Putting Away the Stethoscope for Good? Raymond W. Pong, PhD

Putting Away the Stethoscope for Good? Raymond W. Pong, PhD
Putting Away the Stethoscope for Good?
Toward a New Perspective on Physician Retirement
Raymond W. Pong, PhD
Centre for Rural and Northern Health Research, Laurentian University
Spending and Health Workforce
Who We Are
Established in 1994, CIHI is
an independent, not-for-profit
corporation that provides essential
information on Canada’s health
system and the health of Canadians.
Funded by federal, provincial and
territorial governments, we are guided
by a Board of Directors made up
of health leaders across the country.
Our Vision
To help improve Canada’s health
system and the well-being of Canadians
by being a leading source of unbiased,
credible and comparable information
that will enable health leaders to make
better-informed decisions.
Table of Contents
List of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii
Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Chapter 1: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 The Context. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Organization of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Chapter 2: Physician Retirement: Known, Unknown and Questions
2.1 How Do We Know Whether Someone Is Retired? . . . . . . . . .
2.2 What Is Known About Physician Retirement? . . . . . . . . . . . .
2.3 Data Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. .5
. .6
. .8
. 11
Chapter 3: Data Sources . . . . . . . . . . . . . . . . . . . .
3.1 2007 National Physician Survey . . . . . . . . .
3.2 Scott’s Medical Database . . . . . . . . . . . . . .
3.3 National Physician Database . . . . . . . . . . .
3.4 Canadian Medical Association Master File .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 15
. 16
. 17
. 17
. 18
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 4: Extent of Physician Retirement . . . . . . . . . . . . . . . . . . . . . . . . . 19
Chapter 5: Older Physicians Who Were Minimally Active . . . . . . . . . . . . . . 29
Chapter 6: An Alternative Perspective . . .
6.1 Changing Activity Level . . . . . . . .
6.1.1 First Case Study . . . . . . . . . . .
6.2 Changing Clinical Practice Profile
6.2.1 Second Case Study . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 35
. 36
. 38
. 41
. 48
Chapter 7: Discussions and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . 53
7.1 Summary of Major Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
7.2 Limitations and Future Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
7.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Appendix A: Questions Regarding Changes to Medical Practice
in the 2007 National Physician Survey . . . . . . . . . . . . . . . . . . 61
Appendix B: NPDB Physician Specialty Categories . . . . . . . . . . . . . . . . . . 63
Appendix C: Data for Chapter 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Appendix D: Estimating the Number of Older Physicians
Who Were Minimally Active . . . . . . . . . . . . . . . . . . . . . . . . . . 69
Appendix E: Data for Chapter 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
List of Abbreviations
ii
CIHI
Canadian Institute for Health Information
CMA
Canadian Medical Association
FPs/GPs
family physicians/general practitioners
FFS
fee for service
FTE
full-time equivalent
GPs
general practitioners
NPDB
National Physician Database
NPS
National Physician Survey
OHIP
Ontario Health Insurance Plan
SMDB
Scott’s Medical Database
Acknowledgements
Acknowledgements
Without the analytical support, insightful advice, critical reviews and helpful
comments from many people, the successful completion of this study on
physician retirement would not have been possible. The author of the study
is especially grateful to the following staff of the Canadian Institute for Health
Information (CIHI) for performing data extraction and fact-checking and for
their expertise, cooperation and patience:
Mr. Geoff Ballinger, Manager,
Health Human Resources
Mr. Walter Feeney, Acting Program
Lead, Health Human Resources
Mr. Robert Kyte, Program Lead,
Health Human Resources
Mr. Joshua Mirasol, Analyst,
National Physician Database
Ms. Yvonne Rosehart, Program
Lead, Health Human Resources
Other CIHI officials provided useful comments on earlier drafts of the
report, including
Mr. Jean-Marie Berthelot,
Vice President, Programs
Ms. Francine Anne Roy, Director,
Health Spending and Clinical
Registries
Still others worked diligently behind the scene to prepare the study
for publication.
A number of individuals not affiliated with CIHI also played important roles in
ensuring the success of the study:
Ms. Lynda Buske, Director,
Workforce Research, Canadian
Medical Association, who kindly
made data available from the
Canadian Medical Association Master
File, read drafts of the report and
offered helpful insights on many
physician workforce issues
Mr. Artem Safarov, College of Family
Physicians of Canada, who reviewed
sections of the study based on the
2007 National Physician Survey to
ensure accuracy
iii
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
As well, other reviewers of earlier drafts of the report offered valuable
suggestions for improvement:
Dr. Marmoru Watanabe of the
Faculty of Medicine, University of
Calgary
Dr. Elizabeth Wenghofer of the
Centre for Rural and Northern Health
Research, Laurentian University
Dr. Roger Pitblado of the Centre for
Rural and Northern Health Research,
Laurentian University
It should be noted that the interpretations contained in the study are those
of the author and do not necessarily reflect the opinions of the reviewers or the
organizations they belong to. Likewise, the reviewers and their organizations
are not responsible for any shortcoming the study may have.
Finally, the author wishes to thank Mr. Vince Guerin of the Centre for Rural
and Northern Health Research, Laurentian University, for preparing the tables
and graphs.
iv
Executive Summary
Executive Summary
Health workforce issues, especially the future supply of physicians and other
health workers, have become a major preoccupation among health policymakers and administrators. The apprehension about not having enough
physicians to meet future medical care needs has been reinforced by the
realization that the medical workforce is aging, with the possibility that many
physicians will exit the medical workforce in the coming years. It is commonly
assumed that as more and more physicians approach the traditional
retirement age of 65, the number of doctors retiring will grow. But does this
mean that most physicians will put away their stethoscopes for good at age 65?
This study is an attempt to understand how aging affects physicians’ work,
including staying in or leaving clinical practice. The study begins with a
review of the pertinent literature, which seeks to find out what is known about
retirement in general and physician retirement in particular. This is then
followed by three interrelated sets of empirical analysis: estimating the extent
of physician retirement; estimating the number of older physicians who are
minimally active and could, therefore, be considered retired from a health
workforce planning perspective; and exploring an alternative approach to
understanding how aging affects physicians’ clinical practice.
The study uses data from several sources—the 2007 National Physician
Survey (NPS), Scott’s Medical Database (SMDB), the National Physician
Database (NPDB) and the Canadian Medical Association (CMA) Master
File—to paint a composite and more complete picture of the practice profile
of older physicians. This is also an attempt to ensure that the findings are
based on multiple sources of information because, without a uniform definition
of retirement, each database may have captured retirement information
somewhat differently and counted the number of active or retired physicians
somewhat differently. The study also examines both retirement intentions and
behaviours, without assuming that they are the same phenomenon.
Do physicians typically retire when they turn 65? Are physicians increasingly
opting for early retirement? Evidence from existing studies suggests that
Canadian physicians tend to quit work later than average workers. Also, as
far as physicians are concerned, retirement is anything but an either/or issue.
Instead of dropping out of the medical workforce abruptly and completely at
age 65, many older physicians choose to remain in clinical practice, though
they do not necessarily maintain the same activity level or do the same kind
of work as when they were younger.
Depending on the source of data used, one gets different rates of physician
retirement. Most likely this is because various databases define retirement
differently, count the number of retired physicians differently and/or have
v
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
different criteria for including physicians in or excluding them from the base
population. On the basis of self-reported retirement intentions obtained from
the 2007 NPS, about 3.2% of all physicians planned to retire in each of the two
years following the survey. On the other hand, the estimated average annual
retirement rates were 0.54%, based on three years of SMDB data (including
semi-retirement), and 0.79%, based on data from the CMA Master File for the
same three-year period. The estimated retirement rate based on retirement
intentions (using 2007 NPS data) was substantially higher than those derived
from administrative databases (using NPDB and CMA Master File data). If one
uses these estimated retirement rates as projection parameters to forecast
the size of the medical workforce 25 or 30 years into the future, one is likely to
obtain substantially different workforce projections, when all other variables
are held constant. On the basis of these projections, one could come to very
different conclusions about the future sufficiency of physician supply in Canada.
An attempt was then made to estimate the number of older physicians
who were minimally active in clinical practice. Depending on what “older
physicians” refers to and where the full-time equivalent (FTE) threshold is set,
different proportions of older physicians could be considered minimally active.
For 2007, if the FTE threshold was set at 33% or less of previous workload, the
proportion of physicians considered minimally active would range from 7.3%
of physicians age 55 and older to 11.9% of physicians age 65 and older. If the
FTE threshold was set at 15% or less of previous workload, the range of those
considered minimally active would be 3.3% to 4.9% for physicians age 55 and
older and those age 65 and older, respectively. Whether these minimally active
physicians should be considered retired for the purpose of medical workforce
projections or planning is not just a technical issue but also a policy matter
that needs further consideration and deliberation by stakeholders.
Because retirement is a fuzzy concept, especially for physicians, and since
there is as yet no consensus on what physician retirement means and
how it should be measured, this study has suggested a different way of
understanding how aging affects the way physicians work. If retirement is
understood to mean the complete cessation of medical practice, it should
be seen as the end point of a continuum of changes in medical practice as
a physician gets older. But prior to exiting the medical workforce through
retirement or death, many other changes in medical practice may have taken
place, such as reduction in workload, scope-of-practice compression or
greater involvement in non-clinical work, which may also have implications
for medical care provision and physician workforce planning.
vi
Executive Summary
While physicians in the baby-boom generation will exit the medical workforce
in greater numbers in the coming years, many physicians age 65 and older
are likely to remain active in clinical practice, based on trends from the recent
past. But their workload, as measured by FTE values from physicians’ fee-forservice payments, is likely to decline as they become older. Also, their scope
of practice will tend to narrow, as older physicians relinquish some types
of clinical work while retaining others. The case of older family physicians/
general practitioners (FPs/GPs) was used as an illustration. Although there
were no major differences between FPs/GPs in different age groups with
respect to such core clinical activities as office assessment and mental health
care, the older the FPs/GPs became, the less likely they were to engage in
such activities as hospital inpatient care, obstetrics, anesthesia and services
requiring advanced procedural skills.
The study concludes by examining the implications of the findings from
a physician workforce planning perspective and by identifying several
knowledge gaps. It argues that knowing what older physicians do and how
much they do is just as important as figuring out how many doctors retire
each year. The latter task will continue to be a challenge as administrative
rules and social norms regarding retirement become increasingly fluid. Also,
the task of deciding who is or is not retired will not be easy until there is an
agreed-upon definition of retirement and until the right kinds of data are
collected for analysis.
vii
Chapter 1
Introduction
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
1.1 The Context
Population aging has been getting a lot of attention. We have been told that
the Canadian population, similar to populations in other developed nations,
is getting older. This has a lot to do with the post–World War II baby boom,
the subsequent baby bust and the fact that people are living longer than
in previous generations. The very first wave of the baby-boom cohort is
expected to reach the age of 65 by 2011, and the size of the population 65
and older will be considerably larger in the foreseeable future. Although
longer life expectancy is a desirable development, population aging will affect
Canadian society in many ways, ranging from expected declining crime rates
and changing consumer preferences to potential labour shortages, rejigged
government spending priorities and shifting social values.
Concerns expressed by health policy-makers about population aging
have mostly been on two interrelated issues: expected rising demands for
health and medical care, particularly in the areas of long-term care, chronic
disease management and disability; and possible shortages of health care
personnel. It is a common belief that elderly people tend to use health
services more heavily than younger people.1, 2 At the same time, the health
workforce is getting older and there will be greater attrition as a result of
retirement and death. There is also a concern that Canadian workers, including
those in the health care sector, are increasingly inclined to opt for early
retirement—the Freedom 55 phenomenon. Will there be enough doctors,
nurses, physiotherapists and other practitioners to look after the rapidly
growing number of increasingly frail and sick senior citizens? Not surprisingly,
physicians have been the focus of much of the attention and debate. Mass
media reports about family physicians not accepting new patients and
long waits for specialist services are increasingly common and have been
compounded by a growing unease about the aging of the medical workforce.
The public has been reminded that physicians age 65 and older will make
up 20% of the medical workforce in 2026.3
Concerns about physicians giving up medical practice due to retirement
appear to be a fairly recent phenomenon in Canada. From the early 1980s
to the early 1990s, there were worries, at least among federal and provincial
ministries of health, about an over-supply of physicians, resulting in the
introduction of various measures to regulate their numbers. While many of
those measures targeted medical students (such as reductions in medical
school enrolment), some were designed to trim the number of aging doctors.
Several provinces, such as New Brunswick, Nova Scotia and Quebec, offered
retirement buyout packages for older physicians, and British Columbia
introduced mandatory retirement for physicians at age 75.1, 4 But there
was a change of official position by the late 1990s. The more recent view
is that Canada needs more physicians to meet present and future medical
2
Chapter 1: Introduction
care needs. This is reflected by a substantial expansion of medical school
enrolment in recent years (including the opening of a new medical school
in northern Ontario in 2005), attempts to repatriate Canadian physicians who
have gone abroad and efforts to enable more foreign medical graduates
to practise medicine in Canada.
These demographic trends and emerging public perceptions have caused
considerable unease among policy-makers and health human resources
planners. For instance, various medical organizations have attempted to raise
public awareness about the implications of large-scale physician retirement.
In releasing the results of the 2004 National Physician Survey (NPS), the
Royal College of Physicians and Surgeons of Canada, the College of Family
Physicians of Canada and the Canadian Medical Association (CMA) drew
attention to the issue in a joint press release:
The NPS identified two other significant shifts in the physician population
that are changing the face of medicine in the country. First, a large number
of physicians are reaching retirement. If the survey data is translated to the
physician population as a whole, as many as 3,800 doctors plan to retire
entirely in the next two years alone. This is more than double the current
rate of retirement.5
The release of the 2007 NPS results prompted similar expressions of concern:
First, 6% of NPS 2007 respondents plan to retire from clinical practice
over the next 2 years as the baby-boom generation begins leaving the
work force with a vengeance . . . If 4,000 physicians retire, the number
of new physicians the country is producing will barely be large enough
to replace them.6
Are such concerns justified? What do we know about physician retirement in
Canada? Do most physicians quit working at the traditional retirement age of
65? Are more and more physicians opting for early retirement? Do physicians’
retirement patterns mirror those of other workers? This study is an attempt to
better understand physician retirement in Canada, a phenomenon that has not
received policy and research attention commensurate with growing concerns
about the future supply of physicians.
3
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
1.2 Organization of the Study
The next chapter discusses what retirement means, particularly where
physicians are concerned. Are physician retirement patterns similar to those
of other Canadian workers? By means of a review of the relevant literature,
Chapter 2 examines what is known about physician retirement and discusses
whether conventional notions of retirement are applicable to medical
practitioners. It should be noted that although physicians do a variety of
things, the focus of this study is on the provision of clinical services and the
potential impact aging of the physician workforce may have on medical care.
(Moreover, existing physician databases have little or no information on other
aspects of physicians’ work like education, research and administration.)
The data analysis part of the study begins in Chapter 3 with a description
of the sources of data used in the analysis. This is followed in Chapter 4 by
an examination of physician retirement, using secondary data from several
sources. The purpose is to determine whether different sources of data
yield consistent estimates of physician retirement. As will be shown later on,
many physicians continue to work past age 65, albeit at lower activity levels.
Thus, Chapter 5 is an attempt to estimate the number of older physicians
who are minimally active. The question is whether they should be counted
as retired. The study then explores, in Chapter 6, an alternative approach
to understanding physicians’ transition from active to non-active status as
a result of aging. It is argued that this new perspective represents a more
realistic understanding of the impact of aging on the physician workforce
and a more viable workforce planning approach. Bringing the study to a
close, Chapter 7 summarizes the major findings, points out the limitations of
the study and concludes by noting the implications of the study and possible
future actions.
In the following chapters, data is presented mostly in the form of graphs.
Tables containing more detailed data appear in the appendices.
4
Chapter 2
Physician Retirement: Known, Unknown
and Questions
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
2.1 How Do We Know Whether
Someone Is Retired?
There is not a universally accepted or officially sanctioned definition of
retirement in Canada. It is generally assumed that most Canadians retire at
age 65 but, as will be shown later on, this assumption has increasingly been
called into question. Consequently, measuring retirement is not as simple a
task as one might think. Those studying retirement are likely to ask, as Gower
did, “How does one decide who is retired and who is not? Is it necessary
to be in receipt of a pension? Can a person who has a part-time job still be
considered retired?”7 Retirement has been variously defined. For instance,
some understand retirement as complete withdrawal from the labour force.
Deschênes and Stone consider a person retired when he or she leaves the
labour market for good and receives retirement income from the Canada
Pension Plan, the Quebec Pension Plan or a private pension plan.8 According
to Bowlby, Statistics Canada has a standard definition of retirement, which
refers to someone who is age 55 or older, is not in the labour force and
receives 50% or more of his or her total income from retirement-like sources.9
Others are somewhat more lenient and willing to consider a person retired
when his or her gainful employment is reduced. But reduced to what level?
In a comprehensive review of retirement studies, Denton and Spencer
noted different criteria or measures used to determine retirement.10, 11 One
is the complete absence of labour force participation. A person could
also be considered retired when time worked (and, by extension, income
earned) is reduced. Another possibility is being in receipt of retirement
income. Alternatively, retirement status can be based on self-assessment.
