site of cancer cases who left Alberta after diagnosis 1985–1993
Age
Site
0–44 yrs
45–64 yrs
65–74 yrs
75–84 yrs
85+ yrs
Total
Left/Diagnosed % Left/Diagnosed % Left/Diagnosed % Left/Diagnosed % Left/Diagnosed % Left/Diagnosed
Prostate
Female breast
0/10
0.0
60/1,924
3.1
69/3,892
1.8
38/2,772
1.4
10/717
1.4
177/9,315
105/1,892
5.5
212/4,751
4.5
58/2,355
2.5
32/1,379
2.3
11/431
2.6
418/10,808
Lung
12/317
3.8
83/3,670
2.3
35/3,470
1.0
27/2,000
1.4
4/367
1.1
161/9,824
Colorectal
29/404
7.2
108/2,871
3.8
49/2,593
1.9
33/2,129
1.6
16/768
2.1
235/8,765
Melanoma skin
88/928
9.5
47/776
6.1
4/303
1.3
3/166
1.8
0/65
0.0
142/2,238
NHLa
40/493
8.0
25/910
2.7
13/578
2.2
8/413
1.9
1/106
0.9
87/2,500
Leukemia
40/640
6.3
16/654
2.4
12/524
2.3
9/399
2.3
1/164
.6
78/2,381
Uterus
8/155
5.2
39/1,005
3.9
13/675
1.9
11/332
3.3
1/60
1.7
72/2,227
Bladder
23/172
13.4
52/876
5.9
22/832
2.6
10/611
1.6
2/177
1.1
109/2,668
Kidney
22/265
8.3
45/795
5.7
13/520
2.5
6/294
2.0
0/81
0.0
86/1,955
Testis
56/584
9.6
4/73
5.5
0/7
0.0
0/1
0.0
0/0
Cervix uteri
58/659
8.8
20/350
5.7
2/143
1.4
0/67
0.0
1/27
3.7
81/1,246
Pancreas
60/665
1/66
1.5
3/614
0.5
7/646
1.1
7/511
1.4
1/229
0.4
19/2,066
Other
262/3,564
7.4
169/5,761
2.9
66/4,174
1.6
25/2,940
0.9
11/1,067
1.0
533/17,506
Total
744/10,149
7.3
883/25,030
3.5
363/20,712
1.8
209/14,014
1.5
59/4,259
1.4
2,258/74,164
a
Non-Hodgkin’s Lymphoma
TABLE 4
Distribution of time from diagnosis to leaving Alberta, by cancer site,
for cancer cases diagnosed in Alberta 1985–1993
0–<1 yr
Interval
1–<2 yrs
2–<3 yrs
3–<4 yrs
4–<5 yrs
# alive
# alive
# alive
# alive
# alive
%
#
at
%
#
at
%
#
at
%
#
at
%
#
at
begin- leaving leaving begin- leaving leaving begin- leaving leaving begin- leaving leaving begin- leaving leaving
in
in
ning of
in
in
ning of
in
in
ning of
in
in
ning of
in
in
ning of
interval interval interval interval interval interval interval interval interval interval interval interval interval interval interval
Site
Prostate
Female breast
9,309
31
0.3
8,302
28
0.3
7,402
30
0.4
6,014
26
0.4
4,731
17
0.4
10,801
51
0.5
10,161
73
0.7
9,415
66
0.7
8,084
52
0.6
6,702
37
0.6
Lung
9,821
71
0.7
3,367
28
0.8
1,929
22
1.1
1,315
9
0.7
960
13
1.4
Colorectal
8,763
45
0.5
6,477
51
0.8
5,385
42
0.8
4,292
28
0.7
3,412
13
0.4
Melanoma skin
2,235
19
0.9
2,138
29
1.4
2,022
31
1.5
1,741
16
0.9
1,441
13
0.9
NHLa
2,498
19
0.8
1,788
17
1.0
1,518
18
1.2
1,232
6
0.5
999
8
0.8
Leukemia
2,380
18
0.8
1,637
16
1.0
1,395
15
1.1
1,152
8
0.7
935
7
0.7
Uterus
2,227
12
0.5
2,050
18
0.9
1,932
9
0.5
1,704
5
0.3
1,480
7
0.5
Bladder
2,667
16
0.6
2,252
21
0.9
1,994
19
1.0
1,763
11
0.6
1,567
9
0.6
Kidney
1,954
17
0.9
1,401
13
0.9
1,247
15
1.2
1,047
10
1.0
862
13
1.5
665
9
1.4
640
14
2.2
615
5
0.8
568
7
1.2
501
3
0.6
Cervix uteri
1,244
14
1.1
1,089
11
1.0
951
16
1.7
821
10
1.2
707
9
1.3
Pancreas
2,066
17
0.8
294
2
0.7
123
0
0.0
77
0
0.0
52
0
0.0
Other
17,499
142
0.8
10,221
102
1.0
8,187
83
1.0
6,728
46
0.7
5,515
47
0.9
Total
74,129
481
0.6
51,817
423
0.8
44,115
371
0.8
36,538
234
0.6
29,864
196
0.7
Testis
a
Non-Hodgkins Lymphoma
2001
15
TABLE 5
Person-years of follow-up with and without
censoring at time of emigration for cancer
cases diagnosed in Alberta, 1985–1993
Total person-years of
follow-up
Not
censored
Censored
Prostate
33,099
32,829
270
0.82
Female breast
43,165
42,542
623
1.46
Lung
11,852
11,673
179
1.54
Colorectal
25,332
24,988
344
1.38
Melanoma skin
9,339
9,078
262
2.88
NHL
7,248
7,079
169
2.38
Leukemia
6,630
6,523
108
1.65
Site
Difference
%
Uterus
9,026
8,898
129
1.45
Bladder
9,726
9,550
176
1.84
Kidney
5,945
5,801
144
2.48
Testis
2,982
2,883
100
3.47
Cervix uteri
4,639
4,497
142
3.15
Pancreas
1,218
1,198
20
1.65
Other
41,137
40,218
919
2.28
Total
211,338
207,755
3,583
1.72
To overcome this problem, the ACR sought other
solutions to determine the current residence and vital
status of those cancer cases not known to have died.
The AHCIP registration file, which AHW maintains,
provided ACR with the required information. Deterministic linkages were used in preference to probabilistic
linkages because of the availability of the AHCIP
number. Evaluation of the linkage confirmed that it
provided a high degree of accuracy (98.8% among cases
resident in Alberta). However, this information is subject
to the problems inherent in using administrative data in
an application for which such data were not originally
planned.6 Some of these problems have been identified,
but not resolved, in this study. There are small
percentages of cases in which the vital status is different
in the two files (1.6%), in which the recorded residency
patterns do not seem to concur with the stated residency
in the ACR (1.6%), and in which the death dates do not
agree (1.7%). The discrepancy in vital status could be
explained by the potential delay in registering deaths on
the AHCIP registration file for those cases that are dead
in the ACR data and alive in the AHCIP data. For those
cases that are alive in the ACR data and dead in the
AHCIP data, the absence of the PHN on the ACR record
may have prevented appropriate linkage to the Alberta
Registries vital statistics death data, or these may be
Alberta residents who died outside the province. These
cases would not be included in the Alberta Registries
vital statistics computer files. However, AHW may have
16
Chronic Diseases in Canada
been notified of the death when payment to AHCIP
ceased.
The main aim of the project was to identify the
residency and vital status of those cases that were not
identified as dead so that cases leaving the province
could be censored at the date of leaving in any survival
analysis. The linkage has shown that cancer cases do in
fact move out of the province, although they are less
likely to migrate than the general population.4 The
probability of their emigration depends on the site of
their cancer, their age, and the time since diagnosis.
As would be expected, those who are diagnosed at a
younger age, and are diagnosed with a site for which
there is a good prognosis, are more likely to move out
of the province. The low emigration rate (1.9 %) for
prostate cancer, which has a relatively good prognosis,
may in part be explained by the advanced age at
diagnosis. Both the good prognosis and the younger age
at diagnosis can explain the high emigration rates for
cancer of the testis, cervix and melanoma. Among those
cases that migrate, the proportion of live cases that
migrate on an annual basis does not vary appreciably
among the cancer sites.
The implications of these findings on the results of
survival analysis may be sizeable depending on the
reasons that the cases migrate. If the cases that migrate
were the ones for whom the prognosis is good compared
to others with the same diagnosis, the survival rates
would tend to be underestimated. However, if the ones
who leave have poor prognoses compared to others with
the same diagnosis then the survival rates would be
overestimated. The cause-specific survival of foreign
residents of Geneva diagnosed with cancer is superior
than that of native residents, which may be due to a
combination of the healthy immigrant bias and/or the
repatriation of those cases with poor prognosis
(unhealthy emigrant bias).7 The Alberta data show that
the proportion of cases migrating from Alberta tends to
be related to the overall prognosis of the diagnosis. The
ACR has not collected staging information, and therefore
is not able to address the issue of the prognosis of those
cases who emigrate relative to those who remain in
Alberta. However, the similarity of the distribution of
time to emigration among the cancer sites for those who
migrate would not indicate any systematic emigration
patterns within site, based on prognosis.
Censoring of the emigrants at the time they leave the
province decreases the overall person-years at risk by
1.7%. Although this figure is small, it may have a
marked effect on survival estimates, particularly where
survival is short or emigration is high. In the Eurocare II
study, active follow up of lung cancer cases not known
to be dead after five years resulted in decreases in fiveyear survival of up to 2.5%.5
Alberta is able to identify the cases who emigrate
because the AHCIP registration file is updated regularly,
in part to ensure that appropriate health care premiums
are paid. In other provinces, the health care registration
Vol 22, No 1
file may either not be available to the cancer registry or
may not be sufficiently up to date. Thus the national
death clearance which is currently being undertaken is
essential for provincial cancer registries to improve their
survival analysis. However, this death clearance will not
identify cases who die outside of Canada and thus
linkage with the AHCIP registration may still be of
value.
The economy in Alberta is largely dependent on the
oil and natural gas industry, which fluctuates with world
economic conditions. The economy tends to drive the
emigration pattern for the younger age group, while that
of the older age groups may be driven by the desire to
seek more clement winter conditions in other provinces
or jurisdictions. Given these patterns of emigration, the
results may not be applicable to other jurisdictions, but
do indicate that emigration of cancer cases is an issue
that should be addressed when undertaking survival
analysis.
References
1. Jensen OM, Parkin DM, MacLennan R, Muir CS, Skeet
RG. Cancer Registration: Principles and Methods. Lyon:
International Agency for Research on Cancer, 1991; IARC
Scientific Publications No 95.
2001
2. Parkin DM, Wagner G, Muir, CS. The Role of the Registry
in Cancer Control. Lyon: International Agency for
Research on Cancer, 1985; IARC Scientific Publications
No 66.
3. Government of Alberta. The Alberta Cancer Programs Act
(1992). Chapter C-1, Part 1.1.
4. Statistics Canada. Interprovincial migrants 5 years and
over (place of residence 5 years ago) by age group, sex
and mother tongue, showing province or territory of
residence 5 years ago for Canada, Provinces and
Territories 1991 and 1996 Censuses (20% sample data).
Ottawa, 1998; Cat. 93F0028XDB96010.
5. Berrino F, Sant M, Verdecchia A, Capocaccia R,
Hakulinen T, Esteve J. Survival of Cancer Patients in
Europe: The EUROCARE Study. Lyon: International
Agency for Research on Cancer, 1995; IARC Scientific
Publications No 132.
6. Tennis P, Andrews E, Bombardier C, Wang Y, Strand L,
West R, et al. Record linkage to conduct an epidemiologic
study on the association of rheumatoid arthritis and
lymphoma in the province of Saskatchewan, Canada.
J Clin Epid 1993;46:685–95.
7. Raymond L, Fischer B, Fioretta G, Bouchardy C.
Emigration bias in cancer survival rates. J Epi & Biostat
1996;3:167–173. ■
17
The Economic Burden of Mental Health Problems in Canada
Thomas Stephens and Natacha Joubert
Abstract
This study provides a comprehensive estimate of the economic burden of mental health
problems in Canada in 1998. In particular, it estimates the cost of non-medical services that
have not been previously published and the value of short-term disability associated with
mental health problems that was previously underestimated, according to the approach used
here. The costs of consultations with psychologists and social workers not covered by public
health insurance was $278 million, while the value of reduced productivity associated with
depression and distress over the short term was $6 billion. Several data limitations suggest
that these are underestimates. The estimated total burden of $14.4 billion places mental
health problems among the costliest conditions in Canada.
