OBJECTIVE

We assessed the impact of a diabetes incentive code introduced for primary care physicians in Ontario, Canada, in 2002 on quality of diabetes care at the population and patient level.

RESEARCH DESIGN AND METHODS

We analyzed administrative data for 757,928 Ontarians with diabetes to examine the use of the code and receipt of three evidence-based monitoring tests from 2006 to 2008. We assessed testing rates over time and before and after billing of the incentive code.

RESULTS

One-quarter of Ontarians with diabetes had an incentive code billed by their physician. The proportion receiving the optimal number of all three monitoring tests (HbA1c, cholesterol, and eye tests) rose gradually from 16% in 2000 to 27% in 2008. Individuals who were younger, lived in rural areas, were not enrolled in a primary care model, or had a mental illness were less likely to receive all three recommended tests. Patients with higher numbers of incentive code billings in 2006–2008 were more likely to receive recommended testing but also were more likely to have received the highest level of recommended testing prior to introduction of the incentive code. Following the same patients over time, improvement in recommended testing was no greater after billing of the first incentive code than before.

CONCLUSIONS

The diabetes incentive code led to minimal improvement in quality of diabetes care at the population and patient level. Our findings suggest that physicians who provide the highest quality care prior to incentives may be those most likely to claim incentive payments.

Diabetes accounts for an increasing proportion of the global burden of disease and currently is the fifth or sixth most common cause of death in most developed countries (1). It is well established that appropriate monitoring and treatment can significantly reduce the incidence of diabetes complications and improve overall morbidity and mortality (26). However, numerous studies, both globally and in Canada, have shown that the quality of diabetes care, measured by adherence to recommended processes or attainment of treatment goals, consistently falls short of evidence-based guidelines (7).

Over the last decade, many countries have implemented pay-for-performance programs in an effort to improve the quality of health care, but there still is limited evidence to support the effectiveness of this approach (8,9). In 2002, the government in Ontario, Canada, introduced a new fee code for primary care physicians to encourage regular, comprehensive management of diabetic patients (10). When introduced, this code could be billed a maximum of three times a year per patient at a value of $37.00 (Canadian) per visit and required maintenance of a diabetes flow sheet that tracked cholesterol, hemoglobin A1c (HbA1c), retinal eye examination, blood pressure, weight, and other parameters relevant to diabetes management (11). It is unclear, however, whether the new incentive code has had any impact on the quality of care provided.

In this study, we aimed to investigate the quality of diabetes care, measured by receipt of three evidence-based monitoring tests, and to assess the impact of the new diabetes incentive code in Ontario on quality of care at the population and patient level. We also sought to identify patient and physician characteristics associated with higher quality care.

We used available administrative claims data to examine the use of the diabetes incentive code and assess receipt of evidence-based monitoring tests among individuals with diabetes in Ontario. Data were accessed through a comprehensive research agreement with the Ontario Ministry of Health and Long-Term Care. Prior to data analysis, all patient and provider identifiers were removed and replaced with unique encrypted numbers. This study was approved by the research ethics board of Sunnybrook Health Sciences Centre in Toronto, Ontario.

Outcome measures

Diabetes incentive code.

We assessed use of the Diabetes Management Assessment fee code using physician service claims to the Ontario Health Insurance Plan (OHIP). OHIP coverage is extended to all permanent residents of Ontario. When the code was introduced in 2002, it could be billed a maximum of three times a year per patient at a value of $37.00 (Canadian) per visit and required maintenance of a diabetes flow sheet that tracked key parameters such as cholesterol, eye exams, and HbA1c.

Quality of diabetes care.

We assessed the receipt of three evidence-based monitoring tests for diabetes care using data from OHIP service codes: frequency of retinal eye examination (using optometry and ophthalmology claims to OHIP), frequency of HbA1c measurement, and frequency of cholesterol measurement. We defined the optimal screening frequency during the 2-year study period as follows based on recommendations from the Canadian Diabetes Association 2003 Clinical Practice Guidelines (12): one retinal eye exam, four HbA1c tests, and two cholesterol tests.

Study design and population

Patient and physician characteristics associated with quality of diabetes care.