If individuals describe themselves as retired, they are retired, regardless of
what others think. The authors have grouped retirement definitions into two
major categories: those that are based on a single criterion and those that
use multiple criteria to determine whether a person is retired. Among the
former, the defining characteristics most commonly used are non-participation
in the labour force, a reduction in work hours or income and self-assessed
retirement status. Most definitions based on multiple characteristics include
receipt of pension income in combination with earnings or hours worked
below a specified threshold or a reduction in labour force participation,
including non-participation. Needless to say, the way retirement is defined
affects the size of the retired population.
As noted earlier, there are concerns that Canadians are retiring at an
increasingly younger age. A study by Ibbott, Kerr and Beaujot examined
declining labour force participation among older Canadians.12 Some research
has documented a gradual decline in the median retirement age of Canadian
workers in recent decades, from 64.9 between 1976 and 1980, to 62.2
6
Chapter 2: Physician Retirement: Known, Unknown and Questions
between 1991 and 1995, to 61.0 between 1996 and 2000.13 Gower has
reported similar trends based on his research.7 However, there are indications
that the trend may have shifted in the last several years. A more recent study
by Schellenberg and Ostrovsky, which compared retirement intentions using
data from the 1991 Survey of Aging and Independence and the 2002 and 2007
General Social Surveys, showed that older workers are pushing back their
retirement plans.14 Between 1991 and 2007, the proportion of Canadians age
45 to 49 planning to retire before age 60 decreased by four percentage points,
while the proportion planning to retire at age 65 or older increased by about
seven percentage points. Similar patterns are found among those age 50 to 54
but not among those age 55 to 59. The authors also noted that evidence from
the Labour Force Survey points in the same direction: the average retirement
age of male employees in the private sector reached a low of 61.4 in 2000, and
then rose to 62.3 by 2007. The average retirement age of female employees
increased from 60.7 to 61.7 in the same period. This suggests that declining
age of retirement may not be an inexorable trend. A host of factors may shape
how people view retirement and their retirement decisions.
Questions concerning what retirement means for physicians are even more
difficult to answer, for several reasons. First of all, compulsory retirement
requirements are gradually becoming a thing of the past. As early as 1978,
the General Council of the CMA passed a resolution urging the abolition
of mandatory retirement at age 65.15 In 1982, the Manitoba Court of Appeal
ruled unanimously that the bylaw requiring physicians to retire at age 65
was invalid because it violated the provisions of the Manitoba Human Rights
Act on age discrimination in employment.15 In December 2006, Ontario’s
Ending Mandatory Retirement Statute Law Amendment Act, 2005 became law
and the definition of age in the Ontario Human Rights Code was amended
to prohibit discrimination against an employee who is 65 or older.16 As a
result, mandatory retirement bylaws and policies in relation to physicians
are now considered a human rights violation. Manitoba and Quebec have
also abolished compulsory retirement requirements, and other provinces are
likely to follow suit. What all this means is that one can no longer assume that
workers, including physicians, automatically retire when they turn 65. i
Second, self-employed persons often exhibit retirement patterns that are quite
different from those of employees. According to Bahrami and associates, at
any given age, self-employed individuals are more likely than wage-earners
to continue working full time and, when they do decide to retire, are less
likely to leave the labour force in one move.17 Gower has likewise reported
that self-employed people tend to retire later than employees and that those
with unincorporated businesses are likely to be the last to retire.7 Since most
i.
It should be noted that hospital privileges can still be denied to physicians at any age based
on performance and competence.
7
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
physicians are independent, self-employed practitioners, their retirement
patterns are likely to be more complex than those of salaried workers. This
may have prompted Foot and colleagues to caution that “because there is no
rule that physicians retire at a certain age, such as 65, it is always difficult to
project retirement.”18
Finally, the transition from gainful employment to retirement is not necessarily
a quick one. As Deschênes and Stone have observed, retirement is not so
much an event as a process that may extend over many months or even
years.8 They call this the “transition to retirement” process, which comes
to an end when retiree status is permanently established. Similarly, Denton
and Spencer have pointed out that there is often a long period of time during
which a person can be described as both “retired” and “working.”10 According
to McDaniel, “the transition from employment to retirement . . . is far from the
smooth transition that . . . has long been presumed . . . . Multiple transitions
occur into and out of employment and into and out of the labour force.”19
All this complicates the analysis of retirement behaviours.
Because mandatory retirement at a certain age is no longer a requirement
in some provinces, and since most physicians are self-employed, it is quite
possible that their transition-to-retirement process takes longer or tends
not to follow a set pattern. In other words, the road to retirement for many
physicians may not be a well-charted pathway. This situation may also be true
for an increasing number of other workers as employment conditions and
social institutions become more fluid. Chappell and colleagues described the
retirement scene in the 21st century this way: “In short, retirement does not
always represent an abrupt transition from work to nonwork: it can be gradual,
it can involve multiple exits, and it may never happen.”20 Likewise, it is the view
of Denton and Spencer that “retirement can be voluntary or involuntary;
it can be gradual or sudden; and it can be temporary or permanent.”11
2.2 What Is Known About
Physician Retirement?
The short answer to this question is that we do not know very much. ii To be
more exact, we know quite a bit about the aging of the physician workforce
but not much about physician retirement. Much of the discussion about
physician retirement and its workforce implications involves inferences based
on the changing age structure of the physician population. For instance,
Tyrrell and Dauphinee have predicted that physician retirement will accelerate
ii.
8
This is true not just in Canada but in other parts of the world. The World Health Report 2006,
published by the World Health Organization, notes that “information about the retirement rate
of health workers is very scarce.”21
Chapter 2: Physician Retirement: Known, Unknown and Questions
over the coming years because the number of physicians older than 55 is
expected to increase from about 26% in 1999 to about 43% by 2021.2 Some
studies assume that physicians retire at age 65, with little or no supporting
evidence.22, 23 The rest focus on the legal aspects of mandatory retirement
rules as applied to physicians or the practical aspects of preparing
for retirement.4, 15, 24, 25
If the extent of retirement is strictly a function of the number of physicians
reaching age 65, the task of projecting retirement trends is quite
straightforward, as the changing age structure of the physician population
has been well documented. Chan, for example, pointed out that the physician
workforce was at its most youthful in 1988, when 22% of all physicians were
younger than 35.1 There was a gradual decline in the proportion of physicians
younger than 35 from 1988 to 1993, and a steeper decline from 1993 onwards.
By 2000, only 13% of all physicians were younger than 35. iii Conversely, the
number of physicians age 60 and older has continued to climb. According to
the Canadian Institute for Health Information, the average age of physicians
in Canada increased from 47.0 to 49.6 between 1998 and 2007.26 On average,
female physicians are younger than their male counterparts. During the same
10-year period, the average age of male physicians increased from 49.0 to
51.9, and the average age of female physicians increased from 42.0 to 45.4.
But do physicians typically put away their stethoscopes for good when
they reach age 65? Are physicians more likely to take early retirement in
the coming years? Although there are few systematic studies of physician
retirement in Canada, evidence gleaned from related studies suggests that
physicians tend to retire later than the general working population. There
were indications as early as the mid-1960s that many physicians delayed their
retirement. Studies conducted on behalf of the Royal Commission on Health
Services suggested that the average age of retirement among Canadian
physicians at that time was close to 70.27
In her study on seniors at work in Canada, using data from the 1996 Canadian
census, Duchesne found that 20 occupations accounted for half of the
total employment among workers age 65 and older.13 Farmers and farm
managers made up 17.7% of this total, with 45,205 employed seniors in 1996.
Family physicians and general practitioners (FPs/GPs) were among the 20
occupations, accounting for 1.1% of all seniors at work in 1996. Additionally,
there were 22 occupations with at least 6% of workers age 65 and older in
each of the occupations. Judges topped the list; one in five judges was at
iii.
The decline in the proportion of younger physicians could also be due to the elimination of
rotating internships in the early 1990s. As a result, it takes longer for medical school graduates to
become family physicians, and specialists can no longer practise until they reach full certification
in their specialties.
9
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
least 65 years old. In 1996, there were 1,625 physician specialists age 65 and
older, accounting for 7.6% of all specialists, and there were 2,820 FPs/GPs age
65 and older, accounting for 7.5% of all FPs/GPs in Canada.
According to the 2002 Baseline Study of the CMA, 85% of physicians age 55
to 64 were working full time, 9% were working part time and 4% were retired.
In comparison, 42% of physicians age 65 and older were working full time,
28% were working part time and 28% were retired. 28 Other studies using
different sources of data have come to a similar conclusion: many physicians
are working beyond the traditional retirement age of 65. A study by Chan
showed that the average retirement age of physicians in Canada was 70.8 and
that the retirement age remained relatively constant during the period from
1981 to 2000.1
Analyzing the work patterns of older physicians in Ontario, Trent reported that
older physicians represented a sizeable portion of the medical workforce. 25
Slightly more than 2,000 physicians older than 65 billed the Ontario Health
Insurance Plan (OHIP) for about $225 million in professional fees each year.
About 1,100 physicians between the age of 65 and 69 billed a total of $150
million; the 600 physicians age 70 to 74 billed another $50 million; and the 350
doctors age 75 and older billed OHIP for about $25 million annually. A study
by Chan and associates of physicians who billed fee for service (FFS)
in Ontario showed that in 1995–1996, there were 20,149 physicians in Ontario,
of whom 2,055 (or about 10%) were 65 and older. 29 There were 18,841 full-time
equivalent (FTE) doctors, of whom 1,321 (or 7.0%) were age 65 and older.
These figures suggest that many older physicians were still active in clinical
practice, though they tended to take on a lighter workload (as reflected by the
fact that the 2,055 physicians who billed FFS translated to 1,321 FTEs). iv
A survey of 107 Saskatchewan physicians age 70 and older who were
registered with the College of Physicians and Surgeons of Saskatchewan
found that 93% of the 102 respondents were still medically active in 2004.
Of those 102 physicians, 87% were between age 70 and 79, and 13% were 80
and older. Because of the way the survey was conducted, the study cannot
tell us the proportions of older physicians who had or had not retired, but
it does show that many physicians in Saskatchewan who were past the
traditional retirement age were still active in medical practice.30
Retirement deferral by older physicians is not a uniquely Canadian
phenomenon. Studies have shown that many physicians in the United States
also delay their retirement.27 In the case of Australia, a study by Fletcher and
Schofield showed that most Australian psychiatrists continue to work until late
in life, with only 18% retiring before age 65.31 Another study, by Schofield and
iv.
10
A discussion of full-time equivalence can be found in Section 3.3 in Chapter 3.
Chapter 2: Physician Retirement: Known, Unknown and Questions
Beard, reported that Australian general practitioners tend to work beyond the
traditional retirement age of 65, though they are likely to work fewer hours than
their younger counterparts.32
What are the reasons behind physicians’ desire to postpone retirement?
As noted earlier, studies conducted for the Royal Commission on Health
Services documented postponement of retirement by some doctors; they
further speculated that this might be due to inadequate savings and pension. 27
According to Bahrami et al., those physicians whose identity is closely
linked to their profession may continue to work indefinitely.17 Collier similarly
concluded that some older Canadian doctors who define themselves by
their profession can be notoriously averse to retirement.3 A very old doctor
interviewed by the author was quoted as saying, “I think a doctor is like a
clergyman. This is my calling. As long as I can keep going, I’ll keep going.”3
In 2000, the CMA Physician Resource Questionnaire asked physicians what
might prevent them from retiring at their planned age of retirement. Insufficient
personal savings was the most frequently cited reason, mentioned by twothirds of the respondents. Slightly more than 16% of those surveyed said
that failure to find a suitable replacement could impede their retirement plans.
Not surprisingly, more rural physicians than urban doctors mentioned
this reason.33
It thus appears that while many physicians are expected to exit the Canadian
medical workforce in the coming years based solely on age, it is unlikely that
most physicians will retire at age 65 or before, given past and current trends.
In fact, the studies reviewed suggest that many physicians will continue to work
beyond the age of 65, though they may take on a less demanding workload.
2.3 Data Issues
Two types of data are typically used to document or estimate the extent of
physician retirement: data on retirement intentions based on surveys and
secondary data from administrative, registration or FFS billing databases. The
former includes such sources as the CMA Physician Resource Questionnaire
and the NPS. More will be said about the nature of each of these databases
in the next chapter.
Some surveys ask physicians when they plan to retire. For instance,
Maguiness and colleagues conducted a Canadian dermatology workforce
survey, in which they asked dermatologists about their retirement plans.34
The survey results showed that the average dermatologist planned to retire at
age 64, and 13% of the survey respondents planned to retire within the next
five years. Similarly, Macadam and associates conducted a Canadian plastic
surgery workforce survey in 2004–2005.35 On average, the respondents
planned to retire at age 63. Twenty-eight percent of the surveyed plastic
11
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
surgeons indicated that they planned to retire within five years. In 2000, the
CMA Physician Resource Questionnaire asked physicians to indicate the age
at which they planned to retire. The average age of planned retirement was 63.
Interestingly, the average age of planned retirement increased progressively
by age group. Those younger than 35 said they planned to retire, on average,
at age 58; the average ages of planned retirement were 63 for those age 45
to 54 and 66 for those age 55 to 64. Those age 65 and older planned to retire,
on average, at age 72.33 It appears that the older the physician, the more likely
he or she is to plan a later retirement. Another survey that contains physician
retirement information is the NPS, which asks physicians, among other things,
if they plan to retire in the next two years (data on retirement plans from the
2007 survey is presented and discussed in the next chapter).
But to what extent should credence be given to anticipated age of retirement
obtained from surveys? Although no Canadian studies can be found that
compare physician retirement intentions with actual retirement behaviours,
studies conducted in the United States suggest that caution is needed when
interpreting survey data on retirement plans. For instance, Bahrami and
associates have warned that expected retirement age is not a perfect proxy
for actual retirement age, though it is widely used by researchers.17 Likewise,
having compared data from the 1997 Physician Worklife Survey with data
from follow-up surveys and the 2003 American Medical Association Physician
Master File, Konrad and Dall have come to the conclusion that physicians’
retirement intentions are not accurate predictions of their future retirement
behaviours.36 This does not imply that physicians are not truthful when
answering questionnaires. Discrepancies between intentions and actions may
be due to a variety of factors, such as changes in personal circumstances or
the external environment. v
Retirement estimates can also be made using secondary data from such
sources as Scott’s Medical Database (SMDB), the CMA Master File and
registration databases of various provincial/territorial colleges of physicians
and surgeons. But such estimates also come with their own methodological
challenges. First, none of the data sources in Canada are specifically
designed to capture information on retirement; therefore, retirement estimates
typically have to be inferred or derived. Second, different data sources may
define retirement differently, thus making comparisons difficult. Finally,
because the retirement rate is obtained by dividing the number of individuals
retiring by the number of individuals in the base population, the rate is a
v.
12
For instance, the economic downturn in 2008–2009 appears to have had an impact on Canadian
physicians’ retirement plans, as an article in The Medical Post suggested: “The Medical Post has
heard from many doctors, particularly those nearing retirement, who say as a result of the financial
meltdown they have seriously re-calibrated their retirement plans. The majority say they are taking
on more work, putting in longer hours at the office or even effectively starting second careers
in medicine.”37
Chapter 2: Physician Retirement: Known, Unknown and Questions
function of how retirees are counted and who is included in or excluded from
the base population. Typically, different databases have different inclusion
or exclusion criteria, again making comparisons of retirement rates based
on different data sources problematic.
Other sources of data that are not exclusively about physicians may also
be used to study some aspects of physician retirement. For example, the
Canadian censuses conducted by Statistics Canada have some information
about occupations and labour force activities of Canadians, including
physicians. The aforementioned study by Duchesne is an example of census
data–based research, which sheds some light on physician retirement.13
However, the usefulness of such sources of data for studying physician
retirement tends to be limited.
All this suggests that difficulties encountered in studying physician retirement
in Canada stem not just from lack of agreement about how retirement is
understood or defined but also from the quality of data available for analysis.
Generally speaking, existing sources of data are less than adequate in
analyzing the extent of physician retirement. More will be said about this
in the following chapters.
13
Chapter 3
Data Sources
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Data from several sources was used to paint a complex picture of the changes
that take place when physicians grow older. Four databases were used: the
2007 National Physician Survey (NPS), Scott’s Medical Database (SMDB),
the CMA Master File and the National Physician Database (NPDB). A brief
description of each appears below.
3.1 2007 National Physician Survey
The most recent triennial NPS with data available for analysis was conducted
in 2007. The 2007 NPS was jointly conducted by the College of Family
Physicians of Canada, the CMA and the Royal College of Physicians and
Surgeons of Canada. It surveyed all practising physicians, second-year
residents and medical students in Canada about what they were doing or
intended to do in their practice. Only data from practising physicians was
used in this analysis. The 2007 NPS used multiple questionnaires: a core
questionnaire and two versions of the detailed questionnaire—one for
FPs/GPs and one for specialists. All of the content captured on the core
questionnaire was also captured on the detailed questionnaires.
Physicians were assigned to strata based on their province/territory, broad
specialty (family medicine/general practice or other specialties) and sex. For
Alberta, British Columbia, Ontario and Quebec, one in three physicians in
each stratum received the detailed questionnaire, while all other physicians
received the core (shorter) questionnaire. For the other provinces and the
territories, two out of three physicians in each stratum received the detailed
questionnaire, while all others received the core questionnaire. Respondents
had the option of completing the survey electronically or on paper. Of the
60,811 eligible physicians surveyed, 19,239 completed the survey, for an
overall response rate of 31.6%.