Key words: Canada; depression; distress; economic cost; population
Introduction
The objective of this study is to provide a comprehensive estimate of the economic burden of mental health
problems in Canada. In so doing, we seek to build on
estimates published in Health Canada’s Economic
Burden of Illness in Canada, 1993 (EBIC, 1993)1 and to
address some of the data issues identified in this complex
analytical task. While direct and indirect economic costs
are only one aspect of the burden of disease, they can
provide a valuable perspective for planning programs
and setting priorities.
A recent study by Health Canada’s Cancer Bureau of
the former Laboratory Centre for Disease Control (LCDC),
estimates that the economic burden of mental disorders
in Canada was $7.8 billion in 1993,1 or $8.4 billion in
1998 dollars. Mental disorders ranked seventh among
the 20 disease categories for which cost estimates were
published. Direct costs for treating medically diagnosed
mental disorders totalled $6.3 billion (1998), comprising
$3.9 billion for hospital care, $887 million for other
institutional care, $854 million for physician care, and
$642 million for prescription medications. Additional
indirect costs totalling $3.0 billion were made up of
short-term sick days ($866 million), long-term disability
($1,707 million), and premature death ($400 million),
although these latter amounts were not restricted to
diagnosed disorders.
These estimates were based on a societal perspective
and thus incorporated both direct (internal) and indirect
(external) costs, using conventional assumptions for the
calculations. For example, the value of lost productivity
due to early retirement was based on the present value of
the lifetime earnings of the person who retires early due
to a mental disorder. While there are some limitations to
the approach (for example, health care savings arising
from early death were not considered), it is consistent
across disease categories and allows for a reasonably fair
comparison of the economic burden of diseases.
However, with respect to the economic burden of
mental health problems in particular, there are some
more serious limitations to this approach. First, it
includes only medically treated, diagnosed disorders in
the direct costs (ICD-9 codes 290–319). By definition,
these are problems that come to the attention of the
health care system and that do not include states such as
distress or depression that are untreated by physicians or
other health professionals providing publicly insured
health services.
Large numbers of Canadians with mental health
problems treated outside the medical system are missing
in such medically based calculations of the direct cost of
illness. According to data from the 1996/97 National
Population Health Survey (NPHS),2 only 21% of
Canadians who consulted a psychologist about their
mental health also consulted a family doctor or a
psychiatrist in the previous year, while 29% who
consulted a social worker also consulted a physician.
Since 4% of Canadians were depressed and 20% were
Author References
Thomas Stephens, Department of Public Health Sciences, University of Toronto; and Faculty of Administration, University of Ottawa; and Thomas
Stephens & Associates.
Natacha Joubert, Mental Health Promotion Unit, Health Canada, Ottawa
Correspondence: Thomas Stephens & Associates, PO Box 837, Manotick, Ontario K4M 1A7; Fax (613) 692-1027; E-mail: [email protected]
18
Chronic Diseases in Canada
Vol 22, No 1
classified as distressed in 1996/97,3 the direct costs
associated with their mental health problems could be
considerable, but most of these would not be included in
the EBIC, which is based on publicly insured services.
A second limitation of the EBIC, which affects the
indirect cost of mental health problems, is the method
used to attribute short-term disability to specific disease
categories. Unlike the direct costs, the indirect costs are
not limited to diagnosed disorders in the EBIC analysis,
but include any health-based reason for cutting down
on normal activity. The attribution to specific disease
categories is then made on the basis of data from the
Quebec Health Survey.1 Although the Quebec data are
the only available basis for attributing short-term
disability to disease categories, they have important
shortcomings with respect to their validity and applicability. First, the validity of reports attributing activity
reduction to mental health problems is questionable
because a significant proportion of these attributions is
based on third-party reports for other members of the
household. Second, even if these reports were of unquestioned accuracy, the application of these 1992/93 Quebec
data to all of the Canadian population is doubtful: in
1994/95, Quebec residents were the least likely to report
that distress affects their life – 13% versus an average of
17% for the other provinces.3 This low level of attribution
of effects to distress in Quebec leads to underestimating
the short-term disability costs of mental health problems such as distress, estimated in the EBIC, 1993 as
$811 million ($866 million in 1998 dollars).
The objective of this analysis is to address these
shortcomings and provide a more complete estimate of
the economic burden of mental health problems in
Canada. In so doing, we are building on one of the
recommendations in the EBIC, 1993,1 namely to
“improve data sources and refine methods for direct and
indirect cost components to provide more comprehensive
information for specific diseases” (p. iv).
Methods
Data source
The source of data for the original analyses in this
study was the 1996/97 NPHS “share” file. The share file
is virtually identical to the public use file, but includes
some detail removed from the latter for reasons of
respondent confidentiality. Population estimates and
dollar values were adjusted to 1998 figures.4,5
The NPHS is the biennial survey conducted by
Statistics Canada to describe health status and health
determinants; the 1996/97 sample is representative of the
household population of Canada. Data collection for the
mental health indicators in the present study was by
personal interview of approximately 77,000 persons aged
twelve and over.2
a
Definitions of mental health problems
We used the NPHS questions on depression and distress as evidence of mental health problems. The distress
scale includes many symptoms of anxiety (e.g., feeling
nervous, restless or fidgety) and, with the depression
scale, provides a reasonably comprehensive view of
population mental health problems. Depression was
defined according to the Statistics Canada definition2 as
a probability of 90% or greater of a major depressive
episode in the previous year; the overall prevalence rate
is 4%. Unlike depression, there is no independently
verified definition of “high distress” for the measure
used in the NPHS. We used as a definition a response of
“a lot” or “some” to the question “How much do these
(distressing) experiences interfere with your life or
activities?” regardless of level of distress on the 24-item
scale preceding the question on impact. By this definition, 15% of Canadians can be regarded as distressed.
There is a fairly high association between depression
and distress: 53% of depressed persons also reported
distress, and 24% of distressed persons were depressed.
In order to avoid double-counting these persons, all
analyses in this paper consider two groups in turn – all
depressed persons, then distressed persons who are free
of depression.
Direct costs
The EBIC, 1993 uses a “top-down” approach to
estimating the direct costs of illness. That is, estimates
are based on a known total for health care costs, which is
then allocated to various disease categories, according to
the principal diagnosis for the care received. In contrast,
this study is obliged to use a “bottom-up” approach,
estimating the volume of non-medical health care
associated with mental health problems, and then the
associated cost.
The NPHS ascertained the number of consultations
with each of psychologists, social workers, physicians,
and other health professionals in the previous 12 months,
for reasons of “physical, emotional or mental health.”
The survey also separately identified visits to psychologists, social workers, physicians and others, for reasons of
one’s “emotional or mental health,” but did not ascertain
the number of these mental health visits. To estimate the
number of social worker and psychologist visits for mental
health reasons, we combined data from these two separate
questions. Further, to exclude publicly insured consultations with psychologists and social workers (e.g., in
hospitals) already included in EBIC estimates, in the
absence of survey data on the location of the consultation, we adjusted the total of visits to reflect the proportion provided by psychologists or social workers in the
absence of any physician consultation. As noted above,
this is 79% of those consulting a psychologist and 71%
of persons seeing a social worker.a
For psychologists, this corresponds reasonably well to an estimate of 69% of service hours spent in private, as distinct from institutional, settings by
1,065 respondents to a 1999 survey of the 3,240 psychologists registered with the Canadian Register of Health Service Providers in Psychology.6
2001
19
The number of psychologist and social worker visits
by all depressed and distressed/not depressed persons was
obtained from the NPHS, adjusted for the proportion of
institutional visits and further adjusted for a population
growth of 1.4% between 1996/97 and mid-1998,4 and
then multiplied by the average cost of such visits ($125).b
With respect to medications, there are severe limitations to the NPHS data. They are restricted to reports of
any use in the previous month and there is no information that would permit an estimate of the annual
frequency of use. Unless frequency can be obtained from
some other recent and comparable source, the cost of
these medications has to be limited to the cost estimated
in EBIC, 1993, which is confined to prescriptions arising
from medical care.
Indirect costs
The indirect costs of mental health problems not fully
accounted for in the EBIC, 1993 analysis are, for reasons
described above, those due to short-term work loss associated with depression or distress. In the present study,
short-term work loss is calculated from the NPHS using
the questions on two-week disability days (cut-down
days + bed-days). Excess time off associated with
depression is obtained by comparing the disability days
of depressed vs. non-depressed persons and then, similarly, distressed vs. non-distressed persons. Although the
exact health reason for the time off was not ascertained
and has to be assumed to be mental-health related, this is
analogous to the procedure that attributes excess sickdays to smokers.7
Since the prevalence of depression and distress varies
according to labour force status, we estimated work-loss
days separately for part-time workers, full-time workers,
and non-employed persons. As the NPHS does not identify on which days of the week the reduced activity
occurred, the proportion that are work days was estimated
by assuming that the probability is equal for any day of the
week being a sick day, and multiplying the total by 5/7
(usual work days/week). Assuming two weeks of annual
holidays, the two-week total was then multiplied by 25 to
give an annual estimate for work-days of restricted activity. In the absence of an exact report of hours worked
per week, we weighted the work-loss of part-time workers
by a factor of 0.5 in estimating their contribution to the
total for the worker population. The dollar value of this
lost time was calculated using average employment
income for full- and part-time workers as published by
Statistics Canada,8 expressed in 1998 dollars.
To maintain consistency with the approach used in
EBIC, 1993, the disability-days of persons outside the
work force was also calculated. A proportion of the
disability days of full-time workers (two of seven days)
and of part-time workers (four and a half of seven days)
b
was added to this total, to account for their activity
restriction outside of the usual working days. The value
of this lost time was obtained by assuming unpaid work
is worth $15,000 annually, based on an hourly wage of
$7.50 and 2000 hours of work annually. This is consistent with the “generalist” approach to the value of unpaid
work as used in EBIC, 1993.9
Next, total disability days were adjusted for the fact
that most are not days of complete inactivity, but only of
reduced activity. For working persons, such “cut-down”
days constitute 74.1% of all two-week disability days.2 If
cut-down days are weighted as 0.5 of a bed-day, then the
adjustment required to take account of the proportion of
cut-down days is (74.1 × 0.5 + [1 − 74.1] × 1.0) = 0.6285.
Results
Direct costs – visits to non-medical mental health
professionals
In 1996/97, depressed persons age 12 and older who
sought professional help for mental health reasons made
almost 1.5 million visits to social workers and more than
850,000 visits to psychologists (Table 1). The 1998 equivalent is estimated at 2.38 million visits in total, after
adjusting for population growth of 1.4%.4 In addition,
1.6 million Canadians reported being distressed without
being depressed. While the vast majority of them did not
seek care from any mental health professional, there
were approximately 280,000 visits to social workers and
328,000 visits to psychologists. These consultations are
the equivalent of 616,000 visits in 1998.
For depression and distress combined, there were
almost 3 million visits to psychologists and social
workers in 1998. An estimated 2.2 million of these visits
took place on a fee-for-service basis outside institutions
(Table 1). At $125 each, the total cost for these visits
exceeds $278 million.
Indirect costs – days off work
In 1998, almost 678,000 employed Canadians accumulated more than 39,000 excess person-years of shortterm reduced activity associated with depression and
another 2 million had over 115,000 person-years of time
off associated with distress (Table 2). Among unemployed
Canadians, there were more than 76,000 days of reduced
activity associated with depression and 224,000 associated
with distress.
After adjusting for part-time work, inflation, and the
preponderance of cut-down days over bed-days, the total
value of lost work time was $2.16 billion. An amount
equivalent to $3.86 billion in unpaid work was similarly
reported by depressed and distressed persons. The total
value of paid and unpaid work lost associated with these
conditions was $6.02 billion in 1998 (Table 2).
Provincial bodies that license psychologists and social workers were contacted for their fee schedules: six replied, representing 85% of the Canadian
population. While fees range widely ($60-$180/session) among and within provinces, the weighted average is $125. This figure was confirmed as a
reasonable estimate by the Canadian Register of Health Service Providers in Psychology (P. L-J. Ritchie, personal communication, October 13,
2000).