We conducted a cross-sectional analysis, from 1 April 2006 to 31 March 2008, to assess patient and physician characteristics associated with use of the diabetes incentive code and quality of diabetes care. We identified Ontarians aged ≥40 years using the Ontario Diabetes Database, a registry of Ontario patients diagnosed with diabetes generated using a validated administrative data algorithm (13). Patients were included if they were diagnosed with diabetes on or prior to 31 August 2006. Patients were excluded if they resided in a long-term care facility, registered with the OHIP after 31 March 2006, or died before 31 March 2008. We included primary care physicians in active practice in August 2008.

Impact of the diabetes incentive code at the population level.

We examined trends in quality of diabetes care over time in Ontario. We stratified people with diabetes based on the number of diabetes incentive codes they received between 2006 and 2008. These cohorts of patients were then followed from 1998 to 2008 to examine differences in receipt of evidence-based testing over time. Our primary analysis included all people diagnosed with diabetes on or prior to 31 August 2006. We assessed testing for each 2-year period from 1998 to 2008 and included individuals in the denominator only if they had been diagnosed with diabetes at the start of the 2-year period. We also conducted a secondary analysis that included only those who had diabetes on or before 1 September 1998.

Impact of the diabetes incentive code at the patient level.

We created a cohort of people with diabetes who had at least one diabetes incentive code billed between 1 January 2002 and 31 August 2006, who had diabetes for at least 6 years prior to billing of the first code, and who were alive 2 years after billing of the first code. We stratified this cohort into two groups: those with one to two incentive codes billed and those with three or more codes billed from 2002 to 2006. For this cohort, we assessed the quality of diabetes care in the 2-year period after billing of the first diabetes incentive code, the 2-year period immediately before, the period 2–4 years before, and the period 4–6 years before. We compared the quality of care in each 2-year time period to the quality of care in the 2 years immediately prior.

Patient and physician characteristics

Patient age, sex, and place of residence were obtained from the Registered Persons Database, the registry for all people covered by OHIP. Coverage is extended to all permanent residents of Ontario. We derived neighborhood income quintiles by linking 2006 census data to the patients’ residential postal code. Rurality was assessed using the Rurality Index of Ontario (14). We used OHIP registration after 31 August 1998 as a proxy for recent immigration. The prevalence of mental health problems was assessed using a validated algorithm based on ambulatory physician visits (15). Mental health was included as a variable because of the relatively high rates of diabetes and related complications in this population and known barriers to care (16). Patient comorbidity was determined using Aggregated Diagnosis Groups (ADGs) from the Johns Hopkins Adjusted Clinical Groups (ACG) Case-Mix System (17) (scores included diabetes as a comorbidity). Physician billing claims to OHIP linked with the hospital Discharge Abstract Database from the Canadian Institute for Health Information were used to identify patient diagnoses.

We used the Corporate Provider Database, current to 31 August 2008, to identify demographic data on practicing physicians in Ontario and the Institute for Clinical Evaluative Sciences Physician Database, current to the 2007 fiscal year, for country of graduation. Client Agency Program Enrollment tables as of 31 August 2008 were used to identify patients formally enrolled to a primary care physician practicing in a patient enrollment model. These models consisted of enhanced fee for service, nonteam capitation, and team-based capitation. A virtual roster method was used to assign people with diabetes who were not enrolled in a model to a primary care physician based on the maximum value of 18 common primary care fee codes.

Statistical analysis

We used multivariable generalized estimating equation (GEE) models to examine the association between patient and physician characteristics and the number of incentive codes billed as well as whether patients received recommended testing using Poisson and logistic regression, respectively. GEE models are similar to typical regression models in their interpretation; however, they adjust for the correlation in outcomes that are observed between patients who see the same physician and of physicians in the same enrollment group (18). Independent variables for these models at the patient level included age, sex, neighborhood income quintile, recent registrant status, rural residence, mental illness, comorbidity, duration of diabetes, region, and enrollment model. We ran these models with and without two variables (whether the patient was seen by an endocrinologist or general internist and the number of primary care physician visits) so that the independent effects of these two variables could be examined. At the physician level, the models included primary care physician age, sex, years since graduation, and Canadian graduate status. We excluded patients from regression models if we could not assign them to a primary care physician (n = 22,954 [3.0%]).