When a sample is selected for a survey with unequal probabilities (as was the
case for the 2007 NPS detailed questionnaire sample), weights are typically
used when making estimates so the weighted sample is representative of
the population. Censuses (a census was attempted for the 2007 NPS core
questions) are subject to non-response, and weights are used in estimation
to reduce possible non-response biases. In the case of the 2007 NPS,
Canada-level estimates for the detailed questions for the entire population of
physicians are within 3.2 percentage points, 19 times out of 20. vi
vi.
16
For a more detailed discussion of the survey and weighting methodologies, visit the NPS website at
www.nationalphysiciansurvey.ca/ nps/2007_Survey/pdf/2007.NPS.Methodology.and.Generalizability.
of.results.FINAL.pdf.
Chapter 3: Data Sources
Some of the 2007 NPS data was used to identify the proportion of physicians
who said they planned to retire within the next two years and the proportion of
FPs/GPs who said they had reduced their scope of practice in the two years
before the survey and who said they planned to reduce their scope of practice
in the next two years (see Appendix A for questions used in the 2007 NPS).
3.2 Scott’s Medical Database
The SMDB (formerly known as the Southam Medical Database) provides
information on the demographic and other characteristics of Canadian
physicians. Researchers and health care planners can use SMDB data to
examine the supply, distribution and migration patterns of physicians and
historical changes in the Canadian medical workforce.
Scott’s Directories maintains a database on physicians to produce the
Canadian Medical Directory and mailing lists for commercial purposes. Each
year, the Canadian Institute for Health Information (CIHI) acquires a copy
of this database for the purpose of maintaining and updating the SMDB.
The database contains information on physicians’ name, sex, year of birth,
province/territory, activity status, specialty, hospital affiliation status and
so forth. All data collection and updates are done by Scott’s Directories.
Information and updates are collected from organizations such as physician
licensing authorities, the 17 medical schools in Canada, the Royal College
of Physicians and Surgeons of Canada, the College of Family Physicians of
Canada, the Collège des médecins du Québec and hospitals. In addition,
an annual questionnaire is sent by Scott’s Directories to all active physicians
and all new medical school graduates to confirm or update the information on
record. Physicians may also contact Scott’s Directories throughout the year to
provide new information. At CIHI, once the file has been received from Scott’s
Directories, it is processed through a series of edit checks on the database.
Scott’s Directories is interested in collecting information on all physicians in
Canada, regardless of their type of practice, as long as sufficient information
is available. The SMDB defines physicians as active if they have a medical
doctorate (MD) degree and a valid mailing address, but semi-retired and
retired are not formally defined. These variables are subjective in nature as
they are self-reported by physicians.
3.3 National Physician Database
The NPDB provides information on the demographic profiles of physicians
and their levels of activity within the Canadian medicare system. The NPDB is
managed by CIHI, but data on demographic characteristics and activity levels
of physicians is provided by provincial and territorial health care insurance
17
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
plans. Information such as demographic characteristics and activity levels
of physicians is used by CIHI to generate derived variables like total FFS
payments, total services, average FFS payments per physician and an FTE
physician measurement. These derived variables are currently based solely on
FFS information.
FTE values are calculated for all physicians contained in the NPDB and
are used as a measure of relative workload. FTE values are calculated by
comparing the total annual FFS payments of individual physicians with defined
FTE payment benchmarks. Unique upper and lower national FTE benchmarks
are defined for each provincial medical specialty group. FTE values, therefore,
take into consideration variations in average FFS payments across specialty
groups and across provincial and territorial health care insurance plan fee
schedule prices. Physicians’ individual FTE values are calculated using the
following formula:
{
FTE =
B / B 40
If the physician’s total FFS payments (B) are below the total FFS
payment value at the 40th percentile (B40) for the physician’s
province-specific specialty group, then <1 FTE.
1
If the physician’s total FFS payments (B) are between the total
FFS payment values at the 40th and 60th percentiles for the
physician’s province-specific specialty group, then = 1 FTE.
If the physician’s total FFS payments (B) are above the total
1 + log (B / B60) FFS payment value at the 60th percentile (B60) for the
physician’s province-specific specialty group, then >1 FTE.
3.4 Canadian Medical Association
Master File
The CMA collects demographic information from individual physicians, both
CMA members and non-members. This is supplemented with information
provided by provincial/territorial associations that, in turn, receive data from
the jurisdictional licensing bodies. The CMA also receives information from
the Canadian certifying bodies for family medicine and other specialists.
The CMA Master File is an anonymized annual extract of this file that is used
for research and planning purposes, such as providing data for physician
workforce forecasting.
18
Chapter 4
Extent of Physician Retirement
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
The extent of physician workforce attrition due to retirement and death is
very much related to the age structure of the physician population. A growing
proportion of older physicians typically means greater attrition from the
medical workforce through retirement and death. The changing age structure
of the population of active physicians in Canada based on data from the
SMDB is shown in figures 1 to 3. Over a period of two decades, the proportion
of physicians age 55 and older increased from 25.4% in 1987 to 33.2% in
2007. The proportion of physicians age 65 and older was 7.8% in 1987 and
10.8% in 2007; conversely, the proportion of those younger than 35 dropped
from 21.9% in 1987 to 9.8% in 2007. Medical and surgical specialists, as a
group, were slightly older than those in family or general practice (specialties
grouped under the broad categories of Family Physicians and General
Practitioners, Medical Specialists and Surgical Specialists can be found in
Appendix B).
Figure 1: Percentage Distribution of Physicians, by Broad Specialty
and Age Group, Canada, 1987
35
30
Percentage
25
20
15
10
5
0
<35
35–44
45–54
55–64
65–74
75+
Age Group
FPs/GPs
Medical Specialists
Surgical Specialists
Source
Scott’s Medical Database, Canadian Institute for Health Information.
20
All Physicians
Chapter 4: Extent of Physician Retirement
Figure 2: Percentage Distribution of Physicians, by Broad Specialty and Age
Group, Canada, 1997
35
30
Percentage
25
20
15
10
5
0
<35
35–44
45–54
55–64
65–74
75+
Age Group
FPs/GPs
Medical Specialists
Surgical Specialists
All Physicians
Source
Scott’s Medical Database, Canadian Institute for Health Information.
Figure 3: Percentage Distribution of Physicians, by Broad Specialty and Age
Group, Canada, 2007
35
30
Percentage
25
20
15
10
5
0
<35
35–44
45–54
55–64
65–74
75+
Age Group
FPs/GPs
Medical Specialists
Surgical Specialists
All Physicians
Source
Scott’s Medical Database, Canadian Institute for Health Information.
21
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
The following presents findings about retirement intentions and behaviours
using different sources of data. Information about physicians’ intention to
retire was obtained from the 2007 NPS. Estimates of actual retirement were
derived from the CMA Master File and the SMDB, assuming that incidents
of retirement are accurately reported and recorded.
Figure 4: Percentage of Family Physicians/General Practitioners and Specialists
Planning to Retire From Clinical Practice in the Next Two Years, by Age
Group, Canada, 2007
40
35
30
Percentage
25
20
15
10
5
0
<35
35–44
45–54
55–64
65–74
75+
Total
Age Group
FPs/GPs
Specialists
Source
CFPC/CMA/RCPSC National Physician Survey Database, 2007. “Protected by Copyright.”
Figure 4 (more detailed data can be found in Table C1 in Appendix C) shows
that 6.2% of FPs/GPs and 6.5% of all specialists indicated in the 2007 NPS
that they planned to retire in the two years following the survey. If intentions
translated into action, on average, about 3.2% of all physicians would have
retired from clinical practice in each of the two years following the survey.
Generally speaking, the difference between male and female physicians with
respect to intention to retire was not substantial. FPs/GPs age 55 and older
were somewhat more likely than older specialists to express an intention
to retire within the next two years.
The data above is self-reported, prospective information indicating retirement
intentions in the near future. What about retirement behaviours? Actual
retirement rates can be estimated from the SMDB, which contains information
22
Chapter 4: Extent of Physician Retirement
about semi-retirement and retirement, and the CMA Master File, which provides
information about retirement. Three years of data were used in each of the
following analyses to even out possible random variations from year to year.
Figures 5 to 7 provide the estimated actual retirement rates derived from
SMDB data (detailed data can be found in Table C2 in Appendix C). Three
years of data—2007, 2006 and 2005—are presented. As shown in the second
part of Table C2 in Appendix C, of those physicians who were active in 2005,
114 (or 0.19%) became semi-retired, 333 (0.56%) retired and 96 (0.16%)
passed away in 2006. Similarly, the first part of the table shows that of those
physicians who were active in 2006, 65 (0.11%) changed their activity status
to semi-retired, 177 (0.29%) retired and 84 (0.14%) died in 2007. If the semiretired and retired physicians are combined, vii 242 (or 0.40%), 447 (or 0.75%)
and 286 (or 0.48%) physicians captured by the SMDB exited the medical
workforce through retirement in 2007, 2006 and 2005, respectively. These
actual overall retirement rates, derived from SMDB data, are substantially
lower than the overall rates of intended retirement (an average of 3.2% of all
physicians in a year) obtained from the 2007 NPS.
Percentage
Figure 5: Percentage of Semi-Retired and Retired Physicians Who Were Active
in 2004, by Age Group and Sex, Canada, 2005
Semi-Retired in 2005
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
<35
Retired in 2005
35–44 45–54 55–64 65–74 75+ Total
<35 35–44 45–54 55–64 65–74 75+ Total
Age Group
Male
Female
Source
Scott’s Medical Database, Canadian Institute for Health Information.
vii. Since retired and semi-retired are self-defined and self-reported by physicians in the SMDB, there
is no clear-cut demarcation between the two categories. To be more conservative in estimating the
extent of physician retirement, it was decided to consider semi-retired as retired.
23
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Figure 6: Percentage of Semi-Retired and Retired Physicians Who Were Active
in 2005, by Age Group and Sex, Canada, 2006
Semi-Retired in 2006
Retired in 2006
7.0
Percentage
6.0
5.0
4.0
3.0
2.0
1.0
0.0
<35
35–44 45–54 55–64 65–74 75+ Total
<35 35–44 45–54 55–64 65–74 75+ Total
Age Group
Male
Female
Source
Scott’s Medical Database, Canadian Institute for Health Information.
Figure 7: Percentage of Semi-Retired and Retired Physicians Who Were Active in
2006, by Age Group and Sex, Canada, 2007
Semi-Retired in 2007
Retired in 2007
3.0
Percentage
2.5
2.0
1.5
1.0
0.5
0.0
<35
35–44 45–54 55–64 65–74 75+ Total
<35 35–44 45–54 55–64 65–74 75+ Total
Age Group
Male
Female
Source
Scott’s Medical Database, Canadian Institute for Health Information.
24
Chapter 4: Extent of Physician Retirement
Actual retirement rates can also be derived from CMA Master File data.
Figure 8 shows the estimated retirement rates for each of three years: 2005,
2006 and 2007 (more detailed data can be found in Table C3 in Appendix C).
The base population of physicians is the total count of physicians on the CMA
Master File at the beginning of each calendar year. The number of retired and
deceased physicians is as of the end of that calendar year (that is, the number
of physicians who retired or died between January 1 and December 31 of the
year). While the overall death rates were quite consistent over the three years,
the overall retirement rates fluctuated considerably from year to year. Similar
to the results based on SMDB data, the actual overall annual retirement rates
derived from CMA Master File data are considerably lower than the estimated
overall rates of intended retirement (an average of 3.2% of all physicians
in a one-year period) obtained from the 2007 NPS. viii
Figure 8: Percentage of Physicians Retiring in 2005, 2006 and 2007,
by Age Group and Sex, Canada
14.0
2005
2006
2007
l
35 44 54 64 74 + ta
< 35– 45– 55– 65– 75 To
l
35 44 54 64 74 + ta
< 35– 45– 55– 65– 75 To
12.0
Percentage
10.0
8.0
6.0
4.0
2.0
0.0
l
35 44 54 64 74 + ta
< 35– 45– 55– 65– 75 To
Age Group
Male
Female
Source
Canadian Medical Association Master File, Canadian Medical Association.
viii. It should be noted that whereas the retirement rates derived from the SMDB and CMA Master File
data are for 2005, 2006 and 2007, the estimated retirement rates derived from the 2007 NPS are for
2008 and 2009, as the survey asked respondents to indicate whether they planned to retire in the two
years following the survey.
25
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
The SMDB and the CMA Master File yielded different estimated retirement
rates. The estimated overall retirement rates based on the SMDB were 0.40%
in 2007, 0.75% in 2006 and 0.48% in 2005, with an average overall retirement
rate of 0.54% per year. The estimated overall retirement rates based on the
CMA Master File were 0.77%, 1.06% and 0.53% in 2007, 2006 and 2005,
respectively, with an average overall retirement rate of 0.79% per year, which
was somewhat higher than that based on the SMDB. But both estimates
pale in comparison to the overall rate of intended retirement over a one-year
period—about 3.2%—based on responses to the 2007 NPS. An inevitable
question arises from these disparate findings: Why such differences?
The answer to this question may be due to differences between the databases.
As noted earlier, data sources understand or define retirement differently.
Retirement information in the CMA Master File is based on information
provided by physician licensing bodies across the country, either directly to the
CMA or indirectly through one of the provincial/territorial medical associations.
In addition, the CMA uses responses to the screening questions on its surveys
to alter the status of a physician on the CMA Master File to “retired” if he or she
has indicated retirement or has reported no activities. With respect to the base
physician population, the CMA Master File includes every person with an MD
degree who is licensed and has a valid Canadian address but who is not a
medical student or resident and is not age 80 or older. Also excluded are nonCMA members who are older than 70 and who do not have a business address
in the public directory of their province of residence. The CMA information may
include someone who has left medicine for a sufficiently long period of time
to let their licence lapse and female physicians who are taking an extended
maternity leave. As well, without information on semi-retirement, it is difficult to
determine how active many of the licensed physicians are.38
The SMDB, on the other hand, has several medical activity codes, such as
active, retired and semi-retired. A physician is deemed active if he or she is
registered with a provincial college of physicians and surgeons, has provided
the SMDB with a mailing address and is listed in the directory. However, there
are no formal definitions for retired or semi-retired. These classifications are
what physicians have indicated in the annual questionnaire sent to them
by Scott’s Directories.39 What separates active from semi-retired and what
distinguishes between semi-retired and retired may be entirely subjective, and
there are no clear demarcation lines between the activity categories. A more
detailed analysis comparing the CMA Master File and the SMDB is required to
fully understand why retirement figures differ between these two databases. ix
ix.
26
It can be observed from tables C2 and C3 in Appendix C that the SMDB and the CMA Master File
also differ considerably in the numbers and percentages of deceased (20% or more)—a status that
is more clear-cut and less subject to individual interpretations—suggesting that the two databases
are quite different in the way they capture information.
Chapter 4: Extent of Physician Retirement
The difference between the SMDB and the CMA Master File on the one
hand and the 2007 NPS on the other is much more substantial with respect
to estimated retirement rates. The most important difference is the fact that
whereas the former are based on the actual number of physicians retiring
(regardless of what retiring means), the latter reflects prospective behaviours
based on self-reported intentions. As pointed out earlier, there are concerns
about the extent to which stated intentions translate into actual behaviours.
Another possible difference is that, whereas the 2007 NPS refers to “plan
to retire from clinical practice,” the SMDB and the CMA Master File refer to
“retired.” It is possible that some of those who planned to retire from clinical
practice would continue to be active in other types of medical work, but
those who were retired according to the SMDB and the CMA Master File were
no longer involved in any medically related activity. Further investigation is
needed to ascertain the extent to which this is true.
One could ask another question: If a physician works a few hours or sees a
few patients a week, should he or she be considered active, retired or semiretired? Since there is not a uniform definition of retirement that is universally
accepted, some physicians who are minimally active may not consider
themselves retired. This may inflate the number of active physicians in some
medical workforce databases. This factor is examined in the next chapter.
27
Chapter 5
Older Physicians Who Were Minimally Active
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
This chapter is about older physicians who continue to do clinical work but on
a very limited scale. The purpose of this analysis is to find out how many more
physicians could be considered retired if those who work below a certain
activity threshold were no longer deemed active from a medical workforce
planning perspective. It has been suggested that retirement rates based
on data from such sources as the SMDB and the CMA Master File are low
because they include many older physicians who keep their medical licences
and continue to work at a very low activity level. The argument is that if those
minimally active older physicians were counted as retired, a more accurate
picture of physician retirement would emerge.
There are many reasons why some physicians limit their clinical practice.
For instance, some young doctors may work fewer hours to raise a family.
Others may focus on administration or may be pursuing further training. Still
others may minimize their clinical work due to poor health. If older physicians
begin to work fewer hours, it is quite possible that they are transitioning to full
retirement. But should older physicians who have a very limited practice be
counted as active physicians? This is not just an academic debate about when
active medical practice ends and when retirement begins. A large number
of minimally active older physicians may bias medical workforce statistics
(by over-counting the number of active physicians) and, as a result, skew
workforce projections.
To gain a better understanding of this problem, data from the NPDB was
used to estimate the number of older physicians who had decreased their
workload and consequently had an FTE that was 33% or less of what it used
to be. Physicians’ active FTE values were calculated by taking the average
of all their FTE values for the consecutive years they were above the “active”
threshold (see Appendix D for more details). For the purpose of this analysis,
older physicians are those who were 55 and older. Three age groups were
used in the analysis: 55 and older, 60 and older and 65 and older. In addition,
three FTE thresholds were specified: those whose FTE value for at least
three consecutive years was 15% or less, 25% or less or 33% or less of the
average FTE values of their average active FTE values. The specification
of three consecutive years was used to avoid mislabelling as retired those
physicians who took some time off for family, health or other reasons but
returned to regular work subsequently. The analysis was done separately for
male and female physicians and separately for FPs/GPs, medical specialists
and surgical specialists, using 2007 and 2006 as the reference years. Two
years—2007 and 2006—of data were analyzed to show that the results are
generally consistent across years.