20
Chronic Diseases in Canada
Vol 22, No 1
TABLE 1
Number of consultations with social workers and psychologists for reasons of mental health,
Canada, age 12+, 1998
Condition
Social workera
Depressed
1,491,423
279,634
– all settings
1,771,057
– fee basis only
1,257,450
Distressed (not depressed)
Psychologista
Totala
1998
858,223
2,349,646
2,382,541
327,604
607,238
615,739
1,185,827
2,956,884
2,998,280
936,803
2,194,253
2,224,973
Both conditions
a
Source of unadjusted data: National Population Health Survey, 1996/97 share file
TABLE 2
Indirect cost of depression and distress, Canada, age 15+, 1998
Excess time lost
Condition
Population affecteda
Average days in
2 weeksa
Total person-years
Cost $
Depressed
– paid work
677,625
0.84
39,075
451,676,778
– unpaid workb
536,221
2.00
76,393
967,268,150
– paid work
2,043,168
0.82
115,397
1,711,976,531
– unpaid workb
2,341,064
1.34
224,126
2,892,781,577
– paid work
2,720,793
0.83
154,472
2,163,653,309
– unpaid workb
2,877,285
1.51
300,519
3,860,049,727
Total
5,598,078
1.28
454,991
6,023,703,036
Distressed/not depressed
Both conditions
a
b
Source: National Population Health Survey, 1996/97 share file
Includes the value of unpaid work of full- and part-time workers while not at work
Summary
Table 3 summarizes the direct and indirect costs for
depression and distress as estimated in the present study,
and the direct costs and indirect costs for medically
treated mental disorders as estimated in EBIC, 1993. Our
estimate for the economic burden of mental health
problems increases previously published estimates1 by
71% – after adjustment for inflation between 1993 and
1998. The total in 1998 was $14.4 billion.
Discussion
This attempt to provide a comprehensive estimate of
the economic burden of mental health problems reveals
that previous estimates1 may be far too low, primarily
due to attributing too small a proportion of lost
productivity to mental health problems. However, the
current study may also be an underestimate of the true
value, due to several limitations. The principal ones are:
• The mental health problems newly accounted for in
this analysis are limited to depression and distress.
These conditions are important, but they are not
2001
exhaustive; others such as phobias could not be
accounted for, although some anxiety symptoms are
part of the distress scale.
• The NPHS definition of depression is conservative: it
counts only those persons who report feeling “sad,
blue or depressed for two weeks or more in a row”
during the past 12 months and whose responses to a
symptom checklist indicate a probability of a major
depressive episode during the past year of 90% or
more. This definition would exclude anyone with
transient feelings of depression; such persons might
well have taken time away from work or other usual
activities, however.
• It was not possible to estimate the cost of over-the-
counter medications possibly used in response to
depression and distress. As collected by the NPHS,
these would be sleeping pills, painkillers, stomach
remedies and laxatives, but person-level data on frequency of use needed to calculate annual consumption
are not available.
21
TABLE 3
Summary of costs related to mental health
problems, Canada, 1998 ($ million)
Cost
Sourcea
Treatment
– of diagnosed disorders
– medications
642
EBIC
– physicians
854
EBIC
3,874
EBIC
887
EBIC
278
This study
– hospitals
– other institutions
– of depression and distress
– non-publicly insured mental
– health professionals
Total
6,257
Lost productivity
– short-term disability
6,024
This study
– long-term disability
1,708
EBIC
400
EBIC
– early death
Total
8,132
Total
14,389
a
EBIC estimates from Reference 1, adjusted for inflation of 6.68% between 1993 and
1998.5
• It is not clear that the true extent of reduced productiv-
ity due to distress and depression is captured by the
question used in the NPHS, “[During the last 14 days],
did you stay in bed/cut down on normal activities
because of illness or injury?” It seems unlikely that
distress, in particular, would be universally regarded
as an “illness.”
• Long-term work loss was not included in this analysis
due to data limitations, including the strong possibility
that depression and distress are the result as much as
the cause of activity restriction.
• The value of reduced productivity of non-employed
persons has been set at the equivalent of $15,000 per
year – a very conservative figure.
• No account has been taken of the cost of violence and
early school departure that may accompany depression
and distress. Nor have we included the costs of
smoking, drug and alcohol abuse that may be used to
cope with depression and distress, nor the cost to
family and friends of providing support to persons in
need. Further, no estimate has been made here of the
large amounts of time devoted to personal crisis
counselling by other professionals, e.g., guidance
counsellors in schools, EAP staff in work settings,
and clergy in the community.
All of the foregoing limitations would produce
underestimates of the true economic burden. Only one
limitation – co-morbidity – might inflate these estimates.
22
Chronic Diseases in Canada
While it is possible to estimate the excess days off
work associated with depression and distress (Table 2),
there is no way of knowing whether mental health
problems are the primary cause of the lost productivity,
or whether a co-existing condition might be the cause,
since the NPHS does not provide this detail (and, as
noted, other sources such as the Quebec Health Survey
are inadequate for this purpose). If conditions other than
mental health problems lie behind the reduced productivity associated with depression and distress in this
analysis, however, they appear to be limited: depressed
persons report half the number of co-existing physical
conditions (1.8) of persons with chronic physical conditions (3.3) (G. Torrance, personal communication, May 1,
2000). While there is an association between depression
and number of physical health problems, and similarly
between distress and physical health, it is modest for
those with one or two physical conditions.3
The upshot of all these limitations is that the estimates
presented in this paper are likely quite conservative. We
can thus conclude with fair confidence that the economic
burden of mental health problems – both medically
treated and not – is $14.4 billion annually, at a minimum.
Implications
The major implications of this study are similar in
many ways to those described in a recent analysis of the
mental health status of the Canadian population.3 The
new element is the dollar figures.
These results strongly suggest that promoting the mental
health of Canadians would be a sound investment, not
only to prevent mental health problems but also to reduce
the staggering economic burden associated with them.
This analysis demonstrates that these are much higher
than suggested by previous studies,1 and indeed, are
likely considerably higher than the available data
suggests, due to the many limitations described above.
Moreover, the number of persons in distress may
increase, in tandem with current trends in child poverty,
income disparities, involuntary part-time work, singleparenting, youth unemployment, and declining expenditures on health, welfare and education.3,10 It is striking
that youth now exhibit the highest distress levels in the
population, when they had the lowest levels 20 years
ago.3,10 This trend raises the possibility of lifelong
problems for the current youth cohort, exacerbated by
the sharp decline in the support provided by the community and the mental health system. This situation will
continue to deteriorate as long as individual support
networks and the broader social safety net are not repaired
and maintained. Whatever benefits children and youth
may get from various programs can be very difficult to
sustain when there is a lack of support from the family
and the community.11 Social support can be increased by
fostering the development of meaningful relationships in
families and social environments – in schools,
workplaces, the community and institutions.12
Vol 22, No 1
It is clear that offering only more “services” will not
respond effectively to the population’s mental health
needs. Since approximately 60% of people with mental
health problems do not receive care from a health professional, the apparent gap in services is simply too big
to fill. What is evidently needed is a different kind of
investment to promote the population’s mental health.
Generally speaking, this could take the form of developing individual and community resourcefulness, and
promoting resilience among individuals of all ages.13,14
The significant contribution of mental health problems
to the global burden of disease is being addressed by a
growing number of countries, including Canada, the
United States, Australia, New Zealand, and the Member
States of the European Union, all of which are developing
national plans of action or other initiatives to promote
the mental health of their populations.11 The dollars thus
invested would represent a small figure compared with
the economic burden if nothing is done, judging by our
analysis.
This analysis also has implications for research,
especially for the collection of data in future population
surveys. In order to calculate both direct and indirect costs
adequately, it is essential to determine the respondent’s
assessment of (a) the reason for cutting down on normal
activities, in a manner that can deal with co-morbidity,
(b) whether any mental health services received were
covered by public health insurance, and (c) the frequency
and dosage of over-the-counter medications taken for
mental health reasons. To enhance the validity of these
reports, they should be obtained directly from the
respondent; third-party reports should not be accepted.
Even with these improvements in data, the economic
burden of mental health problems will likely continue to
be underestimated until they are reported as openly as
are physical health problems.
Acknowledgments
This study was supported financially by the Mental Health
Promotion Unit of Health Canada; George Torrance of Health
2001
Canada provided access to the NPHS 1996/97 share file.
George Torrance, Doug Angus and anonymous reviewers
provided constructive comments on earlier drafts.
References
1. Moore R, Mao Y, Zhang J, Clarke K, Laboratory Centre
for Disease Control. Economic Burden of Illness in
Canada, 1993. Ottawa: Health Canada, 1997.
2. Statistics Canada. National Population Health Survey,
1996/97. Share file and public use data file documentation.
3. Stephens T, Dulberg CS, Joubert N. Mental Health of the
Canadian population: A comprehensive analysis. Chronic
Dis Can 1999;20 (3)118–126.
4. Statistics Canada. CANSIM Matrices 6367-6378 and
6408-6409.
5. Statistics Canada. Consumer price index, 1996
classification, average annual all-items indexes, Canada,
historical summary. CANSIM, Matrix 9957.
6. Doody K. CRHSPP registrants speak: “Tell us more about
the CRHSPP.” Rapport 2000, 7 (1), 10.
7. Choi BK, Robson L, Single E. Estimating the economic
costs of the abuse of tobacco, alcohol and illicit drugs: A
review of methodologies and Canadian data sources.
Chronic Dis Can 1997; 18(4):149–165.
8. Statistics Canada . Employment Income. Catalogue No.
13-217-X1B.
9. Statistics Canada. Unpaid Work of Households. The Daily,
Dec 20, 1995.
10. Stephens T. Population Mental Health in Canada. Ottawa:
Mental Health Promotion Unit, 1998.
11. Joubert N. Promoting the best of ourselves : Mental health
promotion in Canada. International Journal of Mental
Health Promotion 2001;3:35–40.
12. Cohen S, Underwood LG, Gottlieb BH. Social Support
Measurement and Intervention. New York: Oxford
University Press, 2000.
13. Pransky J. Prevention: The Critical Need. Springfield:
Burrel Foundation and Paradigm Press, 1991
14. Durlak J & Wells AM. Primary prevention mental health
programs for children and adolescents : A meta-analysis
review. American Journal of Community Psychology
1997; 25:115–152. ■
23
The Storage of Household Long Guns: The Situation
in Quebec
Michel Lavoie, Lise Cardinal, Antoine Chapdelaine and Danielle St-Laurent
Abstract
This survey on the storage of household firearms in Quebec was conducted in 1994. At that
time, 35% (175/504) of survey participants who kept long guns in their homes had failed to
comply with Canadian firearm storage regulations. In most cases (85%; n = 149), this was
because at least one stored long gun was found to be both operable and accessible. Thirtyseven per cent of participants stated that no one, including themselves, had used their
firearm(s) in the 12 months preceding the survey. These findings point to two possible ways of
dealing with long guns kept in the home: render these weapons inoperable or inaccessible,
which would increase the level of compliance with the regulations, and dispose of those no
longer in use. The results of this survey have never been published before, and constitute the
only information of this kind with respect to Quebec.