We also used GEE models when analyzing data for the smaller cohort of Ontarians with diabetes who had at least one incentive code billed. We used GEE models to compare the likelihood of a patient receiving recommended testing in a given 2-year period compared with the 2-year period immediately prior (e.g., comparing the 2-year period after the billing of the first incentive code with the 2-year period before billing of the code).

We analyzed data for 757,928 people with diabetes representing ~12% of Ontario’s population aged ≥40 years. When compared with the general population, people with diabetes were more likely to be older, male, low income, and long-term residents of Ontario. They also were more likely to reside in an urban area, have a psychotic mental illness, have more visits to primary care physicians, and be enrolled to a primary care physician. Approximately 42% saw an endocrinologist or general internist during the 2-year study period. Physicians looking after diabetic patients were more likely to be an international medical graduate, older, male, and in practice for >30 years.

Use of the diabetes incentive code

Only 25% of people with diabetes had a diabetes incentive code billed during the 2-year study period, with only 8% receiving at least four of the maximum six codes permitted during the timeframe. After controlling for patient- and physician-level characteristics, individuals enrolled to primary care physicians practicing in a team or nonteam capitation model were 25% more likely to receive an incentive code than patients enrolled to an enhanced fee-for-service model (Table 1).

Table 1

Patient- and physician-level characteristics associated with receipt of the diabetes incentive code

Patient- and physician-level characteristics associated with receipt of the diabetes incentive code
Patient- and physician-level characteristics associated with receipt of the diabetes incentive code

Quality of diabetes care

Overall, 67% of people with diabetes received a retinal eye exam during the 2-year study period. Seventy-eight percent received at least one HbA1c test and 37% received the optimal number of HbA1c tests (four tests). For cholesterol testing, 80% received at least one test and 59% received the optimal number of cholesterol tests (two tests). Only 27% of individuals received the optimal number of all three monitoring tests.

After adjusting for patient and physician characteristics, people with diabetes aged 65–79 years were more likely to receive optimal testing than their younger counterparts (Table 2). The greatest difference was in retinal eye exams where those aged 65–79 years were just over two and a half times more likely to receive the test than those aged 40–64 years. People with diabetes who did not live in a rural area, those without a mental illness, and those seeing a diabetes specialist over the 2-year period were more likely to receive each of the three monitoring tests. Individuals enrolled to a team or nonteam capitation model were slightly more likely to receive optimal monitoring than those enrolled to the enhanced fee-for-service model, whereas those not enrolled to any model were less likely to receive any of the three monitoring tests.

Table 2

Multivariable regression analysis of the association between patient- and physician-level characteristics and receipt of recommended testing for diabetes

Multivariable regression analysis of the association between patient- and physician-level characteristics and receipt of recommended testing for diabetes
Multivariable regression analysis of the association between patient- and physician-level characteristics and receipt of recommended testing for diabetes

When we adjusted for patient and physician characteristics, we found that patients who received one to two diabetes incentive code billings were more likely to receive the optimal number of all three monitoring tests (odds ratio 2.23 [95% CI 2.15–2.31]); those who received three or more code billings were even more likely to receive optimal monitoring (5.79 [5.49–6.11]).

Impact of the diabetes incentive code at the population level

From 2000 to 2008, the proportion of people with diabetes receiving optimal monitoring rose gradually from 16 to 27% (Fig. 1). Those who had diabetes incentive codes billed between 2006 and 2008 were historically more likely to receive recommended tests relative to those who had fewer or no incentive codes billed. For example, those who had one diabetes incentive code billed between 2006 and 2008 were more likely to receive recommended testing than those with no incentive codes billed but less likely to receive recommended testing than those with two incentive codes billed, and this relationship held true going back to 2000, prior to introduction of the incentive code. Recommended testing was increasing in the time period prior to implementation of the incentive code at a rate similar to the increase afterward. These relationships also were true when the analysis was limited to those who had diabetes diagnosed on or before 1 September 1998 (results not shown). The pattern for eye exams showed a decrease in all groups starting in 2005, remaining at a low level after that (results not shown). HbA1c and cholesterol testing followed a pattern similar to that for all three tests combined (results not shown).