For physicians in New Brunswick, Newfoundland and Labrador and Prince
Edward Island, both FFS and alternative payments were included in the FTE
calculations. Physician-level alternative payment data was combined with
FFS payment data to generate a new FTE value, using FFS benchmarks as
30
Chapter 5: Older Physicians Who Were Minimally Active
an estimate. For physicians in the other provinces, only FFS payments were
included in the FTE calculations. Physicians in the three territories were not
included in this analysis. Appendix D provides further methodological details.
The numbers of older FPs/GPs, medical specialists and surgical specialists
who could be considered retired in the 2007 reference year are presented
in Table 1.
Table 1: Number of Older Physicians Who Were Minimally Active, by Specialty
and Full-Time Equivalent Threshold, Canada, 2007
Specialty
FTE
Threshold
Age 55+*
Age 60+*
Age 65+*
FPs/GPs
15% or Less of FTE
25% or Less of FTE
33% or Less of FTE
225
364
500
131
227
334
87
153
217
Medical
Specialists
15% or Less of FTE
25% or Less of FTE
33% or Less of FTE
125
196
284
76
129
200
50
90
140
Surgical
Specialists
15% or Less of FTE
25% or Less of FTE
33% or Less of FTE
133
284
275
110
175
237
96
156
211
14,460
9,044
4,792
Total Number
of Physicians
in Age Group
Note
* Age 55+ refers to all physicians who were 55 and older; age 60+ refers to all
physicians who were 60 and older; and age 65+ refers to all physicians who were
65 and older.
Source
National Physician Database, Canadian Institute for Health Information.
Table 1 shows that if “older physicians” refers to those age 55 and older and
if “minimally active” means physicians’ clinical activities over a minimum of
three years, including the 2007 reference year, were at 33% or less of their
previous average active FTE value, then 1,059 physicians (500 FPs/GPs, 284
medical specialists and 275 surgical specialists) could be considered retired,
even though they were still included in the NPDB. This represents 7.3% of all
physicians age 55 and older in the NPDB in 2007. On the other hand, if “older
physicians” is understood to mean those age 65 and older and if “minimally
active” refers to physicians’ clinical activities that were at 15% or less of
their previous average active FTE values, then only 233 physicians (87 FPs/
GPs, 50 medical specialists and 96 surgical specialists) could be considered
retired. This represents 4.9% of all physicians age 65 and older in the NPDB
in 2007. The numbers of minimally active physicians who could be considered
retired vary between these two extremes, depending on how the term “older
physicians” is interpreted and where the FTE threshold is set.
31
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Table 2: Number of Older Physicians Who Were Minimally Active, by Specialty
and Full-Time Equivalent Threshold, Canada, 2006
Specialty
FTE
Threshold
Age 55+*
Age 60+*
Age 65+*
FPs/GPs
15% or Less of FTE
25% or Less of FTE
33% or Less of FTE
169
287
399
99
175
267
69
126
187
Medical
Specialists
15% or Less of FTE
25% or Less of FTE
33% or Less of FTE
96
171
252
55
109
172
36
76
120
Surgical
Specialists
15% or Less of FTE
25% or Less of FTE
33% or Less of FTE
111
178
236
93
150
203
82
134
181
13,673
8,357
4,362
Total Number
of Physicians
in Age Group
Note
* Age 55+ refers to all physicians who were 55 and older; age 60+ refers to all
physicians who were 60 and older; and age 65+ refers to all physicians who were
65 and older.
Source
National Physician Database, Canadian Institute for Health Information.
Table 2 presents similar data for the 2006 reference year. It shows that if “older
physicians” refers to those age 55 and older and if “minimally active” means
physicians’ clinical activities over a period of at least three years, including
the 2006 reference year, were at 33% or less of their previous average FTE
values, then 887 physicians (399 FPs/GPs, 252 medical specialists and 236
surgical specialists) could be considered retired. This represents 6.5% of all
physicians age 55 and older in the NPDB in 2006. But if “older physicians”
is understood to be those age 65 and older and if “minimally active” means
physicians’ clinical activities were at 15% or less of their previous average FTE
values, then only 187 physicians (69 FPs/GPs, 36 medical specialists and 82
surgical specialists) could be considered retired. This represents 4.3% of all
physicians age 65 and older in the NPDB in 2006. The numbers of minimally
active physicians who could be considered retired vary between these two
extremes, depending on how the term “older physicians” is interpreted and
where the FTE cut-off is set.
Whether these minimally active physicians should be considered retired for
the purpose of medical workforce projections or health human resources
planning is a decision that requires further consideration and deliberation
by stakeholder groups. However, it should be noted that these figures are
merely estimates because the FTE conversions were based only on FFS
billing information, with the exception of physicians in three Atlantic provinces.
More and more physicians, especially FPs/GPs and medical specialists, are
shifting to alternative payment programs or blended remuneration schemes.
32
Chapter 5: Older Physicians Who Were Minimally Active
There could be a sizeable number of physicians who are getting a salary,
sessional or capitation payment, plus some FFS income as top-up, and it is
only the FFS top-up that has been captured by the NPDB and converted into
FTEs. In other words, the number of minimally active physicians is likely to
be smaller than what was estimated in the above analyses. This scenario is
more likely to be true for the younger aging physicians (those younger than
65). Thus, when estimating the number of minimally active physicians who
could be considered retired, it is safer to examine physicians who are 65 and
older, as they are less likely to receive both FFS payments and alternative
forms of reimbursement.38, x If this more cautious strategy is followed, the
numbers of minimally active physicians who could be considered retired are
not substantial.
Differences between male and female physicians and differences between
the three broad specialty groups were then examined with respect to their
likelihood of being minimally active. Given that the two criteria, with three
categories for each, generate nine possible scenarios, for reasons of
parsimony and for illustration purposes, only one scenario—using the middle
category of each of the two criteria—is presented as follows. When the age
group of 60 and older and 25% of previous average FTE as activity threshold
were used for the analysis, it was found that, in the 2007 reference year, there
was no major difference between males and females (male physicians: 6.0%;
female physicians: 5.0%). Differences between the three broad specialties,
particularly between FPs/GPs and medical specialists on the one hand and
surgical specialists on the other, were somewhat more pronounced: 5.2%,
4.3% and 10.7% of FPs/GPs, medical specialists and surgical specialists,
respectively, could be considered retired because their clinical activities
during a minimum of three consecutive years, including the 2007 reference
year, were at 25% or less of their previous average FTE values. Data for the
2006 reference year reveals a similar pattern. The difference between male
and female physicians who were minimally active was not substantial (male
physicians: 5.3%; female physicians: 4.4%). There were greater differences
between the three categories of physicians: 4.4% of FPs/GPs, 3.9% of medical
specialists and 9.5% of surgical specialists could be considered retired
because their clinical activities during a minimum of three consecutive years,
including the 2006 reference year, were at 25% or less of their previous
average FTE values.
x.
There is some recent data to back up this suggestion. Results from a CIHI study on the demographic
profile of physicians on different payment programs in three Atlantic provinces show that the
proportion of physicians receiving 90% or more of their income from alternative payment programs
declined with age, from 57.6% among those younger than 40 to 41.1% among those age 40 to 59 and
to 28.3% among those age 60 or older.40
33
Chapter 6
An Alternative Perspective
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
After examining retirement patterns of Canadian workers, Bowlby has
resignedly come to the conclusion that “the concept of retirement is fuzzy.”9
The fuzziness problem may be particularly true for physicians, most of whom
are independent, self-employed practitioners, and many of whom continue to
work beyond the traditional retirement age of 65, as other studies have found
and as data presented in the previous chapters has shown. The problem is
made more intractable by the fact that existing physician databases do not
provide adequate information to study physician retirement because they
define retirement loosely or inconsistently and because they lack uniformity
in gathering retirement-related information. As well, the base populations
in different databases may not be the same due to different inclusion or
exclusion criteria. Not surprisingly, estimated retirement rates derived from
different databases vary considerably, and physician supply forecasts using
these retirement rates as projection parameters may produce quite divergent
results years into the future. This leads one to wonder how useful the concept
of retirement is, as far as medical workforce planning is concerned. Maybe a
different perspective on physician aging and medical practice is needed.
It is suggested that retirement should be seen as just one aspect in a
continuum of changes in medical practice as old age sets in. Retirement—in
the sense of complete cessation of medical practice—represents the end
point of this continuum, which may or may not occur at age 65. But prior to full
retirement, many other changes could have occurred, which may also have
implications for medical care provision and physician workforce planning. It
may be more meaningful, for example, to measure how much work physicians
do when they become older. It is because of the recognition of the importance
of activity levels that some health workforce researchers and planners have
shifted to using FTEs, as well as head counts, in their analysis, since FTEs
take into account different levels of clinical activity, as reflected by FFS billing
volume. xi Changes in FTE value in different stages of a physician’s medical
career reflect changing levels of clinical activity as he or she gets older.
6.1 Changing Activity Level
The NPDB makes it possible to use two measures to gauge a physician’s
clinical activity level: number of clinical services provided and FTEs. The two
are related because the former measures the volume of work and the latter
determines the monetary value of that work as reflected by FFS payments.
xi.
36
It should be noted that the methodology for calculating FTEs is still being refined. Additionally, the
current FTE methodology includes only those physicians who are on FFS reimbursement (about 78%
of physicians); the proportion of physicians, especially FPs/GPs and medical specialists, who are not
on FFS is growing.
Chapter 6: An Alternative Perspective
Figure 9: Average Number of Services Provided by Physicians Who Billed Fee
for Service, by Sex, Specialty and Age Group, Canada, 2006
8,000
Number of Services
7,000
6,000
5,000
4,000
3,000
2,000
1,000
<40
40–44
45–49
50–54
55–59
60–64
65–69
70–74
75–79
80+
Age Group
Family Medicine—Male
Medical Specialty—Male
Surgical Specialty—Male
Family Medicine—Female
Medical Specialty—Female
Surgical Specialty—Female
Source
National Physician Database, Canadian Institute for Health Information.
Figure 10: Average Full-Time Equivalent Values for Physicians Who Billed Fee
for Service, by Sex, Specialty and Age Group, Canada, 2006
1.20
1.00
FTE
0.80
0.60
0.40
0.20
0.00
<40
40–44
45–49
50–54
55–59
60–64
65–69
70–74
75–79
80+
Age Group
Family Medicine—Male
Medical Specialty—Male
Surgical Specialty—Male
Family Medicine—Female
Medical Specialty—Female
Surgical Specialty—Female
Source
National Physician Database, Canadian Institute for Health Information.
Figure 10:
37
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Figure 9 shows the average numbers of FFS services provided by physicians
by age, sex and broad specialty in 2006, and Figure 10 shows the average
FTE values of physicians who billed FFS by age, sex and broad specialty
in the same year (more detailed data can be found in tables E1 and E2 in
Appendix E).
It can be seen from Figure 10 that the average FTE values of physicians
who billed FFS varied by age, sex and specialty. Male physicians tended to
have higher FTEs than female physicians. In only a few instances did female
physicians surpass their male counterparts in FTE value (such as female
surgical specialists age 60 to 69). Generally speaking, both medical and
surgical specialists generated higher FTE values than physicians in family or
general practice. A comparison of age groups revealed that physicians who
were younger than 40 tended to have lower average FTE values. The age
at which a physician achieves his or her highest FTE value is a function of
both sex and specialty. For instance, male FPs/GPs generated their highest
average FTE values between ages 50 and 59. The average FTE values then
tended to decline progressively as physicians got older. But for those FPs/GPs
age 75 to 79 who stayed clinically active, their average FTE values were still at
0.52 for both males and females. Medical and surgical specialists followed a
similar pattern, but they achieved their peak FTE values at a younger age; this
was especially true for surgical specialists. Male surgical specialists had very
high average FTE values between ages 40 and 54 (for example, the average
FTE value of those age 40 to 44 reached 1.03), but they tended to have much
lower average FTE values after age 64 relative to medical specialists and FPs/
GPs. In 2004, 52,813 physicians (excluding unknowns) billed FFS and were
included in the calculation of FTE values. Of the total number of physicians
who billed FFS, 5,461 were age 65 and older.
6.1.1 First Case Study
One way to illustrate the impact of aging on the production of medical services
is to examine how a cohort of physicians who billed FFS changed over time
with respect to the number of physicians remaining in clinical practice and
the amount of FFS-billable work they performed. This case study involves
six age–specialty cohorts: male internists age 50 to 54 in 1989 (n = 483),
male internists age 55 to 59 in 1989 (n = 391), male general surgeons age
50 to 54 in 1989 (n = 267), male general surgeons age 55 to 59 in 1989 (n = 276),
male obstetricians/gynecologists age 50 to 54 in 1989 (n = 182) and male
obstetricians/gynecologists age 55 to 59 in 1989 (n = 159). These three
specialties were chosen because they were among the largest specialties.
Female physicians were not included in this analysis because there were very
few older female specialists. For example, there were only 17 female internists
age 55 to 59 in 1989 who billed FFS, and the number dwindled to fewer than
38
Chapter 6: An Alternative Perspective
5 by 2005. Similarly, there were fewer than 5 female general surgeons age 50
to 54 in 1989 who billed FFS. Small numbers present the potential problem of
privacy infringement and the possibility of distortion by a few outliers.
The analysis entails tracking these six cohorts of physicians xii over an 18-year
period, from 1989 to 2006 (for presentation reasons, data for only 10 years
is shown in Table E3 in Appendix E), to see how many individuals remained
in FFS clinical practice until the end of the study period and how the average
FTE value of the still-active cohort members changed over time. General
surgeons who billed FFS for more than zero dollars and who were age 50 to
54 in 1989 were included in the General Surgery, 50 to 54 cohort; internists
who billed FFS for more than zero dollars and were age 55 to 59 in 1989
were included in the Internal Medicine, 55 to 59 cohort; and so forth. The
tracking involved identifying the number of physicians in a particular cohort
who billed for more than zero dollars in 1989 and in each of the subsequent
years up to 2006. It should be pointed out that clinically active, in the present
context, means billing FFS for clinical work. Those cohort members who had
given up clinical work after 1989 but were involved in other activities, such as
administration, research or teaching, were no longer tracked, as the NPDB
contains no information on such activities. As well, some could still have been
clinically active but have moved to alternative payment programs (non-FFS)
and were, therefore, no longer captured by the NPDB.
The findings are presented in figures 11 and 12. The Internal Medicine, 50 to
54 cohort provides an illustration. There were 483 internists in this cohort in
1989. By 1997, when this cohort of physicians reached the age range of 58 to
62, 423 (or 87.6%) were still in FFS clinical practice. By 2006, when the cohort
reached age 67 to 71, 276 members of the original group (or 57.1%) were still
billing FFS for the clinical services they provided (see Figure 11). The average
FTE value changed from year to year (see Figure 12). Already at a high level
of 0.94 in 1989, the average FTE value climbed slowly to 1.00 in 1997. After
that year, the FTE value dropped fairly rapidly to 0.73 in 2006. In other words,
in 2006, slightly more than half of the original cohort of 483 FFS internists,
who were already beyond the traditional retirement age of 65, were still
clinically active, but the average workload (as reflected by FTE values) of
those still billing FFS was slightly more than 25% lower than that in 1997.
A few of those who had dropped out might have died; others might have
left Canada, retired or stopped practising as a clinician but were otherwise
medically active. Some might still have been clinically active but were
reimbursed in a way other than FFS.
xii. Individual physicians within a cohort were tracked over time. In other words, all physicians who were
tracked in years after 1989 were part of the 1989 cohort. Physicians age 50 to 54 or 55 to 59 who
started practising or re-entered the Canadian medical workforce after 1989 (such as newly licensed
international medical graduates or Canadian physicians returning from overseas) were excluded
from this analysis.
39
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Figure 11: Number of Active Physicians Who Billed Fee for Service,
Six Age–Specialty Cohorts, Canada, 1989 to 2006
Number of Physicians
600
500
400
300
200
100
0
1989
1991
1993
1995
1997
1999
2001
2003
2005
2006
Year
Internal Medicine, 50–54 (Male)
Internal Medicine, 55–59 (Male)
General Surgery, 50–54 (Male)
General Surgery, 55–59 (Male)
Ob/Gyn, 50–54 (Male)
Ob/Gyn, 55–59 (Male)
Source
National Physician Database, Canadian Institute for Health Information.
Figure 12: Average Full-Time Equivalent Values of Physicians Who Billed Fee
for Service, Six Age–Specialty Cohorts, Canada, 1989 to 2006
1.20
1.00
FTE
0.80
0.60
0.40
0.20
0.00
1989
1991
1993
1995
1997
1999
2001
2003
2005
Year
Internal Medicine, 50–54 (Male)
Internal Medicine, 55–59 (Male)
General Surgery, 50–54 (Male)
General Surgery, 55–59 (Male)
Ob/Gyn, 50–54 (Male)
Ob/Gyn, 55–59 (Male)
Source
National Physician Database, Canadian Institute for Health Information.