Key words: firearms; home; storage; survey
Introduction
Between 1989 and 1997, an average of 1,252 firearmrelated deaths were reported each year in Canada. Of
these, 80% were suicides, 15% were homicides, 4% were
“accidents” and, in 1% of cases, the cause was unknown.1
Approximately one third of these deaths (30%) occurred
in Quebec.2 Most firearm-related deaths (at least three in
four) in Quebec were linked to the discharge of long
guns (shotguns and rifles) or, more rarely, handguns
(pistols or revolvers).3 A Quebec study of 425 cases of
firearm-related suicide occurring between September 1
and September 31, 1996 indicated that 30% of the victims
were not the owners of the gun and that in most cases
the gun had not been safely stored.4
In 1992, Canada had the sixth highest rate of firearmrelated deaths (rate per 100,000 inhabitants, adjusted for
age) among 26 countries deemed to have a high gross
national product (World Bank classification). That year,
Canada’s rate of firearm-related deaths was 4.31,
compared to 14.24 in the United States, which had the
highest rate, and 0.05 in Japan, which had the lowest.5
In 1993, the direct and indirect costs associated with
firearm-related deaths and injuries in Canada was
estimated at $6.6 billion.6
Case-control studies conducted in the United States
have shown that the presence of a firearm in the home
increased the risk of firearm-related death for household
members and relations: the members of households
where firearms were kept had a 4.7-time greater risk of
committing suicide7 and a 2.7-time greater risk of being
the victim of a homicide8 than those living in households
where no firearms were kept. The risk of suicide
increased ninefold (9.0) in cases where the firearm was
stored loaded (compared to homes without firearms), or
threefold (3.0) when the weapon was kept under lock
and key or rendered inoperable.7 Members of households
in which there were firearms were also 22 times more
likely to die from firearm-related injuries (suicide,
homicide or accident) than they were to kill an intruder
in self-defence using a firearm.9
Many experts believe that an important factor in
reducing the number of firearm-related deaths and
injuries is to reduce the accessibility of firearms by
reducing the number of firearms in the home or by
storing household firearms more safely.10,11 In Canada,
the Firearms Act contains provisions pertaining to both
these strategies.12 Provisions determining the rules
governing firearm storage have been in place since
January 1993.13 These provisions stipulate that firearms
Author References
Michel Lavoie, Lise Cardinal and Antoine Chapdelaine, Direction de la santé publique de Québec, and Institut national de santé publique du Québec
(Québec)
Danielle St-Laurent, Institut national de santé publique du Québec (Québec)
Correspondance: Dr. Michel Lavoie, Direction de la santé publique de Québec, 2400 d’Estimauville, Beauport (Québec) G1E 7G9; Fax: (418) 666-2776;
E-mail: [email protected]
24
Chronic Diseases in Canada
Vol 22, No 1
must be stored unloaded, rendered inoperable or
inaccessible and, preferably stored separately from
the ammunition.13
This paper examines the results of a 1994 survey on
household firearm storage practices in Quebec14 in the
context of Canadian firearm storage regulations.13 These
results provide an estimate of the level of compliance
with firearms regulations shortly after these regulations
came into force. They have never been published before
and constitute the only available information on the
situation in Quebec.
Methodology
The target population was composed of Quebec residents who were 18 years of age or older and who owned
at least one firearm that they stored at home as of
September 1, 1994. The total survey population comprised 515 firearm owners drawn from a random sample
of 4,654 households selected from Quebec telephone
directories. These households were selected to reflect the
demographics of the various administrative regions in
Quebec. Only 17% (n = 792) of the households were
found to include an adult who owned a firearm.
Each gun owner in these households was invited to
participate in the survey. A total of 524 firearm owners
agreed to take part, although nine of them ultimately
chose to terminate their participation before the survey
was completed. This represents a participation rate of
65% (515/792). Ninety-eight percent (n = 504) of
participants indicated that they kept at least one long gun
(shotgun or rifle) in their home, but only 7% (n = 36)
indicated that they owned a handgun (pistol or revolver).
Only the results pertaining to the 504 owners of long
guns are presented here. The maximum margin of error
for a sample of this size (n = 504) is plus or minus 4.4 %,
based on an alpha threshold of 0.05.
The data were collected between September 1 and 13,
1994, or approximately one and a half years after the
firearm storage regulations13 of the Canadian Firearms
Act came into force12 (January 1, 1993). Data collection
was entrusted to a well-known professional polling firm,
Le Groupe Léger & Léger Inc. The data were obtained
directly from long gun owners by means of telephone
interviews conducted by professional, bilingual
interviewers using a pre-tested, standardized
questionnaire that appeared on a computer screen. The
data were recorded as the interview proceeded, with each
interview lasting approximately 12 minutes.
The goal was to describe the state of long gun storage
in the context of Canadian firearm storage regulations.13
These regulations stipulate that firearm storage must
meet the three following criteria to be deemed safe: the
firearm must be stored unloaded (first criterion); it must
be rendered inoperable or inaccessible (second criterion);
and the ammunition must be securely stored (third
criterion). A long gun is considered to be unloaded (first
criterion) if there is no ammunition in the cartridge
magazine. It is rendered inoperable (second criterion) by
2001
means of a secure locking device or by the removal of a
part which is essential to its operation, such as the bolt.
In order to be inaccessible (second criterion), a firearm
must be stored in a securely locked place that cannot
readily be broken open or into. This can be a room, a
receptacle or a container. The ammunition (third
criterion) must be stored separately from the firearm, or
together with the firearm, provided that the place in
which it is stored is inaccessible (securely locked and not
readily broken into). In the latter case, the place can be a
container or a receptacle but not a room.
Long gun storage practices were described by asking
each participant about a single weapon. This was done
for practical and methodological reasons, as several
questions must be asked to evaluate the storage
conditions for a single weapon. In cases where a
participant indicated that he/she kept more than one
firearm at home, one firearm was selected at random
using a software program that made the selection from
the list of firearms declared by the participant.
Participants were asked to respond to a series of
questions specifically designed to determine whether the
firearm was equipped with a secure locking device;
whether any of its parts had been removed; whether it
was being stored in a container, a receptacle or a room;
whether the place of storage was locked and whether it
could readily be broken open or into; whether the ammunition was stored with the weapon or in a separate place;
and, finally, whether the weapon was stored loaded.
Each participant’s answers were analyzed during the
interview using a special software program to determine
whether storage practices complied with the three criteria
stipulated in the regulations. Then, each participant was
placed into one of two categories: compliers (firearm
storage complies with all three criteria) or non-compliers
(firearm storage does not comply with at least one of the
three criteria). In other words, the first group supposedly
comprised those who stored their firearms securely,
while the second group comprised those who failed to
meet the safe storage criteria.
Findings
The majority of the participants were between the
ages of 35 and 54 (53%). There were nine times as many
men (n = 465) as women (n = 50). In 96% of cases, the
language spoken was French, and nine times out of ten
the participant was not the sole occupant of the household. A greater percentage of participants were from
rural areas (60%) than from urban centres (40%). The
majority of participants (64%) indicated that they had
completed 12 years or less of schooling.
On average, participants kept 2.7 long guns in their
homes. Thirty-two per cent owned only one gun, 29%
owned two, 18% owned three, and 21% owned four or
more. The three most popular long guns were, in
decreasing order, 12-calibre shotguns (55%), .22-calibre
rifles (42%), and .410-calibre shotguns (24%). In 22% of
cases, participants indicated that other persons had
25
TABLE 1
Reasons given by participants for owning
a firearma
Reason stated
%b
Hunting
87
Target practice
11
Gun collecting
7
Souvenir
5
Self-protection
3
Employment
2
Predators and other pests
1
Other
1
a
b
TABLE 2
Factors associated with compliance with
firearm storage regulationsa
Factors
Noncompliersb
(n = 175)
Compliers
(n = 329)
%
%
%
(n)
c36c
64
86
100
100
(482)
(22)
73
57
57
100
100
(267)
(160)
d (77)d
54
68
100
100
(109)
(394)
Total
(n = 504)
Language spoken
French
Other
14
Aware of the existence
of an Act
Yes
No
DK/NR
Shotgun or rifle.
The total exceeds 100% because some participants (n = 504) mentioned more than one
reason for owning a firearm (n = 593).
27
d43c
43
Firearm accessible to
one other person
access to their firearms. Thirty-seven per cent of
respondents indicated that no one, not even themselves,
had used their firearms in the 12-month period preceding
the survey.
Hunting was the most frequently mentioned reason
(87%) for owning a long gun (Table 1). A small
percentage of participants owned a firearm for selfprotection (3%) or to hunt predators or other pests (1%).
At the time of the survey, 91% of participants stated that
they had received training in the handling of firearms
and 53% were aware of the existence of an act governing
the storage of firearms in Canada.
The study findings show that, at the time of the
survey, 65% of participants (n = 329) had at least one
long gun that was securely stored in their home ( in
accordance with the three criteria). Specifically, almost
all participants indicated that the weapon for which they
were providing information was stored unloaded (99.6%;
first criterion) and that the ammunition for this weapon
was also securely stored (91%; third criterion). Seventy
per cent of the gun owners surveyed stated that their
weapon had been rendered inoperable or inaccessible
(second criterion).
On the other hand, the results also show that 35% of
participants (n = 175) had failed to comply with at least
one of the three criteria of safe storage. The likelihood of
non-compliance was greatest (Table 2) among participants whose spoken language was French (p < 0.05);
those who owned at least one weapon that was accessible to others (p < 0.05); and those who were not aware
that there is a law governing firearm storage in Canada
(p < 0.05). However, the other variables considered were
not associated with firearm storage practices (p ≥ 0.05).
These included age (18–34; 35–54; 55+); gender (male,
female); place of residence (rural, urban); years of
schooling (≤ 12, > 13); family income (< $40,000,
≥ $40,000); living alone (yes, no); the presence of
children under the age of 18 (yes, no); the number of
firearms owned (1, 2, 3, 4+); the reasons for owning
26
Chronic Diseases in Canada
Yes
No
d46c
32
a
Shotgun or rifle
Failure to comply with at least 1 of the 3 storage criteria.
p < 0.05
d Not taken into account.
b
c
firearms (hunting, target practice, gun collecting); the
calibre of weapon (12, .22, .410); weapon(s) not used
during the past 12 months (yes, no); and prior training in
the handling of firearms (yes, no).
Figure 1 represents the survey findings with respect to
the 175 non-compliers. Eighty-five percent (n = 149) did
not comply with Canadian firearm storage regulations by
having at least one stored long gun which was both
operable (not equipped with a secure locking device and
comprising all of its parts) and accessible (stored in an
unlocked place or a locked place that could readily be
broken open or into).
The 149 non-compliers who kept at least one long gun
that was operable and accessible (non-compliance with
the second criterion of firearm storage) were questioned
more closely. Each of the non-compliers (n = 149) was
asked the following question: “If you were to improve
the conditions under which your firearms are stored,
what would be your first step?” (A list of options was
read out, but the respondent was asked to choose only
one). The two options most frequently mentioned by
non-compliers were to render the firearm inoperable, that
is equip it with a secure locking device or remove a part
needed to discharge the weapon (40%; n = 60); and to
ensure that the weapon is inaccessible, in other words,
store it in a place that is locked and cannot readily be
broken into (24%; n = 36).
The non-compliers were then asked the following
questions: “Why is your firearm stored in an operable
condition and in a place where it is accessible? What
would prompt you to render your weapon both
inoperable and inaccessible?” (No answers to this
Vol 22, No 1
FIGURE 1
Distribution of non-compliers (n = 175/504)
according to firearm (shotgun or rifle)
storage criteria studied
Criterion 1:
Firearm loaded
1% (n = 2/175)
Criterion 2:
Firearm operable and accessible
85% (n = 149/175)
to store a gun was also mentioned by a number of noncompliers (8%; n = 12). It is important to note that 15%
(n = 22) of non-compliers could offer no particular
reason why their firearms were left in an accessible
place. Again, the presence of children was most
frequently invoked by non-compliers as a reason that
would prompt them to make their long guns inaccessible
(26%; n = 39). However, a significant percentage of noncompliers (35 %; n = 52) saw no particular reason to
make their firearm inaccessible.
Discussion
0
1
129
1
18
0
26
Criterion 3:
Ammunition removed,
improperly stored
26% (n = 45/175)
question were suggested and only one answer was
recorded for each non-complier). In each case, the
storage condition at issue was clearly defined to ensure
that the respondent fully understood the questions
(operable/inoperable; accessible/inaccessible).
Operable firearm
In explaining why their long guns were operable, the
149 non-compliers stated that they had taken other safety
precautions (19%; n = 28) and that their weapon was
safely hidden away (7%; n = 10). Negligence was also
invoked as an explanation by a number of non-compliers
(7%; n = 10). It should be noted that 15% (n = 22) of
non-compliers could offer no particular reason to explain
why their weapons were stored at home in an operable
condition. The presence of children was most frequently
invoked by non-compliers as a reason that would prompt
them to render their long guns inoperable (28%; n = 42).
An equal percentage of respondents (28%; n = 42) saw
no particular reason to render their firearms inoperable.
Accessible firearms
A significant percentage of the 149 non-compliers
indicated that they left their firearms in an accessible
place either through negligence or force of habit (13%;
n = 19), or because they felt the firearms were well
hidden (9%; n = 13). Not having any other suitable place
2001
Strengths and limitations
Several aspects of this survey must be emphasized:
• the participation rate was 65%, which is satisfactory
given the subject matter (storage of firearms) and the
data collection method (telephone survey);
• the participants were drawn from a random sample of
4,654 households, the distribution of which was based
on the demographic weight of the various administrative regions of Quebec, which is likely to ensure more
representative results;
• the state of long gun storage was described with
respect to federal regulations on firearm storage,
which not only constituted a first, but also provided a
basic measure of the level of compliance with these
regulations in Quebec;
• data on firearm storage came from the owners
themselves rather than a third party, which ensured
greater validity;15,16
• each participant was questioned on the storage of a
single long gun at a specific point in time (the time of
the interview), which tended to reduce the kind of
information bias that can result from relying on
memory;
• the final judgment on compliance with firearm storage
regulations was made by the researchers rather than
the gun owners, which was an advantage given the
relative complexity of this type of judgment (where
three criteria must be considered).