Figure 1

Population-level analysis. Proportion of Ontarians with diabetes receiving recommended testing from 1998 to 2008 stratified by number of diabetes incentive codes billed from 2006 to 2008 (n = 757,928). Recommended testing was defined as receipt of one retinal eye examination, four HbA1c tests, and two cholesterol tests in the previous 2 years. Therefore, data for 2000 represents testing done from 1998 to 2000. ♦ and dotted line, six or more incentive codes billed from 2006 to 2008 (n = 17,223); ◇ and dashed line, five incentive codes (n = 19,064); ▲ and dash-dot line, four incentive codes (n = 24,285); △ long dash line, three incentive codes (n = 30,312); ● and dotted line, two incentive codes (n = 39,518); ○ and dashed line, one incentive code (n = 63,071); ■ and solid line, overall (n = 757,928); □ and dash-dot line, no incentive codes (n = 564,455).

Figure 1

Population-level analysis. Proportion of Ontarians with diabetes receiving recommended testing from 1998 to 2008 stratified by number of diabetes incentive codes billed from 2006 to 2008 (n = 757,928). Recommended testing was defined as receipt of one retinal eye examination, four HbA1c tests, and two cholesterol tests in the previous 2 years. Therefore, data for 2000 represents testing done from 1998 to 2000. ♦ and dotted line, six or more incentive codes billed from 2006 to 2008 (n = 17,223); ◇ and dashed line, five incentive codes (n = 19,064); ▲ and dash-dot line, four incentive codes (n = 24,285); △ long dash line, three incentive codes (n = 30,312); ● and dotted line, two incentive codes (n = 39,518); ○ and dashed line, one incentive code (n = 63,071); ■ and solid line, overall (n = 757,928); □ and dash-dot line, no incentive codes (n = 564,455).

Impact of the diabetes incentive code at the patient level

We analyzed data for a cohort of 58,927 individuals who had at least one diabetes incentive code billed between 2002 and 2006, had diabetes for at least 6 years prior to the billing of the first code, and were alive 2 years after billing of the first code. In that cohort, 38,127 individuals had one to two codes billed and 20,800 had three or more codes billed. Demographics and comorbidities were similar in both groups. Figure 2 shows that overall there was improvement in the proportion of patients receiving recommended testing in the 2-year period after the first incentive code was billed compared with the previous 2 years (relative risk 1.22 [95% CI 1.21–1.23]) but that this improvement was not larger than in the time periods before the code was billed (1.31 [1.30–1.32] for the 2-year period before the code was billed compared with the 2-year period prior; 1.33 [1.33–1.34] for 2–4 years before the code was billed compared with the 2-year period prior). This observation was true both for individuals who received one to two codes and those who received three or more codes.

Figure 2

Patient-level analysis. Proportion of people with diabetes receiving recommended testing before and after billing of the first diabetes incentive code (n = 33,926). ■ and solid line, overall; ○ and dotted line, patients receiving three or more incentive codes from 2002 to 2006; ● and dashed line, patients receiving one or two incentive codes from 2002 to 2006.

Figure 2

Patient-level analysis. Proportion of people with diabetes receiving recommended testing before and after billing of the first diabetes incentive code (n = 33,926). ■ and solid line, overall; ○ and dotted line, patients receiving three or more incentive codes from 2002 to 2006; ● and dashed line, patients receiving one or two incentive codes from 2002 to 2006.

Our findings demonstrate limited impact of a diabetes incentive code introduced to all primary care physicians in Ontario in 2002. Only one-quarter of people with diabetes in Ontario had an incentive code billed between 2006 and 2008. At the population level, the percentage of Ontarians with diabetes receiving three evidence-based monitoring tests at the recommended intervals (retinal eye examination, HbA1c measurement, and cholesterol measurement) rose gradually, reaching just over one-quarter of the population by 2008. The amount of improvement was similar in the preincentive and postincentive time periods. At the patient level, the amount of improvement in recommended testing was no greater after billing of the first incentive code than it was before.