40
2006
Chapter 6: An Alternative Perspective
As would be expected, those age 55 to 59 exited the medical workforce at a
faster rate than those age 50 to 54, and the average FTE values of the former
declined much faster than those of the latter. Those age 55 to 59 in 1989
were 72 to 76 years old at the end of the study period. By 2006, only 68 of
the original 276 members of the General Surgery, 55 to 59 cohort were still
clinically active and billing FFS, and the average FTE value had dropped from
a high of 0.86 in 1991 to 0.45 in 2006. Viewing the results from another angle,
one can see that about a quarter of the original members of this cohort were
still in FFS clinical practice at age 72 to 76. With an average FTE value of 0.45,
the 68 clinically active physicians were more or less equivalent to 30.6 general
surgeons working full time.
6.2 Changing Clinical Practice Profile
Other changes also take place as a physician gets older. It is equally
important to know what older physicians do and how their medical practice
profiles differ from those of their younger counterparts. Thus, another
aspect that deserves attention is scope of practice, which refers to the
comprehensiveness of a physician’s clinical practice or the basket of clinical
services he or she provides. This is especially relevant for FPs/GPs since, as
primary care practitioners, their scope of practice can be very broad and they
have more leeway in expanding or contracting their scope of practice. Past
studies on differences in scope of practice have also focused mostly
on FPs/GPs. For instance, Chan and Tepper documented a trend of declining
comprehensiveness of services provided by Canadian FPs/GPs over the
years.1, 41 Using survey data, Hutten-Czapski and associates and Pong and
Pitblado have shown that rural FPs/GPs tend to have a much broader scope
of practice than their urban counterparts, possibly to fill some of the service
gaps resulting from a lack of rural specialists.42, 43 Rural FPs/GPs are more
likely to provide services or perform clinical procedures that would typically
be done by specialists in larger urban centres. Similar findings have been
reported in other countries. Humphreys and associates, for example, found
that Australian GPs tend to have more complex practice patterns if
they work in more remote areas.44 A more recent study by Australian
researchers reaffirmed the importance of rural GPs providing procedural
medical services.45
The hypothesis is that, as doctors age, they are more likely to limit their scope
of practice, just as they tend to work fewer hours or see fewer patients. The
2007 NPS provides some indications of an inverse relationship between aging
and scope of practice. The survey asked FPs/GPs if they had reduced their
scope of practice in the two years prior to the survey and if they planned to
reduce their scope of practice in the coming two years (survey questions are
shown in Appendix A). The survey results are presented in Figure 13.
41
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Figure 13: Proportion of Family Physicians/General Practitioners Indicating
Reduction in Scope of Practice in Past Two Years and Plans
to Reduce Scope of Practice in Next Two Years, by Age Group
and Sex, Canada, 2007
Plan to Reduce Scope of Practice in Next Two Years
Reduced Scope of Practice in Past Two Years
30
Percentage
25
20
15
10
5
0
<35 35–44 45–54 55–64 65–74 75+ Total
<35 35–44 45–54 55–64 65–74 75+ Total
Age Group
Male
Female
Sources
CFPC/CMA/RCPSC National Physician Survey Database, 2007. “Protected by Copyright.”
The proportions of FPs/GPs indicating that they had reduced their scope of
practice in the past two years were not large (with the exception of female FPs/
GPs age 75 and older), but those who were older were somewhat more likely
to report scope-of-practice reduction in the past two years (male: younger
than 35, 4.3%; 75 or older, 10.8%; female: younger than 35, 5.8%; 75 or older,
19.5%). On average, 9.2% of male and 7.4% of female FPs/GPs reported a
reduction in scope of practice in the previous two years. When it comes to
plans for the future, the data shows a relationship between aging and scopeof-practice reduction, but only for male physicians. With the exception of those
age 75 and older, as male FPs/GPs got older, they were more likely to indicate
an intention to trim their scope of practice (younger than 35: 8.5%; 65 to 74:
17.9%). There were no clear patterns among female FPs/GPs. On average,
13.2% of male and 7.8% of female FPs/GPs indicated an intention to narrow
their scope of practice in the next two years. xiii
xiii. The 2007 NPS also showed that older specialists, similar to FPs/GPs, were more likely to report that
they had reduced their scope of practice in the previous two years and that they planned to reduce
their scope of practice in the next two years. The relationships between aging and scope-of-practice
reduction were more evident among male than female specialists.
42
Chapter 6: An Alternative Perspective
The above is self-reported information from a survey. What does FFS billing
data from the NPDB tell us about the practice patterns of younger and older
FPs/GPs? Figures 14 to 21 show the percentages of FPs/GPs providing
selected clinical services by age group and sex. The types of services
selected for this analysis are similar to those used in the study by Tepper.41, xiv
Physicians in different age groups did not differ substantially with respect to
some common primary care services, such as office assessments, services
requiring basic procedural skills and mental health services. Between 62.9%
and 76.6% of male FPs/GPs in age groups ranging from younger than 40 to
65 to 69, and between 65.2% and 76.2% of female FPs/GPs in age groups
ranging from younger than 40 to 60 to 64, provided office assessments.
Likewise, between 77.3% and 89.9% of male and female physicians in age
groups ranging from younger than 40 to 65 to 69 offered mental health
services.
However, for other types of service, age does appear to be inversely related
to the likelihood of service provision. For example, fewer than 35% of female
FPs/GPs age 65 to 69 provided hospital inpatient care, compared with 58.8%
of those younger than 40. Similarly, 56.4% of male FPs/GPs age 65 to 69
provided services requiring advanced procedural skills, compared with 77.4%
of those age 40 to 44. It also appears that the progressive narrowing of scope
of practice happened in a more intensive manner among female physicians.
For instance, there was a difference of 10.4 percentage points between
male FPs/GPs who were younger than 40 and those who were 60 to 64 with
respect to providing hospital inpatient care, but there was a difference of 20.9
percentage points between female FPs/GPs who were younger than 40 and
those who were 60 to 64. This suggests that age-related scope-of-practice
compression occurred faster among female physicians than among their
male counterparts.
Although not too many FPs/GPs were involved in providing obstetrical
services, surgical assistance and anesthesia services, similar age-related
scope-of-practice reduction trends were discernible. For instance, while 18.7%
of female FPs/GPs who were younger than 40 provided obstetrical services,
only 6.3% of those age 60 to 64 were still involved. Similarly, while 19.0% of
FPs/GPs younger than 40 provided anesthesia services, only 5.8% of those
age 65 to 69 still provided such services.
xiv. For details about the types of service and their National Grouping System category descriptions,
see Appendix A in Tepper.41
43
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Figure 14: Percentage of Family Physicians/General Practitioners Who Billed
Fee for Service and Were Providing Office Assessment Services,
by Age Group and Sex, Canada, 2006
80
70
Percentage
60
50
40
30
20
10
0
<40
40–44 45–49 50–54 55–59 60–64 65–69 70–74 75–79
80+
Average
Age Group
Male FPs/GPs
Female FPs/GPs
Source
National Physician Database, Canadian Institute for Health Information.
Figure 15: Percentage of Family Physicians/General Practitioners Who Billed
Fee for Service and Were Providing Hospital Inpatient Services,
by Age Group and Sex, Canada, 2006
70
60
Percentage
50
40
30
20
10
0
<40
40–44 45–49 50–54 55–59 60–64 65–69 70–74 75–79
Age Group
Male FPs/GPs
Female FPs/GPs
Source
National Physician Database, Canadian Institute for Health Information.
44
80+
Average
Chapter 6: An Alternative Perspective
Figure 16: Percentage of Family Physicians/General Practitioners Who Billed Fee
for Service and Were Providing Basic Procedural Skills Services, by
Age Group and Sex, Canada, 2006
90
80
70
Percentage
60
50
40
30
20
10
0
<40
40–44 45–49 50–54 55–59 60–64 65–69
70–74 75–79
80+
Average
Age Group
Male FPs/GPs
Female FPs/GPs
Source
National Physician Database, Canadian Institute for Health Information.
F
Figure 17: Percentage of Family Physicians/General Practitioners Who Billed Fee
for Service and Were Providing Advanced Procedural Skills Services,
by Age Group and Sex, Canada, 2006
80
70
Percentage
60
50
40
30
20
10
0
<40
40–44 45–49 50–54 55–59 60–64 65–69 70–74 75–79
80+
Average
Age Group
Male FPs/GPs
Female FPs/GPs
Source
National Physician Database, Canadian Institute for Health Information.
45
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Figure 18: Percentage of Family Physicians/General Practitioners Who Billed Fee
for Service and Were Providing Obstetrical Services, by Age Group
and Sex, Canada, 2006
20
18
Percentage
16
14
12
10
8
6
4
2
0
<40
40–44 45–49 50–54 55–59 60–64 65–69 70–74 75–79
80+
Average
Age Group
Male FPs/GPs
Female FPs/GPs
Source
National Physician Database, Canadian Institute for Health Information.
Figure 19: Percentage of Family Physicians/General Practitioners Who Billed Fee
for Service and Were Providing Mental Health Services, by Age Group
and Sex, Canada, 2006
90
80
70
Percentage
60
50
40
30
20
10
0
<40
40–44 45–49 50–54 55–59 60–64 65–69 70–74 75–79
Age Group
Male FPs/GPs
Female FPs/GPs
Source
National Physician Database, Canadian Institute for Health Information.
46
80+
Average
Chapter 6: An Alternative Perspective
Figure 20: Percentage of Family Physicians/General Practitioners Who Billed
Fee for Service and Were Providing Surgical Assistance Services,
by Age Group and Sex, Canada, 2006
25
Percentage
20
15
10
5
0
<40
40–44 45–49 50–54 55–59 60–64 65–69 70–74 75–79
80+
Average
Age Group
Male FPs/GPs
Female FPs/GPs
Source
National Physician Database, Canadian Institute for Health Information.
Figure 21: Percentage of Family Physicians/General Practitioners Who Billed
Fee for Service and Were Providing Anesthesia Services,
by Age Group and Sex, Canada, 2006
20
18
16
Percentage
14
12
10
8
6
4
2
0
<40
40–44 45–49 50–54 55–59 60–64 65–69 70–74 75–79
80+
Average
Age Group
Male FPs/GPs
Female FPs/GPs
Source
National Physician Database, Canadian Institute for Health Information.
47
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
6.2.1 Second Case Study
The above analysis (figures 14 to 21) relies on cross-sectional data to
demonstrate the relationships between age and the likelihood of engaging in
various types of FFS clinical service by FPs/GPs. A second case study was
conducted using longitudinal data to show how the practice profiles of FPs/
GPs who billed FFS changed as they grew older. The approach is similar to
that used in the first case study, except that the second case study focuses
not just on changes in numbers of active physicians and their FTEs but also
on changes in what they did. Four cohorts of FPs/GPs—males age 50 to 54
(n = 1,527), males age 55 to 59 (n = 1,388), females age 50 to 54 (n = 201)
and females age 55 to 59 (n = 167) in 1989—were identified and followed
over an 18-year period from 1989 to 2006 (for presentation reasons, data for
only 10 years is shown in tables E4 to E7 in Appendix E). For each year, the
number of FPs/GPs in each cohort who still billed FFS was obtained. Also
recorded were the average FTE values for those physicians who billed for
more than zero dollars in the service categories included in this analysis.
Physicians from the territories were not included in the analysis. If physicians
billed in more than one province, they were counted as one physician and
their FTE values were summed.
The tracking involves identifying the number of FPs/GPs in a particular age–
sex cohort who billed for more than zero dollars in 1989 and in each of the
subsequent years up to 2006. In addition to showing changes from year to
year in the number of physicians remaining in the cohort and their FTEs, the
tracking involved quantifying the extent to which FPs/GPs engaged in certain
clinical activities, such as providing hospital inpatient care, services requiring
advanced procedural skills and surgical assistance. The results are presented
in figures 22 to 25.
Figure 22 (and Table E4 in Appendix E) presents findings in relation to the
cohort of male FPs/GPs age 50 to 54 in 1989. There were 1,527 FPs/GPs in
this cohort in 1989. By 2006, when they were age 67 to 71, 785 (or 51.4%) were
still clinically active and billing FFS for more than zero dollars. The average
FTE value of those who were still in clinical practice declined from a high of
1.02 in 1991 to 0.73 in 2006, representing a 28.4% reduction in billable clinical
activities. These findings are similar to those for specialists, as shown in the
first case study. Equally important are the changes, or lack thereof, over time
in the extent to which physicians engaged in certain clinical activities. For
instance, it can be seen in Figure 22 that there were no major changes in the
provision of care requiring basic procedural skills and mental health services.
There was a slight progressive decline in the provision of hospital inpatient
services and services requiring advanced procedural skills. However, there
48
Chapter 6: An Alternative Perspective
was a more substantial progressive decline in the provision of obstetrical
services, from 31.6% in 1989 to 11.4% in 1999 and to 4.2% in 2006, though
the proportion of FPs/GPs doing obstetrics was relatively small at the outset.
Figure 23 (and Table E5 in Appendix E) focuses on the cohort of male FPs/
GPs age 55 to 59 in 1989. About one-third of the cohort members were still
clinically active in 2006, when they reached the age of 72 to 76, and the
average FTE value of those who were still in clinical practice and billing FFS
was 0.59 at the end of the study period. With respect to the relationship
between aging and scope of practice, the findings are similar to those in the
50-to-54 age cohort.
Figures 24 and 25 (and tables E6 and E7 in Appendix E) present findings for
the cohort of female FPs/GPs age 50 to 54 in 1989 and age 55 to 59 in 1989,
respectively. Generally speaking, the patterns are similar to those for their
male colleagues, but a few differences are worth noting. First, the female
cohorts were much smaller in size than the male cohorts. As a result, random
variations or outlier effects may explain some of the year-to-year fluctuations in
the proportions of physicians engaging in some medical services, particularly
those involving relatively few physicians (such as anesthesia and obstetrical
services). Second, the FTE values of female FPs/GPs were consistently
smaller than those of their male counterparts.
Figure 22: Percentage of Active Male Family Physicians/General Practitioners
Age 50 to 54 in 1989 Involved in Eight Types of Clinical Service,
Canada, 1989 to 2006
100
90
Percentage
80
70
60
50
40
30
20
10
0
1989
1991
1993
1995
1997
1999
2001
2003
2005
2006
Year
Providing Office Assessment
Providing Basic Procedural Skills
Providing Advanced Procedural Skills
Providing Anesthesia
Providing Surgical Assistance
Providing Obstetrical Services
Providing Mental Health Services
Providing Inpatient Services
Source
National Physician Database, Canadian Institute for Health Information.
49
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Figure 23: Percentage of Active Male Family Physicians/General Practitioners
Age 55 to 59 in 1989 Involved in Eight Types of Clinical Service,
Canada, 1989 to 2006
100
90
Percentage
80
70
60
50
40
30
20
10
0
1989
1991
1993
1995
1997
1999
2001
2003
2005
2006
Year
Providing Office Assessment
Providing Basic Procedural Skills
Providing Advanced Procedural Skills
Providing Anesthesia
Providing Surgical Assistance
Providing Obstetrical Services
Providing Mental Health Services
Providing Inpatient Services
Source
National Physician Database, Canadian Institute for Health Information.
Figure 24: Percentage of Active Female Family Physicians/General Practitioners
Age 50 to 54 in 1989 Involved in Eight Types of Clinical Service,
Canada, 1989 to 2006
100
90
Percentage
80
70
60
50
40
30
20
10
0
1990
1991
1993
1995
1997
1999
2001
2003
2005
Year
Providing Office Assessment
Providing Basic Procedural Skills
Providing Advanced Procedural Skills
Providing Anesthesia
Providing Surgical Assistance
Providing Obstetrical Services
Providing Mental Health Services
Providing Inpatient Services
Source
National Physician Database, Canadian Institute for Health Information.
50
2006
Chapter 6: An Alternative Perspective
Figure 25: Percentage of Active Female Family Physicians/General Practitioners
Age 55 to 59 in 1989 Involved in Eight Types of Clinical Service,
Canada, 1989 to 2006
100
90
Percentage
80
70
60
50
40
30
20
10
0
1989
1991
1993
1995
1997
1999
2001
2003
2005
2006
Year
Providing Office Assessment
Providing Basic Procedural Skills
Providing Advanced Procedural Skills
Providing Anesthesia
Providing Surgical Assistance
Providing Obstetrical Services
Providing Mental Health Services
Providing Inpatient Services
Source
National Physician Database, Canadian Institute for Health Information.
This case study shows that while some FPs/GPs were no longer clinically
active (or no longer billing FFS) before they reached age 65, significant
proportions of physicians in their late 60s and in their 70s were still clinically
active and billing FFS, though their FTE values declined progressively as they
got older. On the whole, there is considerable empirical evidence to support
the hypothesis that as FPs/GPs get older they are more likely to limit what
they do. Although the provision of some core primary care services remained
largely unaffected over the entire study period, the likelihood of a physician
providing other types of service was very much related to age. As physicians
grew older, they were increasingly less likely to provide such services as
hospital inpatient care, obstetrics and services requiring advanced procedural
skills. The implications of these findings are discussed in the next chapter.