The results of this survey on long gun storage are
subject to three types of bias. The first is that firearm
storage practices constitute a form of reported behaviour.
Some participants may deliberately have indicated that
their firearms were stored more securely than they in fact
were, which would lead the researchers to underestimate
the number of non-compliers.
The second bias is linked to the voluntary aspect of
participation in the survey. It may be that those who chose
to participate were, on average, more likely to comply
with firearm storage regulations than those who refused
to take part, which would also cause the researchers to
underestimate the number of non-compliers.
27
The third bias concerns the fact that participants were
questioned about a single long gun. In cases where
several firearms were being stored at home, it is possible
that the firearm selected as the subject of the questionnaire may have been stored under different conditions
than the other weapons. If storage regulations were being
adhered to solely in the case of the selected firearm, this
too would lead researchers to underestimate the number
of non-compliers; if the opposite were true, there would
be no bias.
Conditions of storage
Based on the results of this survey, at least 35% of
persons who store one or more long gun at home fail
to comply with Canadian firearm storage regulations.13
It should be remembered that 89% (n = 156) of noncompliers stated that they were not the sole occupants of
their household (in 59% of cases the other occupants
were children < 18 years old), and 29% (n = 51) stated
that their gun was accessible to others, usually a spouse
(70%) or a child (40%).
The results of the survey lead us to estimate that at
least 6% of Quebec homes contain at least one long gun
that is improperly stored. This estimate was calculated
by multiplying the percentage of Quebec households
where at least one firearm is stored by an adult (17%)
by the percentage of non-compliers among study
subjects (35%).
The study did not link storage practices with firearms
training as the same number of compliers and noncompliers had been trained. This is likely due to the fact
that most of the participants had never received adequate
instruction on the storage of firearms. In fact, at the time
of the survey (1994), 91% of participants indicated
having received training in the use of firearms in the
past. However, before the Canadian firearm storage
regulations came into effect (January 1993), the safe
storage of long guns was not part of the firearms training
courses. When the regulations came into effect, this
deficiency was remedied, but the “enriched” training was
only required for new gun owners, which likely affected
only a small proportion of the study participants since it
was carried out in 1994.
To our knowledge, there is no basis for comparison of
firearm storage practices in Quebec and elsewhere.
However, a survey conducted in 1999 with 282 long gun
owners from all 10 provinces is somewhat interesting.17
Methodological differences, however, make it rather
risky to draw comparisons between the two surveys
(Appendix 1). The study populations are not truly
comparable, the criteria used to evaluate long gun
storage conditions are considerably different, and the
survey questions did not relate to the same number of
weapons.
The main results of the 1999 survey are nonetheless
presented by way of indication: 99% of the participants
in that study stated that all the long guns they kept at
home were stored unloaded; 83% indicated that all the
28
Chronic Diseases in Canada
long guns they kept at home were either kept under lock
and key or had been rendered inoperable; and 98% of
participants indicated that the ammunition for their
firearms was stored securely, either in a separate place,
or with the weapon, but in a locked compartment.
Approximately 17% of participants had failed to comply
with at least one storage criterion. Interestingly, similar
results are observed in the 1994 survey if the 1999
storage criteria are applied (results not presented).
Courses of action
The study findings suggest some possible methods of
increasing the level of compliance with Canadian firearm
storage regulations.13 In 85% of cases where the owner
of a long gun had failed to comply with federal
regulations (149/175), the weapon in question was both
operable and accessible. To increase the level of
compliance, an important goal should be to encourage
the owners of long guns to render them inoperable (e.g.
by use of a locking device) or inaccessible (e.g. by
storing them in a securely locked place). These two
measures are the ones most frequently identified by the
non-compliers as a way of improving security. The
survey findings also show that to persuade non-compliers
to take these measures, they must be made aware of the
fact that they will be protecting their children, as well as
the children of neighbours and relations.
Study results indicate that reducing the number of
firearms stored at home is another possible solution.
More than a third (37%) of participants indicated that
their gun(s) had not been used by themselves or by
anyone else during the 12-month period preceding the
survey. It should be noted that these percentages were
even higher among non-compliers than compliers. It
would seem important, therefore, to encourage those
who own guns that they do not use to dispose of these
weapons.
Acknowledgements
We wish to thank Michel Lemieux and Pierre Duchesnay of
the Léger & Léger polling firm in Quebec City who oversaw
the data collection. This survey would not have been possible
without the support of the Comité intersectoriel sur la violence
familiale et l’entreposage sécuritaire des armes à feu au
Québec (CCAAF), the Canadian Firearms Centre at the
Department of Justice, and the regional public health branches
of Quebec City and Montérégie.
References
1. Hung CK Firearm statistics, Updated tables and special
tabulations. Ottawa: Department of Justice, March 2000.
Statistics Canada: catalogue 84–208.
2. Bureau du coroner du Québec. Décès par arme à feu,
1990-1998. Quebec: September 1999.
3. Tennina S. Enquête descriptive des décès par arme à feu,
1990-1993. Quebec: Bureau du Coroner du Québec,
October 1994.
4. Saint-Laurent D, Tennina, S. Résultats de l’enquête
portant sur les personnes décédées par suicide au Québec
entre le 1er septembre et le 31 décembre 1996. Québec:
Vol 22, No 1
5.
6.
7.
8.
9.
10.
11.
Ministère de la Santé et des Services sociaux and the
Bureau du coroner du Québec. Bibliothèque nationale du
Québec, 2000.
Krug Étienne G, Powell KE, Dahlberg LL. Firearm-related
deaths in the United States and 35 other high- and uppermiddle-income countries. Int J Epidemiol, 1998;27:214–
221. Drawn from Table 1: p. 216.
Miller TR, Cohen MA. Costs of gunshot and cut/stab
wounds in the United States, with some Canadian
comparisons. Accid Anal and Prev. 1997;29(3):329–41.
Kellermann AL, Rivara FP, Somes G. et al. Suicide in
the home in relation to gun ownership. New Eng J Med
1993;327(7):470.
Kellermann AL, Rivara FP, Rushforth NB et al. Gun
ownership as a risk factor for homicide in the home.
New Eng J Med 1993;329(15):1084–1091.
Kellermann AL, Rivara FP, Lee RK Banton JG. Injuries
and deaths due to firearms in the home. Journal of Trauma,
Injury, Infection and Critical Care 1998;45(2):263–267.
Cukier W. Firearms regulation: Canada in the international
context. Chronic Dis Can 1998;19(1):29–40.
Kellermann AL, Lee RK, Mercy JA, Banton J. The
epidemiologic basis for the prevention of firearm injuries.
Annu Rev Public Health 1991;12:17–40:30–31.
12. Acts of the Parliament of Canada, Chapter 39, An Act
Respecting Firearms and Other Weapons, Bill C-68,
assented to December 5, 1995. p. 137.
13. Department of Justice. Consolidated Regulations
Pertaining to Part III of the Criminal Code: “Firearms and
Offensive Weapons.” Criminal and Social Policy Sector,
Ottawa: October 1993. p. 25–29.
14. Le Groupe Léger & Léger Inc., Lavoie, M. and
Chapdelaine. A. Enquête sur l’entreposage des armes à
feu gardées à domicile au Québec. October 31, 1994. 44 p.
15. Ludwig J, Cook PH, Smith TW. The gender gap in
reporting household gun ownership. Am J Public Health
1998;88:1715–18.
16. Azrael D, Miller M, Hemenway D. Are household
firearms stored safely? It depends on whom you ask.
Pediatrics 2000;103 (3):1–6.
17. Angus Reid Group. Safe storage knowledge and practice:
overview of findings. For the Canadian Firearms Centre.
July 23, 1999. 5 p.
18. Brent DA, Perper JA, Allman C. et al. The presence and
accessibility of firearms in the homes of adolescent
suicides. JAMA 1991; 2666 (21):2989–2995. ■
APPENDIX
Methodological differences between the two surveys: Quebec (1994) and Canada (1999)
Populations not comparable: In the 1999 Canadian survey, only 40 participants out of 282 were from Quebec.
This is significant in that in the 1994 survey, firearm storage conditions were associated with the language spoken
by participants.
Nature of the storage criteria: Contrary to the 1994 survey, the 1999 survey defines a location as being
inaccessible if it is kept under lock and key, regardless of whether it is difficult to break open or into. As a result
of these less stringent requirements with respect to accessibility, a greater number of participants were classified as
being in compliance with two of the three criteria stipulated in federal firearm storage regulations, namely the
criterion that requires long guns to be stored in such a way as to render them inoperable or inaccessible (second
criterion) and the criterion stating that a weapon can be stored with its ammunition provided that they are stored in
an inaccessible compartment or receptacle (third criterion).
Number of long guns considered: In 1994, compliance with the storage criteria was evaluated by questioning
each participant about a single long gun (when more than one firearm was stored at home, one of them was
selected at random), whereas in the 1999 study, participants were asked to consider the storage conditions for all
of the firearms in their possession.
2001
29
Book Review
Design and Analysis of Cluster Randomization Trials
in Health Research
By Allan Donner and Neil Klar
London (England): Arnold Publishers, 2000;
x + 151 pp; ISBN 0 340 69153 0; (hardcover)
The increasing popularity of the cluster randomization
design among health researchers over the past two
decades has led to an extensive body of methodology
and a growing literature that cuts across several disciplines in the statistical, social and medical sciences. This
book is the first to provide a unified and systematic treatment of the topic. It may be used as a reference source
for investigators in the planning or analysis stages of a
study or as a textbook for a graduate level course in
research methodology.
The book includes fairly non-technical chapters summarizing key issues of study design, data analysis and
reporting as well as more technical material describing
extensions of standard regression models (e.g. generalized estimating equations approach, multilevel models)
that are needed to account for the variance inflation due
to clustering.
The book also provides intriguing discussions of the
historical development of cluster randomized trials and
summarizes the unique ethical challenges of cluster
randomization. For example, in community randomized
trials it is typically not possible to obtain informed
consent from all individuals who may be affected by the
intervention prior to random assignment. This would be
the case, for example, in trials evaluating innovative
methods of water treatment for the prevention of
infectious diseases or in trials evaluating smoking
cessation interventions using mass media.
This is a well-written book, which includes data and
worked examples illustrating methods of sample size
estimation and data analysis.
A challenge facing the authors is that the methodology
for cluster randomization trials is undergoing very rapid
development. For instance, since publication of the book,
30
Chronic Diseases in Canada
special issues of two leading journals1,2 have been devoted
to cluster randomization. It is to be hoped that a second
edition is being considered in which the authors can
discuss some of these newer developments. For example,
as the number of trials adopting cluster randomization
has increased, meta-analyses of trials using various units
of allocation are starting to appear in the literature.
However, investigators have very little guidance, as yet,
on how best to conduct such meta-analyses.
Purchasers of the book are entitled to a 25% discount
on ACluster, a computer software compatible with
Windows 3.1, 95, 98 and NT that implements many of the
sample size and analysis formulas presented in the book.
Information concerning the software can be obtained at
<http://www.arnoldpublishers.com/support/cluster/>.
References
1. Campbell MJ, Donner A, Elbourne D. Design and analysis
of cluster randomized trials. Editorial. Stat Med
2001;20(3):329–30.
2. Donner A, Klar N. Cluster randomization. Editorial. Stat
Methods Med Res 2000;9:79–80.
Yang Mao
Chief
Environment Risk and Case Surveillance
Cancer Division
Centre for Chronic Disease Prevention and Control
Population and Public Health Branch
Health Canada
Address Locator: 0601C
Ottawa, Ontario K1A 0L2
Vol 22, No 1
New Resource
NOTICE!
Canadian Cancer Statistics 2001
National Cancer Institute of Canada
Toronto (Ontario), 2001
Canadian Cancer Statistics 2001 is now accessible on
the Internet at <www.cancer.ca/stats>.
You can download and/or print any sections, graphs,
tables, etc. or all of this document from the above Web
site.