Our findings suggest that physicians who provide the best care prior to incentives may be those most likely to claim incentive payments. In a cross-sectional analysis, billing of the incentive code and the number of times it was billed was associated with greater receipt of monitoring tests for diabetes. However, those with the greatest number of codes billed and highest quality of care between 2006 and 2008 also had the highest quality of care prior to when the codes were introduced. Others also have found that financial incentives may largely reward those with higher performance at baseline (19).

Low uptake of the diabetes incentive code in Ontario is likely attributed to several factors, including its relatively low financial value, lack of physician awareness about the code, the introduction of competing incentive codes during the same time period, and the added administrative burden of completing a diabetes flow sheet, particularly for paper-based practices. We found that primary care practices reimbursed via blended capitation were more likely to use the diabetes incentive code than fee-for-service practices. A possible reason is that the diabetes code is one of the few codes paid in full in capitation models versus other codes that are paid at 10% of their value, making the incremental value of the code much greater in capitation than in fee for service ($33.77 Canadian vs. $4.65 Canadian). Research from Australia suggests that administrative paperwork is a major barrier to uptake of diabetes incentives by general practitioners, whereas having a practice nurse and practice computerization facilitates uptake (20).

Only 27% of people with diabetes in Ontario received recommended monitoring. This proportion is far below the target of 80% announced by the Ontario Diabetes Strategy in November 2009. The poor attainment of diabetes process measures in our study is consistent with findings from other published literature examining the Canadian population (2123). Of note, we found a large drop in eye exams in 2005, just shortly after routine eye examinations for those aged 20–64 years were delisted from Ontario’s public insurance plan even though exams for people with diabetes were not delisted. This effect persisted over time. Further investigation of this issue is required to understand the degree to which eye exams are being foregone versus taking place but being paid through private insurance or out of pocket.

Some disparities were found in diabetes care, with those who were younger, living in rural areas, or having mental health disorders less likely to receive each of the three recommended tests, even after controlling for other factors. Patients in the lowest income quintile were less likely to receive eye examinations and cholesterol testing, but we found no disparities in care for new immigrants. As in our study, others also have found that patients with mental illness are less likely to receive recommended diabetes care (2426). Similar to us, Brown et al. (27) found lower rates of dilated eye exams among patients of lower socioeconomic position in the U.S., whereas Khunti et al. (28) found lower compliance with most process measures for practices serving more socioeconomically deprived populations in England. Dallo et al. (29) did not find differences in self-reported receipt of eye examinations, HbA1c testing, or cholesterol testing between foreign-born and U.S.-born individuals with diabetes after controlling for confounders such as age, sex, and socioeconomic status.

Pay-for-performance schemes have become a popular method for improving the quality of care, but their effectiveness still is unclear (8,9,30). The best evidence in primary care comes from the U.K. where the government introduced a comprehensive financial incentive scheme for general practitioners in 2004 that in 2006 included 16 quality indicators specific to diabetes care that assessed both processes and outcomes of care (31). Studies suggest that the introduction of the incentive scheme accelerated the rate of improvement in diabetes care but that the rate has since slowed (32) and that the scheme has reduced the quality gap in primary health care between practices serving patients with low versus high socioeconomic status (33). The financial incentives in the U.K. differ from those in Ontario markedly in that they constitute a significant portion of general practitioner income (estimated at 25%), virtually all general practitioners participate, and by the third year, general practitioners on average achieved >95% of the clinical targets including >95% of the diabetes targets (34).There also are important differences in context. For example, in the U.K., almost all general practitioners use electronic medical records that enable creation of disease registries and tracking of performance measures, and most were on a capitation model prior to the scheme’s introduction. In contrast, ~30% of Ontario primary care physicians had electronic records in 2009.

Financial incentives for general practitioners introduced in Australia in 2001 are more comparable to those in Ontario. These included an incentive specific to diabetes that pays $40 (AUS) per patient per year for a completed annual cycle of care that includes parameters similar to those in Ontario’s diabetes flow sheet (35). Incentive payments for diabetes and other parameters comprise ~10% of general practitioners’ income and most general practitioners were paid by fee for service prior to the introduction of the incentives. As in Ontario, there was fairly low uptake of the incentive, with only 42% of eligible general practitioners in one region claiming the incentive in 2007 (20). Similar to Ontario, there is evidence that general practitioners who participated in the incentive scheme were more likely to order an HbA1c test for a diabetic patient (36) and more likely to have a rigorous approach to diabetes management (20). However, it is unknown whether these differences in care preceded introduction of the incentive scheme.