51
Chapter 7
Discussions and Conclusion
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
7.1 Summary of Major Findings
Policy-makers and health workforce planners have identified aging of the
health workforce, especially the medical workforce, as an important health
human resources issue that deserves close attention. It is commonly believed
that as more and more physicians approach the traditional retirement age
of 65, the number of physicians retiring will grow. But does this mean that
most physicians will put away their stethoscopes for good at age 65? This
study is an attempt to go beyond the conventional notion of retirement to
understand the complexity of the way older physicians work. Starting with
a review of the relevant literature, the study is built on what is known about
retirement in general and physician retirement in particular. It used data from
several sources—the 2007 NPS, SMDB, CMA Master File and NPDB—to paint
a more complete picture of the medical practice profile of older physicians
and to ensure that the findings are not an artefact of a single data source
because, without a uniform definition of retirement, each database may
have captured retirement information somewhat differently and counted the
number of active or retired physicians somewhat differently. It examined both
retirement intentions and behaviours, without assuming that they are the same
phenomenon. As well, it tried to estimate the number of minimally active older
physicians and asked if it would make a significant difference if they were
counted as retired. Most important of all, the study proposed an alternative
approach to understanding the impact of aging on medical practice.
Do physicians typically retire when they turn 65? Are physicians increasingly
opting for early retirement? Evidence from existing studies suggests that
Canadian physicians tend to quit work later than average workers. Also, as
far as physicians are concerned, retirement is anything but an either/or issue.
Instead of dropping out of the medical workforce abruptly and completely at
age 65, many older physicians choose to remain in clinical practice, though
not necessarily maintaining the same activity level or doing the same kind of
work as when they were younger.
Depending on the source of data used, one gets different estimates of the
extent of physician retirement. This may be due to the fact that various
databases define retirement differently, capture the number of retirees
differently and/or have different criteria for including physicians in or excluding
them from the base population. On the basis of self-reported retirement
intentions from the 2007 NPS, about 3.2% of all physicians planned to retire in
each of the two years following the survey. On the other hand, the estimated
average annual retirement rates were 0.54%, based on three years of SMDB
data (including semi-retirement), and 0.79%, based on data from the CMA
Master File for the same three-year period. If one uses these estimated
retirement rates as projection parameters to forecast the size of the Canadian
medical workforce 25 or 30 years into the future, one is likely to obtain
54
Chapter 7: Discussions and Conclusion
substantially different medical workforce projections, assuming that all
other variables remain constant. On the basis of these projections, one could
come to different conclusions about the future sufficiency of physician supply
in Canada.
An attempt was then made to estimate the number of older physicians
who were minimally active in clinical practice. Depending on what “older
physicians” refers to and where the FTE threshold is set, different proportions
of older physicians can be considered minimally active—ranging from 4.9%
of those age 65 and older to 7.3% of those age 55 and older, with 2007
as the reference year. Whether these minimally active physicians should
be considered retired for the purpose of medical workforce projections or
planning is not just a technical but also a policy issue that needs further
consideration and deliberation by stakeholders.
Because the concept of retirement is not clearly defined, especially for physicians,
and since there is as yet no consensus on what physician retirement means
and how it should be measured, this study suggested a different way of
understanding how aging affects the way physicians work. If retirement is
understood to mean the complete cessation of medical practice, it should
be seen as the end point of a continuum of changes in medical practice
as a physician gets older. But prior to exiting the medical workforce, many
other changes in medical practice may have taken place, such as reduction
in workload, scope-of-practice compression or greater involvement in nonclinical work, which may also have implications for medical care provision
and physician workforce planning.
While physicians in the baby-boom generation will exit the medical workforce
in greater numbers in the coming years, many physicians age 65 and older
are likely to remain active in clinical practice, if trends from the recent past
continue. But data presented in Chapter 6 shows that their workload, as
measured by average FTE values among physicians who billed FFS, tends
to decline as they become older. Also, the scope of their practice tends to
become narrower, as older physicians relinquish some types of clinical work
while retaining others. The case of older FPs/GPs was used as an illustration.
Although there were no major differences between FPs/GPs in different age
groups with respect to such core clinical activities as office assessments and
mental health care, the older FPs/GPs became, the less likely they were to
engage in such activities as hospital inpatient care, obstetrics, anesthesia
and services requiring advanced procedural skills. Similar findings have been
reported by other researchers. For example, Chan and associates found that
Ontario FPs/GPs age 65 and older were less likely than those younger than 65
to perform obstetric deliveries (4.6% versus 16.9%), house calls (38.7% versus
60.4%), minor procedures (38.7% versus 62.3%) and emergency department
work (1.1% versus 14.8%).29
55
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
7.2 Limitations and Future Actions
This study has several limitations, which are discussed as follows. Many of the
limitations are related to inadequate data and/or lack of conceptual clarity or
consistency. Each identified limitation calls for remedial actions or may offer
opportunities for further research.
a. Although three sets of estimated retirement rates were produced, each
based on a different source of data, there is no way of knowing which one
is closer to the true physician retirement situation. This is because there is
as yet no consensus on what physician retirement means, and no existing
physician database has come up with a foolproof way of identifying and
counting retired or retiring physicians. Thus, existing data on retirement
and retirement projections should be used with caution and treated as
tentative. It is for this reason that this study has proposed an alternative
perspective on physician retirement, one that takes into account the
complexity of medical practice in the later years of a physician’s career.
b. This study used data from administrative databases (the SMDB and
CMA Master File) and a survey (the 2007 NPS) to estimate the extent of
physician retirement and found that estimated retirement rates based on
the latter are substantially higher than those based on the former. Although
it is premature to recommend which type of data should be used, since
retirement intentions are often taken to mean actual retirement, there is a
need to examine the extent to which intentions translate into behaviours.
It would be useful to conduct studies in Canada similar to those carried
out in the United States by Konrad and Dall and by Rittenhouse and
associates.36, 46 Their findings show that self-reported intentions to quit
clinical practice do not necessarily correspond with actual behaviours.
Studies of a similar nature in Canada would help health workforce planners
more accurately interpret survey data about retirement intentions or plans.
As Rittenhouse and colleagues warned,
increasing reliance on proxy variables for physician attrition such as
intention-to-quit is equally concerning in light of research that suggests
that “intention to . . .” variables are not strongly correlated with actual
behavior. The possibility that current measures of physician attrition
are not valid has important policy implications, particularly if these data
are used in forecasting models that inform policy decisions regarding
physician supply.46
At the same time, there is a need to assess the veracity of retirement
variables in databases such as the SMDB and the CMA Master File. It is not
known at this time the extent to which retired or retiring physicians have
been captured by these and other databases. It is also not known whether
retirement in one database is equivalent to retirement in another database
56
Chapter 7: Discussions and Conclusion
with respect to definition and data capture. Also, although these sources of
data are usually called administrative databases, with the impression that
the data is objectively derived, their retirement information may actually be
subjective in nature. For instance, retirement status in the SMDB is selfdefined and self-reported by physicians, and there is no way of knowing
how they determine whether they are active, semi-retired or fully retired. As
a result, rather than reflecting retirement behaviours, the data may simply
be physicians’ own perceptions of their work status.
c. Much of the discussion in chapters 5 and 6 concerning older physicians
who are minimally active and changing practice profiles is based on FFS
billing information. But not all physicians bill FFS, and a growing number
of physicians are shifting to alternative payment schemes or blended
reimbursement models. This study has also not addressed the possibility
that physicians who bill FFS and physicians who do not bill FFS may have
different practice patterns. Moreover, FFS billing and FTE information from
the NPDB tells us something about billable clinical services but nothing
about other activities, such as administration, teaching, consulting
and research.
However, it is worth noting that the NPDB is meant to be a multi-phase
project. Phase I, which has been completed, is intended to capture,
among other things, information on activity levels of and FFS payments to
physicians. Phase II, which is currently under way, will add data on clinical
activities paid under alternative remuneration schemes (such as salaries
and sessional fees), and Phase III will further augment the database with
information about non-clinical activities, such as administration, teaching
and research.47 Thus, when phases II and III are fully implemented, the
NPDB should be able to shed more light on more aspects of medical
practice for more physicians. At that time, we should be able to get a more
complete picture of what physicians do and how much they do throughout
their entire medical careers.
d. This study has shown that the practice profile of FPs/GPs tends to become
narrower as they get older, but it is not known if there are similar changes
to the practice patterns of aging medical and surgical specialists. Other
researchers are encouraged to fill this knowledge gap by examining
possible differences between younger and older specialists with respect to
the way they practise.
e. Similarly, this study has not explored the implications of scope-of-practice
contraction. For example, it is possible that rural Canadians will feel the
impact much more than urban residents because, as Pong and Pitblado
and others have found, rural Canada relies mostly on FPs/GPs, and
rural FPs/GPs tend to have a broader scope of practice than their urban
counterparts.43 If many rural FPs/GPs reduce their practice scope as well
57
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
as the amount they work as a result of aging, the effects on medical care
provision in rural Canada could be severe. The implications, especially
for special populations such as rural residents, of an aging physician
workforce and concomitant changes in the way medicine is practised
should receive greater policy and research attention.
f. This study has alluded to the fact that some aging physicians may give
up clinical work but not other related activities, such as administration,
research, teaching and consulting. However, it has not examined these
other activities in any detail, mostly because no reliable data is available
on physicians’ work other than clinical practice. This is also an area that
deserves research attention. Furthermore, if information about non-clinical
activities is deemed important for health care planning purposes, efforts
will have to be made to collect such information in a systematic manner
and on a national basis.
7.3 Conclusion
The findings of this study suggest that the impact of an aging medical
workforce should be examined from the broader perspective of changing
medical practice patterns rather than from the narrow focus on retirement.
Aging will impact not only the number of physicians reaching age 65 or exiting
the medical workforce but also what older physicians do and how much they
do. The conventional notion of retirement may no longer be very useful in
medical workforce planning, as there is still no consensus on what it means
and no reliable measurements of the extent of retirement. This is borne out
in this study by the fact that different databases yield different retirement
rates. Likewise, the traditional approach of projecting future physician supply,
which is based in part on unproven retirement assumptions and possibly
unreliable retirement figures, is no longer adequate. As the proportion of older
physicians increases, understanding what they do, how much they do, how
long they stay active and so forth is becoming increasingly important
and urgent.
While it is tempting to suggest that the concept of retirement be retired, it may
be premature or unrealistic, as the notion of retirement is so deeply ingrained
in our thinking and daily discourse and so firmly entrenched in social and
administrative arrangements. Thus, the following actions are suggested as
interim measures:
• Physician data gathering agencies and those involved in physician
workforce planning and research need to work together to clarify what
retirement means as far as physicians are concerned. They may even
have to adopt an arbitrary definition of retirement—like deciding whether
a 70-year-old doctor who spends a morning each week seeing patients
should be considered retired—to achieve a more or less uniform
58
Chapter 7: Discussions and Conclusion
measurement until a better or more permanent solution is found. An attempt
has been made in this study to assess the number of older physicians who
are minimally active. It is hoped that the initial findings will stimulate more
research and discussions.
• As this study has shown, existing sources of data, regardless of whether
they are administrative or survey data, tend to be inconsistent and
problematic in how they capture retirement information and are likely to be
less than adequate in generating reliable counts of retiring physicians and
retirement rates. Ways need to be found so that databases are as consistent
as possible in how they define retirement and how they count retirees,
notwithstanding the fact that different databases serve different purposes.
• In addition to assessing the adequacy of retirement rates used in current
physician workforce projections, researchers need to develop a strategy
to transition from traditional physician workforce planning and projection
methods that focus on head counts to a new approach that takes into
consideration the changing practice patterns of aging physicians.
The subtitle of this study is Toward a New Perspective on Physician Retirement.
The word “toward” implies that the study serves as a point of departure, rather
than the end point. Adopting a new perspective on physician retirement will
require more than collecting better data and conducting more investigations,
even though such activities are important and urgently needed. There are
other equally critical issues that need to be addressed.
Physician workforce enhancement efforts in Canada have tended to focus on
supply issues by increasing medical school enrolments and enabling more
international medical graduates to practise in Canada. It is argued that much
more should be done to encourage older physicians to remain in the medical
workforce. Hall surveyed senior academic pediatricians to find out the extent
to which they wished to carry on working after the usual age of retirement.48
The survey, conducted in both the United States and Canada, found that many
of the respondents wanted to continue to use their skills and experience and
identified several areas of work that they were interested in pursuing, such
as editing and writing, international health, working with research networks,
teaching and consulting. It appears that creating more flexible practice
opportunities is worth considering. As a first step, it may be useful to survey
older physicians in Canada to determine what would motivate them to stay in
clinical practice longer. Would they be willing to continue to work if there were
less onerous on-call requirements, lighter patient loads, greater flexibility for
vacation, easier access to skills-maintenance programs or more opportunities
for teaching or research?
As noted earlier, more and more physicians are giving up FFS and opting
for alternative payment models. It also appears that physicians are becoming
increasingly open to the idea of participating in government-sponsored
59
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
pension plans. xv Will such policies have unintended consequences? Will
the introduction of alternative payment programs or government pension
plans change physicians’ retirement behaviours in the future? Although it is
beyond the scope of the present study to explore such issues, they are clearly
important for policy-makers to consider and for researchers to monitor
and assess.
The present study focused exclusively on physicians. But physicians do not
practise in isolation. They work with and are supported by other health care
providers, many of whom face similar population-aging and workforce-attrition
challenges, xvi though they may respond differently to such challenges. It is
thus critical to examine the future supply and practice patterns of physicians in
the broader context of the aging of the Canadian health workforce in general.
xv. Several news items and opinion pieces provide some indications of an emerging interest in public
pensions for physicians.49, 50, 51
xvi. A series of analyses on the geographic distribution and internal migration of health care workers
by Pitblado showed that many health care occupations are aging faster than the general
Canadian workforce.52
60
Appendix A: Questions Regarding Changes to Medical Practice in the 2007 National Physician Survey
Appendix A: Questions Regarding
Changes to Medical Practice in the
2007 National Physician Survey
The 2007 National Physician Survey asked the following questions regarding
changes to medical practice:
With reference to the last 2 years, please check all of the following changes
you have already made. With reference to the next 2 years, please check all
of the following changes that you are planning to make:
• Reduce weekly work hours (excluding on-call) [asked in both short and
long questionnaires]
• Increase weekly work hours (excluding on-call) [asked in both short
and long questionnaires]
• Retire from clinical practice [asked in both short and long questionnaires]
• Reduce scope of practice [asked in long questionnaire only]
61
Appendix B: NPDB Physician Specialty Categories
Appendix B: NPDB Physician
Specialty Categories
Family Physicians and
General Practitioners
01 Family Medicine
010 Residency
011 General Practice
012 Family Practice
013 Community Medicine/
Public Health
014 Emergency Medicine
Medical Specialists
02 Internal Medicine
020 General Internal Medicine
021 Cardiology
022 Gastroenterology
023 Respiratory Medicine
024 Endocrinology
025 Nephrology
026 Hematology
027 Rheumatology
028 Clinical Immunology
and Allergy
030 Oncology
031 Geriatrics
032 Tropical Medicine
035 Genetics
04 Neurology
040 Neurology and EEG
041 Neurology
042 EEG
05 Psychiatry
050 Psychiatry and Neuropsychiatry
051 Psychiatry
052 Neuropsychiatry
06 Pediatrics
060 Pediatrics
07 Dermatology
065 Dermatology
08 Physical Medicine/Rehabilitation
070 Physical Medicine and
Rehabilitation
071 Electromyography
09 Anesthesia
075 Anesthesia
Surgical Specialists
10 General Surgery
080 General Surgery
11 Thoracic/Cardiovascular
Surgery
086 Thoracic Surgery
087 Cardiovascular Surgery
088 Cardiovascular/
Thoracic Surgery
12 Urology
090 Urology
13 Orthopedic Surgery
095 Orthopedic Surgery
14 Plastic Surgery
100 Plastic Surgery
15 Neurosurgery
110 Neurosurgery
16 Ophthalmology
115 Ophthalmology
116 Ophthalmology/
Otolaryngology
17 Otolaryngology
120 Otolaryngology
18 Obstetrics/Gynecology
126 Obstetrics
127 Gynecology
128 Obstetrics/Gynecology
Note
Although genetics is no longer a subspecialty of internal medicine, it is included in
the internal medicine category because the number of physician records assigned
to this specialty is relatively small.
63
Appendix C: Data for Chapter 4
Appendix C: Data for Chapter 4
Table C1: Percentage of Family Physicians/General Practitioners and Specialists
Planning to Retire From Clinical Practice in the Next Two Years,
by Age Group and Sex, Canada, 2007
FPs/GPs
Specialists
Age Group
Male
Female
Total
Male
Female
Total
<35
1.1%
0.5%
0.7%
1.1%
0.3%
0.7%
35–44
0.7%
0.7%
0.7%
0.7%
0.2%
0.5%
45–54
1.8%
2.1%
2.0%
1.0%
1.9%
1.3%
55–64
11.4%
11.6%
11.4%
9.8%
10.2%
9.6%
65–74
33.5%
26.5%
32.3%
27.5%
29.6%
27.5%
75+
33.1%
*
35.4%
28.2%
*
27.8%
Total
6.2%
6.5%
Note
* Responses suppressed; column number is less than 30.
Sources
2007 National Physician Survey, College of Family Physicians of Canada,
Canadian Medical Association and Royal College of Physicians and Surgeons of Canada.