2001
If you would like to receive a hard copy of this
publication, contact
• your local office of the Canadian Cancer Society,
• your regional office of Statistics Canada
or
• Canadian Cancer Society (National Office)
10 Alcorn Avenue, Suite 200
Toronto, Ontario M4V 3B1
Tel.: (416) 961-7223
Fax: (416) 961-4189
E-mail: [email protected]
31
Calendar of Events
May 28–30, 2001
Toronto, Ontario
Second International Symposium on the
Effectiveness of Health Promotion
Centre for Health Promotion
University of Toronto
<www.utoronto.ca/chp>
June 13–16, 2001
Toronto, Ontario
Congress of Epidemiology 2001
Combined Meeting of American College of
Epidemiology, American Public Health
Association’s Epidemiology Section, Canadian
Society for Epidemiology and Biostatistics and
Society for Epidemiologic Research
<www.epi2001.org>
July 1–6, 2001
Vancouver, British
Columbia
“Global Aging: Working Together in a Changing
World”
17th Congress of the International Association of
Gerontology
Congress Secretariat
Gerontology Research Centre
Simon Fraser University
2800 – 515 West Hastings Street
Vancouver, BC V6B 5K3
Tel.: (604) 291-5062
Fax: (604) 291-5066
E-mail: [email protected]
<www.harbour.sfu.ca/iag>
July 15–20, 2001
Paris, France
“Health: An Investment For a Just Society”
XVII World Conference on Health Promotion and
Health Education
International Union for Health Promotion and
Education
Martine Lapergue
Réjane Jouan
Comite francais d’Education pour la
Sante (C.F.E.S)
XVII World Conference on Health
Promotion and Health Education
2 , rue Auguste Comte - 92174 VANVES
Cedex - FRANCE
Tel.: 33 (0)1 41 09 96 48
Fax: 33 (0) 1 46 45 00 45
E-mail: [email protected]
<www.iuhpe.org>
September 4–7, 2001
Atlanta, Georgia, USA
“Using Science to Build Comprehensive Cancer
Programs: A 2001 Odyssey”
US Department of Health and Human Services
Centers for Disease Control and Prevention
2001 Cancer Conference
Conference Registration: June 1, 2001
American Cancer Society, National
Home Office
Attn: CDC’s 2001 Cancer Conference
1599 Clifton Road, NE
Atlanta, Georgia 30329 USA
September 22–25, 2001
Sydney, Australia
4th International Conference on the Scientific
Basis of Health Services
Sydney Convention and Exhibition Centre –
Darling Harbour
<www.icsbhs.org>
32
Chronic Diseases in Canada
Vol 22, No 1
October 18–21, 2001
Saskatoon, Saskatchewan
“Health Research in Rural and Remote Canada:
Taking the Next Steps”
Second Scientific Conference and Annual
Meeting of The Rural Health Research
Consortium
The Rural Health Research Consortium
103 Hospital Drive
PO Box 120, RUH
Saskatoon, SK S7N OW8
Tel.: (306) 966-7888
Fax: (306) 966-8378
E-mail:
<www.usask.ca/medicine/agmedicine>
November 29–December 1,
2001
Toronto, Ontario
Canada’s Fifth National Conference on Asthma
and Education (ASED 5)
Abstract Deadline: June 30, 2001
A. Les McDonald, Executive Director
Canadian Network for Asthma Care
(CNAC)
1607 – 6 Forest Laneway
North York, Ontario M2N 5X9
Tel.: (416) 224-9221
Fax: (416) 224-9220
E-mail:
<www.cnac.net>
2000 Peer Reviewers
We are extremely grateful to the following
people for their enormous contribution
to Chronic Diseases in Canada as
peer reviewers in 2000.
2001
Nicholas J Birkett
Joan Lindsay
Jean-François Boivin
Sharon M MacDonald
Larry Chambers
Ian W McDowell
Clarence Clottey
Alfred I Neugut
Paul Corey
Edgardo L Pérez
Kitaw Demissie
Ken Potvin
John Frank
Elizabeth J Robinson
Brian Habbeck
Jorge Segovia
Shanna Hudson
Tom Stephens
Philip Jacobs
Robert Spasoff
Murray J Kaiserman
Jean-Pierre Thouez
Wendy Kennedy
Peter P Wang
Betsy Kristjansson
Roy West
John M Last
Russell Wilkins
33
Indexes for Volume 21, 2000
Volume 21 Contents
No 1, 2000
Agreement in Measuring Socio-economic Status:
Area-based Versus Individual Measures ............................
No 2, 2000
1
Kitaw Demissie, James A Hanley, Dick Menzies,
Lawrence Joseph and Pierre Ernst
An Assessment of the Validity of a Computer System
for Probabilistic Record Linkage of Birth and Infant
Death Records in Canada ..................................................
Commentary
Monitoring Tobacco Use in Canada: The Need for a
Surveillance Strategy.......................................................... 50
Roberta Ferrence and Thomas Stephens
8
Martha Fair, Margaret Cyr, Alexander C Allen,
Shi Wu Wen, Grace Guyon and Ralph C MacDonald,
for the Fetal and Infant Health Study Group
Infant Mortality by Gestational Age and Birth Weight in
Canadian Provinces and Territories, 1990–1994 Births ..... 14
Shi Wu Wen, Michael S Kramer, Shiliang Liu,
Susie Dzakpasu and Reg Sauvé, for the Fetal and
Infant Health Study Group
Status Reports
Canadian Strategy for Cancer Control................................ 23
Silvana Luciani and Neil J Berman
Canadian Coalition on Cancer Surveillance ....................... 26
Barbara Foster and Anna Maria Boscaino, for the
CCOCS Management Committee
Canadian Association of Provincial Cancer Agencies ........ 30
Donald R Carlow
National Enhanced Cancer Surveillance System:
A Federal–Provincial Collaboration to Examine
Environmental Cancer Risks .............................................. 34
Kenneth C Johnson
Announcement
Formation of the Canadian Prostate Cancer Research
Initiative............................................................................... 35
Book Reviews
Health Promotion Planning: An Educational and
Ecological Approach (third edition) ..................................... 36
Predictors of Smoking Cessation in an Incentive-based
Community Intervention...................................................... 54
Namrata Bains, William Pickett, Brian Laundry and
Darlene Mecredy
School-based Smoking Prevention: Economic Costs
Versus Benefits................................................................... 62
Thomas Stephens, Murray J Kaiserman,
Douglas J McCall and Carol Sutherland-Brown
Performance of the Composite International Diagnostic
Interview Short Form for Major Depression in a
Community Sample ............................................................ 68
Scott B Patten, Jennifer Brandon-Christie,
Jennifer Devji and Brandy Sedmak
Health-adjusted Life Expectancy at the Local Level
in Ontario ............................................................................ 73
Douglas G Manuel, Vivek Goel, J Ivan Williams and
Paul Corey
Ontario Familial Colon Cancer Registry: Methods and
First-year Response Rates ................................................. 81
Michelle Cotterchio, Gail McKeown-Eyssen,
Heather Sutherland, Giao Buchan, Melyssa Aronson,
Alexandra M Easson, Jeannette Macey, Eric Holowaty
and Steven Gallinger
Book Review
The Burden of Disease Among the Global Poor: Current
Situation, Future Trends, and Implications for Strategy...... 87
Reviewed by Franklin White
New Resource .................................................................... 89
Calendar of Events ............................................................. 90
Reviewed by Paula J Stewart
Quantitative Estimation and Prediction of Human Cancer
Risks ................................................................................... 37
Reviewed by Ian B MacNeill
Epidemiology of Childhood Cancer .................................... 38
Reviewed by Amanda Shaw
New Resources .................................................................. 40
Calendar of Events ............................................................. 41
No 3, 2000
Elements of Mobility as Predictors of Survival in Elderly
Patients with Dementia: Findings from the Canadian
Study of Health and Aging .................................................. 93
Anthanasios Tom Koutsavlis and Christina Wolfson
A Deprivation Index for Health and Welfare Planning in
Quebec ...............................................................................104
Robert Pampalon and Guy Raymond
1999 Peer Reviewers ......................................................... 43
Indexes for Volume 20, 1999.............................................. 44
34
Chronic Diseases in Canada
Vol 22, No 1
A Comparison of Methods for Measuring Socio-economic
Status by Occupation or Postal Area..................................114
Raywat Deonandan, Karen Campbell, Truls Ostbye,
Ian Tummon and James Robertson
Estimation of Youth Smoking Behaviours in Canada .........119
William Pickett, Anita Koushik, Taron Faelker and
K Stephen Brown
The Prevalence of Diabetes in the Cree of Western
James Bay ..........................................................................128
David AL Maberley, Will King and Alan F Cruess
Status Report
Orius Software: Calculation of Rates and Epidemiologic
Indicators, and Preparation of Graphical Output ................134
Long On, Robert M Semenciw and Yang Mao
Book Reviews
Qualitative Research Methods: A Health Focus ................. 137
No 4, 2000
Workshop Report: Physical Activity and Cancer
Prevention...........................................................................143
Loraine D Marrett, Beth Theis, Frederick D Ashbury and
an Expert Panel
Comparing Two Different Approaches to Measuring
Drug Use Within the Same Survey .....................................150
C Ineke Neutel and Wikke Walop
Book Review
Genetics and Public Health in the 21st Century:
Using Genetic Information to Improve Health and
Prevent Disease ................................................................. 157
Reviewed by Lynne Belle-Isle
Spatial Epidemiology: Methods and Applications ............... 159
Reviewed by Alette Willis
New Resource .................................................................... 161
Reviewed by Connie Kristiansen
Social Epidemiology ........................................................... 138
Calendar of Events ............................................................. 162
Reviewed by Cam Mustard
Calendar of Events ............................................................. 140
Volume 21 Subject Index
BOOK REVIEWS
The burden of disease among the global poor: current
situation, future trends, and implications for strategy.
21(2):87–8.
Epidemiology of childhood cancer (IARC Scientific
Publications No. 149). 21(1):38–9.
Genetics and public health in the 21st century: using genetic
information to improve health and prevent disease.
21(4):157–8.
Health promotion planning: an educational and ecological
approach (third edition). 21(1):36–7.
Qualitative research methods: a health focus. 21(3):137–8.
Quantitative estimation and prediction of human cancer risks
(IARC Scientific Publications No. 131). 21(1):37–8.
Ontario familial colon cancer registry: methods and first-year
response rates. 21(2):81–6.
Workshop report: physical activity and cancer prevention.
21(4):143–9.
DEMENTIA
Elements of mobility as predictors of survival in elderly
patients with dementia: findings from the Canadian study of
health and aging. 21(3):93–103.
DIABETES
The prevalence of diabetes in the Cree of western James
Bay. 21(3):128–33.
ENVIRONMENTAL HEALTH
Social epidemiology. 21(3):138–9.
National Enhanced Cancer Surveillance System: a
federal–provincial collaboration to examine environmental
cancer risks [status report]. 21(1):34–5.
Spatial epidemiology: methods and applications.
21(4):159–60.
GEOGRAPHIC VARIATIONS
CANCER
Health-adjusted life expectancy at the local level in Ontario.
21(2):73–80.
Canadian Association of Provincial Cancer Agencies [status
report]. 21(1):30–3.
Canadian Coalition on Cancer Surveillance [status report].
21(1):26–9.
Canadian Strategy for Cancer Control [status report].
21(1):23–5.
National Enhanced Cancer Surveillance System: a
federal–provincial collaboration to examine environmental
cancer risks [status report]. 21(1):34–5.
2001
Infant mortality by gestational age and birth weight in
Canadian provinces and territories, 1990–1994 births.
21(1):14–22.
The prevalence of diabetes in the Cree of western James
Bay. 21(3):128–33.
HEALTH SURVEYS
Agreement in measuring socio-economic status: area-based
versus individual measures. 21(1):1–7.
35
Comparing two different approaches to measuring drug use
within the same survey. 21(4):150–6.
Comparing two different approaches to measuring drug use
within the same survey. 21(4):150–6.
Health-adjusted life expectancy at the local level in Ontario.
21(2):73–80.
Orius software: calculation of rates and epidemiologic
indicators, and preparation of graphical output [status report].
21(3):134–6.
Monitoring tobacco use in Canada: the need for a surveillance
strategy [commentary]. 21(2):50–3.
SENIORS’ HEALTH
Performance of the composite international diagnostic
interview short form for major depression in a community
sample. 21(2):68–72.
Elements of mobility as predictors of survival in elderly
patients with dementia: findings from the Canadian study of
health and aging. 21(3):93–103.