Table C2: 2007, 2006 and 2005 Activity Status of Physicians Who Were Active
in 2006, 2005 and 2004, Respectively, by Age Group and Sex, Canada
Age Group
Active in
2006
Active in
2007
Semi-Retired
in 2007
Retired
in 2007
Deceased
in 2007
<35
Male
Female
2,412
2,727
2,352
2,658
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
35–44
Male
Female
8,901
7,075
8,789
7,005
0 (0.00%)
0 (0.00%)
1 (0.01%)
0 (0.00%)
3 (0.03%)
2 (0.03%)
45–54
Male
Female
12,426
7,055
12,332
6,995
1 (0.01%)
1 (0.01%)
2 (0.02%)
4 (0.06%)
8 (0.06%)
1 (0.01%)
55–64
Male
Female
11,191
3,109
11,018
3,062
13 (0.12%)
4 (0.13%)
29 (0.26%)
13 (0.42%)
25 (0.23%)
1 (0.03%)
65–74
Male
Female
5,141
634
4,908
604
35 (0.69%)
4 (0.65%)
87 (1.72%)
9 (1.46%)
27 (0.53%)
1 (0.16%)
75+
Male
Female
1,208
119
1,130
114
7 (0.59%)
0 (0.00%)
29 (2.45%)
3 (2.56%)
16 (1.35%)
0 (0.00%)
61,998
60,967
65 (0.11%)
177 (0.29%)
84 (0.14%)
Total
65
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Table C2: 2007, 2006 and 2005 Activity Status of Physicians Who Were Active
in 2006, 2005 and 2004, Respectively, by Age Group and Sex,
Canada (cont’d)
Active in
2005
Active in
2006
Semi-Retired
in 2006
Retired
in 2006
Deceased
in 2006
<35
Male
Female
2,482
2,743
2,341
2,642
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
1 (0.04%)
35–44
Male
Female
9,230
7,029
9,045
6,924
0 (0.00%)
2 (0.02%)
2 (0.22%)
2 (0.02%)
4 (0.44%)
0 (0.00%)
45–54
Male
Female
12,523
6,771
12,342
6,649
1 (0.00%)
2 (0.03%)
7 (0.06%)
12 (0.18%)
10 (0.08%)
2 (0.03%)
55–64
Male
Female
11,031
2,790
10,735
2,697
18 (0.17%)
10 (0.37%)
45 (0.42%)
17 (0.62%)
29 (0.27%)
2 (0.07%)
65–74
Male
Female
4,927
596
4,485
540
53 (1.12%)
6 (1.07%)
169 (3.57%)
14 (2.50%)
28 (0.59%)
0 (0.00%)
75+
Male
Female
1,119
107
960
92
21 (1.98%)
1 (1.01%)
59 (5.57%)
6 (6.06%)
20 (1.89%)
0 (0.00%)
61,348
59,452
114 (0.19%)
333 (0.56%)
96 (0.16%)
Total
Active in
2004
Active in
2005
Semi-Retired
in 2005
Retired
in 2005
Deceased
in 2005
<35
Male
Female
2,557
2,803
2,458
2,720
0 (0.00%)
0 (0.00%)
0 (0.00%)
0 (0.00%)
1 (0.04%)
1 (0.04%)
35–44
Male
Female
9,420
7,039
9,254
6,940
1 (0.01%)
0 (0.00%)
0 (0.00%)
2 (0.03%)
2 (0.02%)
2 (0.03%)
45–54
Male
Female
12,726
6,487
12,569
6,396
0 (0.00%)
0 (0.00%)
4 (0.03%)
6 (0.09%)
14 (0.11%)
5 (0.08%)
55–64
Male
Female
10,566
2,487
10,366
2,415
16 (0.15%)
3 (0.12%)
33 (0.32%)
16 (0.62%)
25 (0.24%)
2 (0.08%)
65–74
Male
Female
4,746
537
4,473
498
41 (0.89%)
3 (0.59%)
91 (1.97%)
10 (1.95%)
26 (0.56%)
1 (0.20%)
75+
Male
Female
985
99
881
92
17 (1.78%)
2 (2.12%)
41 (4.19%)
0 (0.00%)
16 (1.67%)
0 (0.00%)
60,452
59,062
83 (0.14%)
203 (0.34%)
95 (0.16%)
Total
Note
Not shown in this table are numbers in the Other category, which includes physicians who were abroad,
in the military, not in practice, on sabbatical, on a leave of absence, temporarily not in practice or
unknown.
Source
Scott’s Medical Database, Canadian Institute for Health Information.
66
Appendix C: Data for Chapter 4
Table C3: Total Number of Physicians and Number of Retired and Deceased
Physicians, by Age Group and Sex, Canada, 2007, 2006 and 2005
Total Count of
Physicians at
Beginning of Year
Number Retired
at End of Year
Number Deceased
at End of Year
2007
<35
Male
Female
1,964
2,246
1 (0.05%)
6 (0.27%)‡
0 (0.00%)
0 (0.00%)
35–44
Male
Female
9,319
7,015
12 (0.13%)
13 (0.19%)‡
6 (0.06%)
5 (0.07%)
45–54
Male
Female
12,751
7,062
17 (0.13%)
15 (0.21%)
16 (0.13%)
1 (0.01%)
55–64
Male
Female
11,403
3,251
72 (0.63%)
39 (1.20%)
28 (0.25%)
6 (0.18%)
65–74
Male
Female
5,647
730
241 (4.27%)
29 (3.97%)
40 (0.71%)
1 (0.14%)
75+*
Male
Female
1,201
110
34 (2.83%)
6 (5.45%)
20 (1.67%)
0 (0.00%)
777
343
0
1
1
1
62,699
485 (0.77%)
123 (0.20%)
2,060
2,323
1 (0.05%)
13 (0.56%)‡
1 (0.05%)
1 (0.04%)
9,526
6,955
7 (0.07%)
27 (0.39%)‡
4 (0.04%)
1 (0.01%)
12,772
6,753
17 (0.13%)
10 (0.15%)
9 (0.07%)
1 (0.01%)
11,177
2,883
77 (0.69%)
19 (0.66%)
30 (0.27%)
1 (0.03%)
5,497
684
302 (5.49%)
50 (7.31%)
38 (0.69%)
2 (0.29%)
1,201
108
122 (10.16%)
13 (12.00%)
20 (1.67%)
1 (0.93%)
718
315
61,939
1
2
658 (1.06%)
0
0
109 (0.18%)
Missing Age
Male
Female
Total†
2006
<35
Male
Female
35–44
Male
Female
45–54
Male
Female
55–64
Male
Female
65–74
Male
Female
75+*
Male
Female
Missing Age
Male
Female
Total†
67
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Table C3: Total Number of Physicians and Number of Retired and Deceased
Physicians, by Age Group and Sex, Canada, 2007, 2006
and 2005 (cont’d)
Total Count of
Physicians at
Beginning of Year
Number Retired
at End of Year
Number Deceased
at End of Year
2005
<35
Male
Female
2,159
2,364
0 (0.00%)
14 (0.59%)‡
0 (0.00%)
0 (0.00%)
35–44
Male
Female
9,597
6,922
7 (0.07%)
33 (0.48%)‡
4 (0.04%)
2 (0.03%)
45–54
Male
Female
12,924
6,429
21 (0.16%)
16 (0.25%)
10 (0.08%)
7 (0.11%)
55–64
Male
Female
10,682
2,544
69 (0.65%)
23 (0.90%)
31 (0.29%)
3 (0.12%)
65–74
Male
Female
5,191
590
117 (2.25%)
17 (2.88%)
30 (0.58%)
1 (0.17%)
75+*
Male
Female
980
102
5 (0.51%)
1 (0.98%)
14 (1.43%)
1 (0.98%)
Missing Age
Male
Female
763
335
1
1
2
1
60,484
323 (0.53%)
103 (0.17%)
Total†
Notes
* Excludes all physicians older than 80.
† Total number of physicians less number of physicians with missing age.
‡ Some of the retired female physicians may have taken an extended maternity leave and may return to
active practice later on.
Source
Canadian Medical Association Master File, Canadian Medical Association.
68
Appendix D: Estimating the Number of Older Physicians Who Were Minimally Active
Appendix D: Estimating the
Number of Older Physicians Who
Were Minimally Active
Physicians are considered minimally active if their full-time equivalent (FTE)
value falls from active to below a designated lower threshold for a period
of at least three years.
Age
Physicians’ age is calculated as the difference between their birthdate and
the final day of the fiscal year being analyzed (March 31).
Active
Active FTE thresholds were calculated for both male and female physician
groups of family medicine, medical specialists and surgical specialists for
each fiscal year. The active FTE thresholds were based on the 40th percentile
FTE value of each designated group for each specific data year. When a
physician’s FTE value was equal to or greater than the calculated active
threshold, for a minimum of three consecutive years, that physician was
classified as active. A single physician’s active FTE value was calculated
as the average of the most recent consecutive active years.
Minimally Active
For physicians to be considered minimally active, they must be considered
active first and then have their FTE value fall below a calculated lower FTE
threshold. The lower FTE threshold was calculated by selecting a proportion
of the active FTE value for a physician. There were three proportional levels
used in calculating a physician’s lower FTE threshold: 15%, 25% and 33%.
The lower threshold would mean that a physician’s workload had dropped
below 15%, 25% or 33% of what his or her average FTE value was during the
active period. A physician’s FTE value would need to be at or below the lower
FTE threshold for at least two consecutive years immediately preceding the
reference year of the analysis for him or her to be considered minimally active.
Data Gaps
Any physicians with data gaps between their active period and minimally
active period were excluded. A physician had to have data in each
consecutive year between active status and the reference year of the analysis
to be included.
69
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Multiple Provinces and Specialties
Physicians working in multiple provinces had their FTE values summed
to ensure that each physician had only one FTE value for each year. If the
physician had FTE values under different specialties, the FTE values were
summed and the physician was assigned the specialty of the greater FTE value.
Total Clinical Payments
Newfoundland and Labrador, New Brunswick and Prince Edward Island each
submit physician-level alternative payment data to the National Physician
Database. The physician-level alternative payment data was combined with
physician-level fee-for-service payment data to calculate a total clinical
payment value for physicians in these three provinces. Fee-for-service
benchmark values were used to estimate an FTE value based on total clinical
payments. These estimated FTE values were used for physicians in these
three provinces.
Visual representations of the methodology are below.
Figure D1 displays an example of a physician’s FTE records, using 2007
as the reference year for analysis. In this situation, the physician would be
included in the analysis and considered retired because he or she had met
the minimum three-year active period (1989 to 1996), had FTE values below
the lower FTE threshold for the reference year and at least two immediate
preceding years (1997 to 2007) and had no data gaps between the active
period and the reference year (1996 to 2007).
Figure D1
1.2
1.0
0.8
0.6
0.4
0.2
FTE
70
Lower FTE Threshold
6
7
200
200
4
200
5
200
200
3
1
200
2
200
9
Active FTE Threshold
200
0
8
199
199
199
7
199
6
4
5
199
3
199
199
2
199
199
1
0
199
198
9
0.0
Appendix D: Estimating the Number of Older Physicians Who Were Minimally Active
Figure D2 displays an example of a physician’s FTE records, using 2007
as the reference year for analysis. In this situation, the physician would be
included in the analysis and considered retired. This graph highlights the fact
that this physician has two time periods of activity above the active cut-off
(1991 to 1994 and 1999 to 2003). The most recent active time period (circled)
is used in calculating the average active FTE value. This most recent average
active FTE value is used when setting the lower benchmark (15%, 25% or 33%).
Figure D2
1.2
1.0
0.8
0.6
0.4
0.2
Active FTE Threshold
7
200
6
5
200
200
4
200
2
3
200
200
0
1
200
9
200
199
8
7
199
199
6
5
199
4
199
199
3
2
FTE
199
199
1
0
199
199
198
9
0.0
Lower FTE Threshold
71
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Figure D3 displays an example of where a physician would not be included
in the analysis because of data gaps, using 2007 as the reference year for
analysis. Although he or she met the minimum active period and retired
period, there were data gaps (in circle) between the two periods. Any physician
without data in consecutive years between the active period and the reference
year was removed from the analysis.
Figure D3
1.2
1.0
0.8
0.6
0.4
0.2
72
Active FTE Threshold
7
200
6
5
Lower FTE Threshold
200
200
4
3
2
200
200
200
0
1
200
200
9
8
199
7
199
199
6
5
199
199
4
199
3
2
FTE
199
199
0
1
199
199
198
9
0.0
Appendix E: Data for Chapter 6
Appendix E: Data for Chapter 6
Table E1: Average Number of Fee-for-Service Services Provided by Physicians,
by Sex, Specialty and Age Group, Canada, 2006
Family
Medicine
Male
Family
Medicine
Female
Medical
Specialty
Male
Medical
Specialty
Female
Surgical
Specialty
Male
Surgical
Specialty
Female
<40
5,231
3,313
3,532
2,088
4,225
3,583
40–44
6,670
4,111
4,455
2,526
5,357
4,612
45–49
6,867
4,601
4,471
2,769
5,444
5,559
50–54
7,044
4,663
4,876
2,842
5,718
4,676
55–59
7,212
5,197
4,825
2,799
5,409
4,992
60–64
7,091
5,276
4,387
2,962
4,620
5,330
65–69
5,794
4,338
3,953
2,887
3,763
4,445
70–74
4,735
2,763
3,229
2,104
2,875
2,085
75–79
3,530
3,919
2,462
1,432
2,139
1,663
80+
2,156
1,358
2,116
2,113
2,074
3,254
Average
Number of
Services
6,400
4,122
4,214
2,512
4,696
4,347
Age Group
Source
National Physician Database, Canadian Institute for Health Information.
Table E2: Average Full-Time Equivalent Values for Physicians Who Billed Fee
for Service, by Sex, Specialty and Age Group, Canada, 2006
Family
Medicine
Male
Family
Medicine
Female
Medical
Specialty
Male
Medical
Specialty
Female
Surgical
Specialty
Male
Surgical
Specialty
Female
<40
0.75
0.55
0.79
0.56
0.86
0.67
40–44
0.86
0.63
0.90
0.66
1.03
0.81
45–49
0.89
0.67
0.93
0.72
1.00
0.84
50–54
0.90
0.67
0.96
0.70
1.01
0.73
55–59
0.90
0.70
0.95
0.70
0.94
0.78
60–64
0.88
0.70
0.88
0.73
0.82
0.83
65–69
0.76
0.60
0.81
0.70
0.62
0.65
70–74
0.65
0.44
0.69
0.67
0.46
0.26
75–79
0.52
0.52
0.59
0.47
0.32
0.32
80+
0.35
0.21
0.48
0.52
0.26
0.38
Average
FTE Value
0.83
0.62
0.86
0.65
0.86
0.74
Age Group
Source
National Physician Database, Canadian Institute for Health Information.
73
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
Table E3: Number of Active Physicians Who Billed Fee for Service,
Six Age–Specialty Cohorts and Corresponding Average
Full-Time Equivalent Values, Canada, 1989 to 2006
1989
1991
1993
1995
1997
1999
2001
2003
2005
2006
Age–Specialty
Cohort
Internal
Medicine,
50–54 (Male)
N
FTE
483
0.94
463
0.99
456
0.96
445
0.97
423
1.00
372
0.96
346
0.89
322
0.82
295
0.78
276
0.73
Internal
Medicine,
55–59 (Male)
N
FTE
391
0.90
381
0.91
366
0.88
343
0.87
305
0.87
230
0.80
203
0.71
176
0.65
150
0.63
136
0.61
General
Surgery,
50–54 (Male)
N
FTE
267
0.93
258
0.95
243
0.95
230
0.96
223
0.92
193
0.87
174
0.79
145
0.69
137
0.60
127
0.57
General
Surgery,
55–59 (Male)
N
FTE
276
0.85
265
0.86
247
0.84
231
0.79
193
0.78
148
0.65
110
0.59
90
0.48
70
0.45
68
0.45
Ob/Gyn,
50–54 (Male)
N
FTE
182
1.00
176
1.01
170
0.98
161
0.94
147
0.93
127
0.82
115
0.73
102
0.67
88
0.61
85
0.54
Ob/Gyn,
55–59 (Male)
N
FTE
159
0.93
150
0.93
147
0.88
133
0.84
111
0.77
83
0.66
69
0.54
58
0.42
44
0.43
41
0.38
Source
National Physician Database, Canadian Institute for Health Information.
74
Appendix E: Data for Chapter 6
2003
2005
2006
1999
1997
1995
1993
1,527 1,456 1,391 1,326 1,243 1,057
2001
Number of FFS
Billings, FPs/GPs
1991
1989
Table E4: Number of Active Male Family Physicians/General Practitioners
Age 50 to 54 in 1989 Who Billed Fee for Service, Average Full-Time
Equivalent Values and Percentage Involved in Eight Types of Clinical
Service, Canada, 1989 to 2006
972
915
841
785
Average FTE Value
1.01
1.02
0.99
0.98
0.97
0.95
0.90
0.83
0.78
0.73
Providing Office
Assessments (%)
58.3
57.1
54.7
52.3
60.9
53.8
53.9
50.6
49.0
59.1
Providing Basic
Procedural
Skills (%)
88.8
88.5
87.1
87.0
84.3
82.8
81.5
79.2
76.8
75.3
Providing
Advanced Procedural Skills (%)
76.1
75.3
72.6
70.5
69.4
67.5
63.9
60.1
56.1
52.2
Providing
Anesthesia (%)
12.5
10.0
9.2
7.5
6.9
7.6
9.2
6.6
7.3
5.2
Providing Surgical
Assistance (%)
47.0
44.4
41.0
38.1
36.0
31.5
28.7
21.4
18.8
18.2
Providing
Obstetrical
Services (%)
31.6
28.3
23.7
18.6
14.6
11.4
8.5
6.3
5.4
4.2
Providing
Mental Health
Services (%)
81.0
81.5
80.8
81.4
81.0
79.2
77.1
75.2
72.8
74.9
Providing Inpatient
Services (%)
75.4
74.7
71.6
69.5
66.5
64.6
60.1
52.6
49.3
42.9
Source
National Physician Database, Canadian Institute for Health Information.