INFANT AND CHILD HEALTH
SOCIO-ECONOMIC ISSUES
An assessment of the validity of a computer system for
probabilistic record linkage of birth and infant death records in
Canada. 21(1):8–13.
Agreement in measuring socio-economic status: area-based
versus individual measures. 21(1):1–7.
Infant mortality by gestational age and birth weight in
Canadian provinces and territories, 1990–1994 births.
21(1):14–22.
MENTAL DISORDERS
A comparison of methods for measuring socio-economic
status by occupation or postal area. 21(3):114–8.
A deprivation index for health and welfare planning in Quebec.
21(3):104–13.
STATUS REPORTS
Performance of the composite international diagnostic
interview short form for major depression in a community
sample. 21(2):68–72.
Canadian Association of Provincial Cancer Agencies [status
report]. 21(1):30–3.
METHODOLOGIC ISSUES
Canadian Coalition on Cancer Surveillance [status report].
21(1):26–9.
An assessment of the validity of a computer system for
probabilistic record linkage of birth and infant death records in
Canada. 21(1):8–13.
Canadian Strategy for Cancer Control [status report].
21(1):23–5.
Ontario familial colon cancer registry: methods and first-year
response rates. 21(2):81–6.
National Enhanced Cancer Surveillance System: a
federal–provincial collaboration to examine environmental
cancer risks [status report]. 21(1):34–5.
POPULATION SURVEILLANCE
Monitoring tobacco use in Canada: the need for a surveillance
strategy [commentary]. 21(2):50–3.
Orius software: calculation of rates and epidemiologic
indicators, and preparation of graphical output [status report].
21(3):134–6.
PUBLIC POLICY
TOBACCO ISSUES
Canadian Association of Provincial Cancer Agencies [status
report]. 21(1):30–3.
Estimation of youth smoking behaviours in Canada.
21(3):119–27.
Canadian Coalition on Cancer Surveillance [status report].
21(1):26–9.
Predictors of smoking cessation in an incentive-based
community intervention. 21(2):54–61.
Canadian Strategy for Cancer Control [status report].
21(1):23–5.
School-based smoking prevention: economic costs versus
benefits. 21(2):62–7.
National Enhanced Cancer Surveillance System: a
federal–provincial collaboration to examine environmental
cancer risks [status report]. 21(1):34–5.
SUMMARY WORKSHOP/CONFERENCE REPORTS
Workshop report: physical activity and cancer prevention.
21(4):143–9.
RESEARCH DESIGN
An assessment of the validity of a computer system for
probabilistic record linkage of birth and infant death records in
Canada. 21(1):8–13.
36
Chronic Diseases in Canada
Vol 22, No 1
Volume 21 Author Index
Allen, Alexander C
Fair M, Cyr M, Allen AC, Wen SW, Guyon G, MacDonald RC,
for the Fetal and Infant Health Study Group. An
assessment of the validity of a computer system for
probabilistic record linkage of birth and infant death
records in Canada. 21(1):8–13.
Aronson, Melyssa
Cotterchio M, McKeown-Eyssen G, Sutherland H, Buchan G,
Aronson M, Easson AM, Macey J, Holowaty E,
Gallinger S. Ontario familial colon cancer registry:
methods and first-year response rates. 21(2):81–6.
Ashbury, Frederick D
Marrett LD, Theis B, Ashbury FD, and an Expert Panel.
Workshop report: physical activity and cancer
prevention. 21(4):143–9.
Bains, Namrata
Bains N, Pickett W, Laundry B, Mecredy D. Predictors of
smoking cessation in an incentive-based community
intervention. 21(2):54–61.
Belle-Isle, Lynne
Belle-Isle L. Genetics and public health in the 21st century:
using genetic information to improve health and prevent
disease [book review]. 21(4):157–8.
Berman, Neil J
Luciani S, Berman NJ. Canadian Strategy for Cancer Control
[status report]. 21(1):23–5.
Boscaino, Anna Maria
Foster B, Boscaino AM, for the CCOCS Management
Committee. Canadian Coalition on Cancer Surveillance
[status report]. 21(1):26–9.
Brandon-Christie, Jennifer
Patten SB, Brandon-Christie J, Devji J, Sedmak B.
Performance of the composite international diagnostic
interview short form for major depression in a community
sample. 21(2):68–72.
Brown, K Stephen
Pickett W, Koushik A, Faelker T, Brown KS. Estimation of
youth smoking behaviours in Canada. 21(3):119–27.
Buchan, Giao
Cotterchio M, McKeown-Eyssen G, Sutherland H, Buchan G,
Aronson M, Easson AM, Macey J, Holowaty E,
Gallinger S. Ontario familial colon cancer registry:
methods and first-year response rates. 21(2):81–6.
Campbell, Karen
Deonandan R, Campbell K, Ostbye T, Tummon I,
Robertson J. A comparison of methods for measuring
socio-economic status by occupation or postal area.
21(3):114–8.
Carlow, Donald R
Carlow DR. Canadian Association of Provincial Cancer
Agencies [status report]. 21(1):30–3.
2001
Corey, Paul
Manuel DG, Goel V, Williams JI, Corey P. Health-adjusted life
expectancy at the local level in Ontario. 21(2):73–80.
Cotterchio, Michelle
Cotterchio M, McKeown-Eyssen G, Sutherland H, Buchan G,
Aronson M, Easson AM, Macey J, Holowaty E, Gallinger
S. Ontario familial colon cancer registry: methods and
first-year response rates. 21(2):81–6.
Cruess, Alan F
Maberley DAL, King W, Cruess AF. The prevalence of
diabetes in the Cree of western James Bay.
21(3):128–33.
Cyr, Margaret
Fair M, Cyr M, Allen AC, Wen SW, Guyon G, MacDonald RC,
for the Fetal and Infant Health Study Group. An
assessment of the validity of a computer system for
probabilistic record linkage of birth and infant death
records in Canada. 21(1):8–13.
Demissie, Kitaw
Demissie K, Hanley JA, Menzies D, Joseph L, Ernst P.
Agreement in measuring socio-economic status:
area-based versus individual measures. 21(1):1–7.
Deonandan, Raywat
Deonandan R, Campbell K, Ostbye T, Tummon I,
Robertson J. A comparison of methods for measuring
socio-economic status by occupation or postal area.
21(3):114–8.
Devji, Jennifer
Patten SB, Brandon-Christie J, Devji J, Sedmak B.
Performance of the composite international diagnostic
interview short form for major depression in a community
sample. 21(2):68–72.
Dzakpasu, Susie
Wen SW, Kramer MS, Liu S, Dzakpasu S, Sauvé R, for the
Fetal and Infant Health Study Group. Infant mortality by
gestational age and birth weight in Canadian provinces
and territories, 1990–1994 births. 21(1):14–22.
Easson, Alexandra M
Cotterchio M, McKeown-Eyssen G, Sutherland H, Buchan G,
Aronson M, Easson AM, Macey J, Holowaty E, Gallinger
S. Ontario familial colon cancer registry: methods and
first-year response rates. 21(2):81–6.
Ernst, Pierre
Demissie K, Hanley JA, Menzies D, Joseph L, Ernst P.
Agreement in measuring socio-economic status:
area-based versus individual measures. 21(1):1–7.
Faelker, Taron
Pickett W, Koushik A, Faelker T, Brown KS. Estimation of
youth smoking behaviours in Canada. 21(3):119–27.
37
Fair, Martha
Fair M, Cyr M, Allen AC, Wen SW, Guyon G, MacDonald RC,
for the Fetal and Infant Health Study Group. An
assessment of the validity of a computer system for
probabilistic record linkage of birth and infant death
records in Canada. 21(1):8–13.
Ferrence, Roberta
Ferrence R, Stephens T. Monitoring tobacco use in Canada:
the need for a surveillance strategy [commentary].
21(2):50–3.
Foster, Barbara
Foster B, Boscaino AM, for the CCOCS Management
Committee. Canadian Coalition on Cancer Surveillance
[status report]. 21(1):26–9.
Gallinger, Steven
Cotterchio M, McKeown-Eyssen G, Sutherland H, Buchan G,
Aronson M, Easson AM, Macey J, Holowaty E, Gallinger
S. Ontario familial colon cancer registry: methods and
first-year response rates. 21(2):81–6.
Goel, Vivek
Manuel DG, Goel V, Williams JI, Corey P. Health-adjusted life
expectancy at the local level in Ontario. 21(2):73–80.
Guyon, Grace
Fair M, Cyr M, Allen AC, Wen SW, Guyon G, MacDonald RC,
for the Fetal and Infant Health Study Group. An
assessment of the validity of a computer system for
probabilistic record linkage of birth and infant death
records in Canada. 21(1):8–13.
Hanley, James A
Demissie K, Hanley JA, Menzies D, Joseph L, Ernst P.
Agreement in measuring socio-economic status:
area-based versus individual measures. 21(1):1–7.
Holowaty, Eric
Cotterchio M, McKeown-Eyssen G, Sutherland H, Buchan G,
Aronson M, Easson AM, Macey J, Holowaty E, Gallinger
S. Ontario familial colon cancer registry: methods and
first-year response rates. 21(2):81–6.
Johnson, Kenneth C
Johnson KC. National Enhanced Cancer Surveillance System:
a federal–provincial collaboration to examine
environmental cancer risks [status report]. 21(1):34–5.
Joseph, Lawrence
Demissie K, Hanley JA, Menzies D, Joseph L, Ernst P.
Agreement in measuring socio-economic status:
area-based versus individual measures. 21(1):1–7.
Kaiserman, Murray J
Stephens T, Kaiserman MJ, McCall DJ, Sutherland-Brown C.
School-based smoking prevention: economic costs
versus benefits. 21(2):62–7.
King, Will
Maberley DAL, King W, Cruess AF. The prevalence of
diabetes in the Cree of western James Bay.
21(3):128–33.
Koushik, Anita
Pickett W, Koushik A, Faelker T, Brown KS. Estimation of
youth smoking behaviours in Canada. 21(3):119–27.
38
Chronic Diseases in Canada
Koutsavlis, Anthanasios Tom
Koutsavlis AT, Wolfson C. Elements of mobility as predictors
of survival in elderly patients with dementia: findings
from the Canadian study of health and aging.
21(3):93–103.
Kramer, Michael S
Wen SW, Kramer MS, Liu S, Dzakpasu S, Sauvé R, for the
Fetal and Infant Health Study Group. Infant mortality by
gestational age and birth weight in Canadian provinces
and territories, 1990–1994 births. 21(1):14–22.
Kristiansen, Connie
Kristiansen C. Qualitative research methods: a health focus
[book review]. 21(3):137–8.
Laundry, Brian
Bains N, Pickett W, Laundry B, Mecredy D. Predictors of
smoking cessation in an incentive-based community
intervention. 21(2):54–61.
Liu, Shiliang
Wen SW, Kramer MS, Liu S, Dzakpasu S, Sauvé R, for the
Fetal and Infant Health Study Group. Infant mortality by
gestational age and birth weight in Canadian provinces
and territories, 1990–1994 births. 21(1):14–22.
Luciani, Silvana
Luciani S, Berman NJ. Canadian Strategy for Cancer Control
[status report]. 21(1):23–5.
Maberley, David AL
Maberley DAL, King W, Cruess AF. The prevalence of
diabetes in the Cree of western James Bay.
21(3):128–33.
MacDonald, Ralph C
Fair M, Cyr M, Allen AC, Wen SW, Guyon G, MacDonald RC,
for the Fetal and Infant Health Study Group. An
assessment of the validity of a computer system for
probabilistic record linkage of birth and infant death
records in Canada. 21(1):8–13.
Macey, Jeannette
Cotterchio M, McKeown-Eyssen G, Sutherland H, Buchan G,
Aronson M, Easson AM, Macey J, Holowaty E, Gallinger
S. Ontario familial colon cancer registry: methods and
first-year response rates. 21(2):81–6.
MacNeill, Ian B
MacNeill IB. Quantitative estimation and prediction of human
cancer risks (IARC Scientific Publications No. 131)
[book review]. 21(1):37–8.
Manuel, Douglas G
Manuel DG, Goel V, Williams JI, Corey P. Health-adjusted life
expectancy at the local level in Ontario. 21(2):73–80.
Mao, Yang
On L, Semenciw RM, Mao Y. Orius software: calculation of
rates and epidemiologic indicators, and preparation of
graphical output [status report]. 21(3):134–6.
Marrett, Loraine D
Marrett LD, Theis B, Ashbury FD, and an Expert Panel.