75
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
1989
1991
1993
1995
1997
1999
2001
2003
2005
2006
Table E5: Number of Active Male Family Physicians/General Practitioners Age 55
to 59 in 1989 Who Billed Fee for Service, Average Full-Time Equivalent
Values and Percentage Involved in Eight Types of Clinical Service,
Canada, 1989 to 2006
1,388
1,312
1,253
1,163
983
795
678
606
509
470
Average FTE
Value
0.96
0.96
0.90
0.88
0.84
0.78
0.72
0.66
0.63
0.59
Providing Office
Assessments (%)
59.1
55.7
53.6
50.6
56.3
48.2
46.5
44.7
45.0
53.4
Providing Basic
Procedural
Skills (%)
86.0
87.3
85.2
83.3
81.3
77.7
74.6
74.1
70.7
69.1
Providing
Advanced Procedural Skills (%)
74.9
74.8
72.2
68.3
65.1
60.5
56.8
52.5
51.5
49.6
Providing
Anesthesia (%)
11.2
9.7
7.6
7.8
7.5
6.9
6.0
5.1
3.7
3.6
Providing
Surgical
Assistance (%)
45.0
41.2
37.1
34.9
33.9
32.6
29.5
23.8
20.8
21.1
Providing
Obstetrical
Services (%)
28.2
24.3
18.4
14.5
11.1
9.1
6.0
4.0
2.8
2.6
Providing
Mental Health
Services (%)
78.8
79.3
77.9
78.2
77.4
73.6
74.5
70.3
69.7
70.4
Providing
Inpatient
Services (%)
73.9
72.9
69.6
66.9
62.9
56.4
49.9
41.6
37.5
35.3
Number of FFS
Billings, FPs/GPs
Source
National Physician Database, Canadian Institute for Health Information.
76
Appendix E: Data for Chapter 6
1993
1995
1997
1999
2001
2003
2006
1991
2005
1989
Table E6: Number of Active Female Family Physicians/General Practitioners
Age 50 to 54 in 1989 Who Billed Fee for Service, Average Full-Time
Equivalent Values and Percentage Involved in Eight Types of Clinical
Service, Canada, 1989 to 2006
Number of FFS
Billings, FPs/GPs
201
187
180
168
154
125
117
111
96
88
Average FTE
Value
0.74
0.80
0.78
0.78
0.79
0.75
0.69
0.61
0.59
0.57
Providing Office
Assessments (%)
48.8
48.1
45.6
44.6
46.1
39.2
35.0
37.8
37.5
46.6
Providing Basic
Procedural
Skills (%)
71.6
75.4
74.4
72.0
69.5
69.6
68.4
57.7
62.5
56.8
Providing
Advanced Procedural Skills (%)
51.7
54.5
55.6
46.2
45.5
43.2
35.9
33.3
32.3
34.1
Providing
Anesthesia (%)
14.9
12.8
9.4
9.5
11.0
7.2
8.5
8.1
8.3
5.7
Providing
Surgical
Assistance (%)
33.3
31.6
31.1
27.4
27.2
27.2
2.05
19.8
19.8
22.7
Providing
Obstetrical
Services (%)
16.9
15.5
11.7
12.5
15.6
8.8
8.5
7.2
4.2
5.7
Providing
Mental Health
Services (%)
80.1
82.9
83.3
82.1
87.0
84.0
84.8
76.6
81.2
78.4
Providing
Inpatient
Services (%)
59.2
61.0
57.8
54.8
54.5
52.0
47.0
36.9
26.0
30.6
Source
National Physician Database, Canadian Institute for Health Information.
77
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
78
2006
1997
2005
1995
2003
1993
Number of FFS
Billings, FPs/GPs
167
155
148
131
112
86
71
56
45
44
Average FTE
Value
0.73
0.74
0.67
0.67
0.65
0.62
0.55
0.53
0.52
0.45
Providing Office
Assessments
(%)
52.1
51.6
47.3
42.0
50.0
41.9
39.4
35.7
40.0
45.5
Providing
Basic Procedural
Skills (%)
67.7
66.5
65.5
61.8
58.0
59.3
56.3
55.4
55.6
50.0
Providing
Advanced Procedural Skills (%)
56.3
50.3
48.0
42.0
45.5
40.7
33.8
33.9
28.9
27.3
Providing
Anesthesia (%)
7.8
8.4
6.8
4.6
4.5
7.0
2.8
7.1
2.2
4.5
Providing
Surgical
Assistance (%)
34.1
33.5
30.4
29.0
22.3
25.6
21.1
16.1
11.1
11.4
Providing
Obstetrical
Services (%)
22.2
18.7
16.2
13.7
10.7
5.8
5.6
1.8
2.2
2.3
Providing
Mental Health
Services (%)
76.6
74.8
77.7
77.1
78.6
86.0
78.8
75.0
75.5
79.5
Providing
Inpatient
Services (%)
58.1
56.8
53.4
48.9
44.6
45.3
38.0
32.1
33.3
29.5
Source
National Physician Database, Canadian Institute for Health Information.
2001
1991
1999
1989
Table E7: Number of Active Female Family Physicians/General Practitioners
Age 55 to 59 in 1989 Who Billed Fee for Service, Average Full-Time
Equivalent Values and Percentage Involved in Eight Types of Clinical
Service, Canada, 1989 to 2006
References
References
1. B. T. B. Chan, From Perceived Surplus to Perceived Shortage:
What Happened to Canada’s Physician Workforce in the 1990s?
(Ottawa, Ont.: Canadian Institute for Health Information, 2002).
2. L. Tyrrell and D. Dauphinee, Task Force on Physician Supply in Canada
(Ottawa, Ont.: Canadian Medical Forum Task Force on Physician Supply
in Canada, 1999).
3. R. Collier, “Diagnosing the Aging Physician,” CMAJ 178, 9 (2008):
pp. 1121–1123.
4. N. Robb, “Interest in Physician-Buyout Packages Grows as More Doctors
Contemplate Retirement,” CMAJ 156, 6 (1997): pp. 882–888.
5. Royal College of Physicians and Surgeons of Canada, College of Family
Physicians of Canada and Canadian Medical Association, National
Physician Survey: Changing Physician Population Pressuring
Healthcare System (media release) (Ottawa, Ont.: RCPSC/CFPC/CMA,
October 27, 2004), accessed from <http://www.nationalphysiciansurvey.
ca/nps/news/PDF-e/National_Physician_Survey_Press_Release_Oct27.04_
updated%20Nov.pdf>.
6. P. Sullivan, “No Quick Fix for Wait Times,” in MD Pulse 2008 (2008):
pp. 27–30, accessed from <www.nationalphysiciansurvey.ca/nps/
MDPulse08/10-MDPulse08_sec2-waittimes.pdf>.
7. D. Gower, “Measuring the Age of Retirement,” Perspectives (summer 1997):
pp. 11–17.
8. N. Deschênes and L. O. Stone, “The Probability of Reaching the State
of Retirement—A Longitudinal Analysis of Variations Between Men and
Women,” in New Frontiers of Research on Retirement, L. O. Stone, ed.
(Ottawa, Ont.: Statistics Canada, 2006), catalogue no. 75-511-XIE.
9. G. Bowlby, “Defining Retirement,” Perspectives on Labour and Income
8, 2 (2007): pp. 15–19.
10. F. T. Denton and B. G. Spencer, Social and Economic Dimensions of an
Aging Population (SEDAP Research Paper No. 231) (Hamilton, Ont.: Social
and Economic Dimensions of an Aging Population Research Program, 2008).
11. F. T. Denton and B. G. Spencer, “What Is Retirement? A Review and
Assessment of Alternative Concepts and Measures,” Canadian Journal
on Aging 28, 1 (2009): pp. 63–76.
79
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
12. P. Ibbott, D. Kerr and R. Beaujot, “Probing the Future of Mandatory
Retirement in Canada,” Canadian Journal on Aging 25, 2 (2006):
pp. 161–178.
13. D. Duchesne, “Seniors at Work,” Perspectives (May 2002): pp. 5–16.
14. G. Schellenberg and Y. Ostrovsky, “2007 General Social Survey Report—
The Retirement Plans and Expectations of Older Workers,” Canadian
Social Trends 86 (2008): pp. 11–34.
15. J. H. Gillies and L. C. Ross, “Physician Retirement: A Case for Concern in
Canadian Hospitals,” CMAJ 131 (1984): pp. 297–299.
16. M. Watts, K. O’Brien and P. Trattner, “The Impact of Ontario’s End of
Mandatory Retirement Legislation on Privileged Physicians,” Osler Update
(December 14, 2006).
17. B. Bahrami, J. Elder and S. Jacobson, “Change in the US Health Care
System: Effects on Physician Retirement and Implications for Health Care
Managers,” Journal of Health and Human Services Administration
25, 3 (2002): pp. 342–370.
18. D. K. Foot et al., “Demographics and Cardiology, 1950–2050,” Journal of
the American College of Cardiology 35, 4 (2000): pp. 1067–1081.
19. S. A. McDaniel, “Work, Retirement and Women in Later Life,” in Rethinking
Retirement, E. Gee, ed. (Vancouver, B.C.: Gerontology Research Centre,
Simon Fraser University, 1995).
20. N. Chappell et al., Aging in Contemporary Canada (Toronto, Ont.:
Prentice Hall, 2003).
21. World Health Organization, The World Health Report 2006: Working
Together for Health (Geneva, Switzerland: WHO, 2006), accessed
from <http://www.who.int/whr/2006/en/>.
22. N. Donen et al., “Canadian Anesthesia Physician Resources: 1996 and
Beyond,” Canadian Journal of Anesthesia 46, 10 (1999): pp. 962–969.
23. H. Yang, R. Byrick and N. Donen, “Analysis of Anesthesia Physician
Resources: Projected Ontario Deficit in 2005,” Canadian Journal of
Anesthesia 47, 2 (2000): pp. 179–184.
24. A. Gilmore, “Should Mandatory Retirement Rules Apply to Doctors?,”
CMAJ 137 (1987): pp. 229–231.
80
References
25. B. Trent, “Mandatory Retirement: Should Older MDs Be Forced to Retire
to Make Way for the New?,” CMAJ 149, 11 (1993): pp. 1696–1699.
26. Canadian Institute for Health Information, Supply, Distribution and
Migration of Canadian Physicians, 2007 (Ottawa, Ont.: CIHI, 2008).
27. H. Grauer and N. M. Campbell, “The Aging Physician and Retirement,”
Canadian Journal of Psychiatry 28 (1983): pp. 552–554.
28. CMA Centre for Physician Health and Well-Being, The Non-Financial
Aspects of Physician Retirement: Environmental Scan and Literature Review
(Ottawa, Ont.: Canadian Medical Association, 2004).
29. B. Chan, G. M. Anderson and M.-E. Thériault, “Patterns of Practice Among
Older Physicians in Ontario,” CMAJ 159, 9 (1998): pp. 1101–1106.
30. L. Thompson, Analysis of a Survey of Older Physicians (unpublished
report) (Saskatoon, Sask.: College of Physicians and Surgeons of
Saskatchewan, 2004).
31. S. L. Fletcher and D. Schofield, “The Impact of Generational Change
and Retirement on Psychiatry to 2025,” BMC Health Services Research
7 (2007): p. 141, accessed from <http://www.biomedcentral.com/14726963/7/141/pre_pub>.
32. D. J. Schofield and J. R. Beard, “Baby Boomer Doctors and Nurses:
Demographic Change and Transitions to Retirement,” Medical Journal of
Australia 183, 2 (2005): pp. 80–83.
33. S. Martin, “‘Freedom 55’ Closer to Age 65 for Physicians,” CMAJ 163, 11
(2000): p. 1499.
34. S. Maguiness et al., “The Canadian Dermatology Workforce Survey:
Implications for the Future of Canadian Dermatology—Who Will Be Your
Skin Expert?,” Journal of Cutaneous Medicine and Surgery 8, 3 (2004):
pp. 141–147.
35. S. Macadam et al., “The Canadian Plastic Surgery Workforce Survey:
Interpretation and Implications,” Plastic and Reconstructive Surgery 119, 7
(2007): pp. 2299–2306.
36. T. R. Konrad and T. Dall, Physician Retirement Intentions and Trends:
Implications for Supply, poster presented at AcademyHealth in San Diego,
California, on June 8, 2004.
37. M. Sylvain, “Doctor Shortage (and Early Retirement) Over Thanks to the
Stock Market,” The Medical Post 45, 3 (2009): pp. 6, 10.
81
Putting Away the Stethoscope for Good? Toward a New Perspective on Physician Retirement
38. Personal communications from Lynda Buske, Director, Workforce
Research, Canadian Medical Association, May 2, 2008, March 9, 2009
and June 7, 2010.
39. Personal communication from Yvonne Rosehart, Program Lead, Health
Human Resources, Canadian Institute for Health Information,
February 27, 2009.
40. Canadian Institute for Health Information, Profiling Physicians by Payment
Program: A Closer Look at Three Provinces (Ottawa, Ont.: CIHI, 2010).
41. J. Tepper, The Evolving Role of Canada’s Family Physicians, 1992–2001
(Ottawa, Ont.: Canadian Institute for Health Information, 2004).
42. P. Hutten-Czapski, R. Pitblado and S. Slade, “Scope of Family Practice
in Rural and Urban Settings,” Canadian Family Physician 50 (2004):
pp. 1548–1550.
43. R. W. Pong and J. R. Pitblado, Geographic Distribution of Physicians in
Canada: Beyond How Many and Where (Ottawa, Ont.: Canadian Institute
for Health Information, 2005).
44. J. Humphreys et al., “The Influence of Geographic Location on the
Complexity of Rural General Practice Activities,” Medical Journal of
Australia 179 (2003): pp. 416–420.
45. M. Robinson et al., “GP Proceduralists: ‘The Hidden Heart’ of Rural and
Regional Health in Australia,” Rural and Remote Health 10 (2010): p. 1402.
46. D. R. Rittenhouse et al., “No Exit: An Evaluation of Measures of Physician
Attrition,” Health Services Research 39, 5 (2004): pp. 1571–1588.
47. Canadian Institute for Health Information, National Physician Database:
Data Submission Specification Manual (Version 3.1) (Ottawa, Ont.:
CIHI, 2003).
48. J. G. Hall, “The Challenge of Developing Career Pathways for Senior
Academic Pediatricians,” Pediatric Research 57, 6 (2005): pp. 914–919.
49. [Author unknown], “CMA Needs to Step Up Lobby Effort for Pensions
With Gov’t,” The Medical Post (February 20, 2009): p. 10.
82
References
50. [Author unknown], “An Open Letter to Canada’s Health Minister: It’s Time
to Consider the Benefits of Physician Pension Not Only to Doctors, but to
the Health-Care System,” The Medical Post (June 2, 2009): pp. 2, 5.
51. [Author unknown], “Many See Pensions as Good Tradeoff for Smaller Fee
Raises,” The Medical Post (July 21, 2009): p. 12.
52. J. R. Pitblado, Distribution and Internal Migration of Canada’s Health Care
Workforce: Summary Report Update to 2006 (Ottawa, Ont.: Canadian
Institute for Health Information, 2007).
83
Production of this report is made possible by financial contributions from Health
Canada and provincial and territorial governments. The views expressed herein
do not necessarily represent the views of Health Canada or any provincial or
territorial government.
All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by
any means, electronic or mechanical, including photocopying and recording,
or by any information storage and retrieval system now known or to be invented,
without the prior permission in writing from the owner of the copyright, except
by a reviewer who wishes to quote brief passages in connection with a review
written for inclusion in a magazine, newspaper or broadcast.
Requests for permission should be addressed to:
Canadian Institute for Health Information
495 Richmond Road, Suite 600
Ottawa, Ontario K2A 4H6
Phone: 613-241-7860
Fax: 613-241-8120
www.cihi.ca
[email protected]
ISBN 978-1-55465-863-3 (PDF)
© 2011 Canadian Institute for Health Information
How to cite this document:
R. W. Pong, PhD, Putting Away the Stethoscope for Good? Toward a New
Perspective on Physician Retirement (Ottawa, Ont.: Canadian Institute for Health
Information, 2011).
Cette publication est aussi disponible en français sous le titre R. W. Pong, Ph. D.,
Accrocher définitivement son stéthoscope? Vers une nouvelle perspective du départ
à la retraite des médecins.
ISBN 978-1-55465-864-0 (PDF)
Talk to Us
CIHI Ottawa
495 Richmond Road, Suite 600
Ottawa, Ontario K2A 4H6
Phone: 613-241-7860
CIHI Montréal
1010 Sherbrooke Street West, Suite 300
Montréal, Quebec H3A 2R7
Phone: 514-842-2226
CIHI Toronto
4110 Yonge Street, Suite 300
Toronto, Ontario M2P 2B7
Phone: 416-481-2002
CIHI St. John’s
140 Water Street, Suite 701
St. John’s, Newfoundland and Labrador A1C 6H6
Phone: 709-576-7006
CIHI Victoria
880 Douglas Street, Suite 600
Victoria, British Columbia V8W 2B7
Phone: 250-220-4100
March 2011
www.cihi.ca
Was this manual useful for you? yes no
Thank you for your participation!

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

Download PDF

advertisement