Workshop report: physical activity and cancer
prevention. 21(4):143–9.
Vol 22, No 1
McCall, Douglas J
Stephens T, Kaiserman MJ, McCall DJ, Sutherland-Brown C.
School-based smoking prevention: economic costs
versus benefits. 21(2):62–7.
McKeown-Eyssen, Gail
Cotterchio M, McKeown-Eyssen G, Sutherland H, Buchan G,
Aronson M, Easson AM, Macey J, Holowaty E,
Gallinger S. Ontario familial colon cancer registry:
methods and first-year response rates. 21(2):81–6.
Mecredy, Darlene
Bains N, Pickett W, Laundry B, Mecredy D. Predictors of
smoking cessation in an incentive-based community
intervention. 21(2):54–61.
Menzies, Dick
Demissie K, Hanley JA, Menzies D, Joseph L, Ernst P.
Agreement in measuring socio-economic status:
area-based versus individual measures. 21(1):1–7.
Mustard, Cam
Mustard C. Social epidemiology [book review]. 21(3):138–9.
Neutel, C Ineke
Neutel CI, Walop W. Comparing two different approaches to
measuring drug use within the same survey.
21(4):150–6.
On, Long
On L, Semenciw RM, Mao Y. Orius software: calculation of
rates and epidemiologic indicators, and preparation of
graphical output [status report]. 21(3):134–6.
Ostbye, Truls
Deonandan R, Campbell K, Ostbye T, Tummon I, Robertson
J. A comparison of methods for measuring
socio-economic status by occupation or postal area.
21(3):114–8.
Pampalon, Robert
Pampalon R, Raymond G. A deprivation index for health and
welfare planning in Quebec. 21(3):104–13.
Patten, Scott B
Patten SB, Brandon-Christie J, Devji J, Sedmak B.
Performance of the composite international diagnostic
interview short form for major depression in a community
sample. 21(2):68–72.
Pickett, William
Bains N, Pickett W, Laundry B, Mecredy D. Predictors of
smoking cessation in an incentive-based community
intervention. 21(2):54–61.
Pickett W, Koushik A, Faelker T, Brown KS. Estimation of
youth smoking behaviours in Canada. 21(3):119–27.
Raymond, Guy
Pampalon R, Raymond G. A deprivation index for health and
welfare planning in Quebec. 21(3):104–13.
Robertson, James
Deonandan R, Campbell K, Ostbye T, Tummon I, Robertson
J. A comparison of methods for measuring
socio-economic status by occupation or postal area.
21(3):114–8.
2001
Sauvé, Reg
Wen SW, Kramer MS, Liu S, Dzakpasu S, Sauvé R, for the
Fetal and Infant Health Study Group. Infant mortality by
gestational age and birth weight in Canadian provinces
and territories, 1990–1994 births. 21(1):14–22.
Sedmak, Brandy
Patten SB, Brandon-Christie J, Devji J, Sedmak B.
Performance of the composite international diagnostic
interview short form for major depression in a community
sample. 21(2):68–72.
Semenciw, Robert M
On L, Semenciw RM, Mao Y. Orius software: calculation of
rates and epidemiologic indicators, and preparation of
graphical output [status report]. 21(3):134–6.
Shaw, Amanda
Shaw A. Epidemiology of childhood cancer (IARC Scientific
Publications No. 149) [book review]. 21(1):38–9.
Stephens, Thomas
Ferrence R, Stephens T. Monitoring tobacco use in Canada:
the need for a surveillance strategy [commentary].
21(2):50–3.
Stephens T, Kaiserman MJ, McCall DJ, Sutherland-Brown C.
School-based smoking prevention: economic costs
versus benefits. 21(2):62–7.
Stewart, Paula J
Stewart PJ. Health promotion planning: an educational and
ecological approach (third edition) [book review].
21(1):36–7.
Sutherland, Heather
Cotterchio M, McKeown-Eyssen G, Sutherland H, Buchan G,
Aronson M, Easson AM, Macey J, Holowaty E,
Gallinger S. Ontario familial colon cancer registry:
methods and first-year response rates. 21(2):81–6.
Sutherland-Brown, Carol
Stephens T, Kaiserman MJ, McCall DJ, Sutherland-Brown C.
School-based smoking prevention: economic costs
versus benefits. 21(2):62–7.
Theis, Beth
Marrett LD, Theis B, Ashbury FD, and an Expert Panel.
Workshop report: physical activity and cancer
prevention. 21(4):143–9.
Tummon, Ian
Deonandan R, Campbell K, Ostbye T, Tummon I,
Robertson J. A comparison of methods for measuring
socio-economic status by occupation or postal area.
21(3):114–8.
Walop, Wikke
Neutel CI, Walop W. Comparing two different approaches to
measuring drug use within the same survey.
21(4):150–6.
Wen, Shi Wu
Fair M, Cyr M, Allen AC, Wen SW, Guyon G, MacDonald RC,
for the Fetal and Infant Health Study Group. An
assessment of the validity of a computer system for
probabilistic record linkage of birth and infant death
records in Canada. 21(1):8–13.
39
Wen SW, Kramer MS, Liu S, Dzakpasu S, Sauvé R, for the
Fetal and Infant Health Study Group. Infant mortality by
gestational age and birth weight in Canadian provinces
and territories, 1990–1994 births. 21(1):14–22.
White, Franklin
White F. The burden of disease among the global poor:
current situation, future trends, and implications for
strategy [book review]. 21(2):87–8.
Willis , Alette
Willis A. Spatial epidemiology: methods and applications
[book review]. 21(4):159–60.
Wolfson, Christina
Koutsavlis AT, Wolfson C. Elements of mobility as predictors
of survival in elderly patients with dementia: findings
from the Canadian study of health and aging.
21(3):93–103.
Williams, J Ivan
Manuel DG, Goel V, Williams JI, Corey P. Health-adjusted life
expectancy at the local level in Ontario. 21(2):73–80.
40
Chronic Diseases in Canada
Vol 22, No 1
CDIC: Information for Authors
Chronic Diseases in Canada (CDIC) is a peer-reviewed
scientific journal published four times a year. Contributions
are welcomed from outside of Health Canada as well as
from within this federal department. The journal’s focus is
the prevention and control of non-communicable diseases
and injuries in Canada. This may include research from
such fields as epidemiology, public/community health,
biostatistics, behavioural sciences and health services.
CDIC endeavours to foster communication about chronic
diseases and injuries among public health practitioners,
epidemiologists and researchers, health policy planners
and health educators. Submissions are selected based on
scientific quality, public health relevance, clarity, conciseness and technical accuracy. Although CDIC is a Health
Canada publication, authors retain responsibility for the
contents of their papers, and opinions expressed are not
necessarily those of the CDIC Editorial Committee or of
Health Canada.
FEATURE ARTICLES
Regular Feature Articles: Maximum 4,000 words for
main text body (excluding abstract, tables, figures,
references) in the form of original research, surveillance
reports, meta-analyses, methodological papers, literature
reviews or commentaries
Short Reports: Maximum 1,200 words (as above)
Status Reports: Describe ongoing national programs,
studies or information systems at Health Canada (maximum
3,000 words)
Workshop/Conference Reports: Summarize workshops,
etc. organized or sponsored by Health Canada (maximum
3,000 words)
Cross-country Forum: For authors outside of Health
Canada to exchange information from research or
surveillance findings, programs under development or
program evaluations (maximum 3,000 words)
ADDITIONAL ARTICLE TYPES
Letters to the Editor: Comments on articles recently
published in CDIC will be considered for publication
(maximum 500 words)
Book/Software Reviews: Usually solicited by the editors
(500–1,300 words), but requests to review are welcomed
SUBMITTING MANUSCRIPTS
Submit manuscripts to the Editor-in-Chief, Chronic
Diseases in Canada, Population and Public Health Branch,
Health Canada, Tunney’s Pasture, CDIC Address Locator:
0602C3, Ottawa, Ontario K1A 0L2.
Since CDIC adheres in general (section on illustrations not
applicable) to the “Uniform Requirements for
Manuscripts Submitted to Biomedical Journals” as
approved by the International Committee of Medical
Journal Editors, authors should refer to this document for
complete details before submitting a manuscript to CDIC
(see <www.cma.ca/publications/mwc/uniform.htm> or
Can Med Assoc J 1997;156(2):270–7).
Checklist for Submitting Manuscripts
G Cover letter: Signed by all authors, stating that all
have seen and approved the final manuscript and
have met the authorship criteria of the Uniform
Requirements and including a full statement
regarding any prior or duplicate publication or
submission for publication
G First title page: Concise title; full names of all
authors and institutional affiliations; name, postal
and e-mail addresses, telephone and fax numbers
for corresponding author; separate word counts for
abstract and text
G Second title page: Title only; start page
numbering here as page 1
G Abstract: Unstructured (one paragraph, no
headings), maximum 175 words (100 for short
reports); include 3–8 key words (preferably from
the Medical Subject Headings (MeSH) of Index
Medicus)
G Text: Double-spaced, 1 inch (25 mm) margins,
12 point font size
G Acknowledgements: Include disclosure of finan-
cial and material support in acknowledgements;
if anyone is credited in acknowledgements with
substantive scientific contributions, authors should
state in cover letter that they have obtained written
permission
G References: In “Vancouver style” (consult
Uniform Requirements and a recent CDIC issue for
examples); numbered in superscript (or within
parentheses) in the order cited in text, tables and
figures; listing up to 6 authors (first 3 and “et al.” if
more); without any automatic reference numbering
feature used in word processing; any unpublished
observations/ data or personal communications
used (discouraged) to be cited in the text in
parentheses (authors responsible for obtaining
written permission); authors are responsible for
verifying accuracy of references
G Tables and Figures: Each on a separate page and
in electronic file(s) separate from the text (not
imported into the text body); as self-explanatory
and succinct as possible; not duplicating the text,
but illuminating and supplementing it; not too
numerous; numbered in the order that they are
mentioned in the text; explanatory material for
tables in footnotes, identified by lower-case
superscript letters in alphabetical order; figures
limited to graphs or flow charts/templates (no
photographs), with software used specified and
titles/footnotes on a separate page
G Number of copies: Four complete copies,
including tables and figures; 2 copies of any related
supplementary material
Chronic Diseases in Canada
a publication of the
Population and Public Health Branch
Health Canada
Acting Editor-in-Chief .................. Debby Baker
Scientific Editor ....................... Christina J Mills
Associate Scientific Editor ............. Gerry B Hill
Associate Scientific Editor ........ Stephen B Hotz
Associate Scientific Editor .......Robert A Spasoff
Assistant English Editor ..............Marion Pogson
Assistant French Editor ...................Pamela Fitch
Desktop Publisher ................... Robert Friedman
CDIC Editorial Committee
Donald T Wigle, Committee Chair
Healthy Environments and Consumer
Safety Branch
Health Canada
Jean-François Boivin
McGill University
Chronic Diseases in Canada (CDIC) is a quarterly scientific journal
focusing on current evidence relevant to the control and prevention of
chronic (i.e. non-communicable) diseases and injuries in Canada. The
journal publishes a unique blend of peer-reviewed feature articles by
authors from the public and private sectors that may include research
from such fields as epidemiology, public/community health, biostatistics,
behavioural sciences and health services. Authors retain responsibility
for the contents of their articles; the opinions expressed are not
necessarily those of the CDIC Editorial Committee or of Health Canada.
Jacques Brisson
Université Laval
Neil E Collishaw
Physicians for a Smoke-Free Canada
Subscription is free upon request.
When submitting change of address,
please enclose your old address label.
Mailing Address:
Clyde Hertzman
University of British Columbia
C Ineke Neutel
Élisabeth-Bruyère Research Dept
Sisters of Charity of Ottawa Health
Service Inc.
Kathryn Wilkins
Health Statistics Division
Statistics Canada
Chronic Diseases in Canada
Population and Public Health Branch
Health Canada, Tunney's Pasture
Address Locator: 0602C3
Ottawa, Ontario K1A 0L2
James A Hanley
McGill University
Telephone:
Editor-in-Chief (613) 957-1767
Scientific Editor (613) 957-2624
Fax
(613) 952-7009
E-mail:
[email protected]
Indexed in Index Medicus/MEDLINE, PAIS
(Public Affairs Information Service) and
EMBASE, the Excerpta Medica database.
This publication is also available online at
<http://www.hc-sc.gc.ca/hpb/lcdc>

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project