Our study has limitations. First and most important was our dependence on administrative data. Adherence to recommended testing was measured using billing data for Ontario’s public insurance system, which may underestimate services actually provided. Specifically, we could not include retinal eye exams paid for privately or laboratory tests done in hospitals. Second, we were only able to measure processes of care for diabetes, which may not translate to outcomes that are important to patients, providers, and payers, such as better blood glucose, cholesterol and blood pressure control, and, ultimately, reduced long-term consequences, such as hospital admissions, cardiovascular disease, and death. Third, our observational study limited our ability to distinguish between association and causation. As a result, reverse causation is a plausible explanation for the associations we found (i.e., physicians who historically provided good care were more likely to use the diabetes incentive code).

In summary, our study found limited impact of a financial incentive code for diabetes 6 years after its introduction in Ontario. At the population level, the proportion of Ontarians with diabetes receiving recommended testing rose gradually after the incentive was introduced but was rising at a similar rate beforehand and remained far below provincial targets. At the patient level, the amount of improvement in recommended testing was no greater after billing of the first incentive code than it was before. Our findings suggest that physicians who already were providing relatively good diabetes care were more likely to bill the incentive code. Ontarians with diabetes who were younger, lived in rural areas, had mental illness, or were not enrolled to a primary care model were less likely to receive recommended testing. Additional research is needed to understand whether and how financial incentives change physician behavior and how incentives influence patient outcomes for diabetes and other chronic diseases. Policy development work and research is also needed to understand and address gaps in diabetes care, particularly for those least likely to receive recommended care.

The opinions, results, and conclusions reported in this article are those of the authors and are independent from the funding sources. No endorsement by Institute for Clinical Evaluative Sciences or the Ontario Ministry of Health and Long-Term Care is intended or should be inferred.

This study was supported by the Institute for Clinical Evaluative Sciences, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. T.K. is supported by a Canadian Institutes of Health Research Strategic Training Fellowship in the ACHIEVE Research Partnership and a New Investigator Award from the Department of Family and Community Medicine, University of Toronto.

No potential conflicts of interest relevant to this article were reported.

T.K. designed the study, conducted the analysis, interpreted the results, and drafted and critically reviewed the manuscript. J.C.V. conducted the analysis, interpreted the results, and critically reviewed the manuscript. A.K. conducted the analysis and critically reviewed the manuscript. B.R.S. conceived and designed the study and critically reviewed the manuscript. R.H.G. conceived and designed the study, conducted the analysis, interpreted the results, and drafted and critically reviewed the manuscript. T.K. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

1.
International Diabetes Federation. IDF Diabetes Atlas, Fourth Edition, 2009. Available from www.eatlas.idf.org. Accessed 9 September 2009
2.
American Diabetes Association
.
Implications of the UK Prospective Diabetes Study
.
Diabetes Care
2002
;
25
:
28
32
3.
Gaede
P
,
Vedel
P
,
Larsen
N
,
Jensen
GV
,
Parving
HH
,
Pedersen
O
.
Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes
.
N Engl J Med
2003
;
348
:
383
393
[PubMed]
4.
Holman
RR
,
Paul
SK
,
Bethel
MA
,
Matthews
DR
,
Neil
HA
.
10-year follow-up of intensive glucose control in type 2 diabetes
.
N Engl J Med
2008
;
359
:
1577
1589
[PubMed]
5.
Pyŏrälä
K
,
Pedersen
TR
,
Kjekshus
J
,
Faergeman
O
,
Olsson
AG
,
Thorgeirsson
G
.
Cholesterol lowering with simvastatin improves prognosis of diabetic patients with coronary heart disease: a subgroup analysis of the Scandinavian Simvastatin Survival Study (4S)
.
Diabetes Care
1997
;
20
:
614
620
[PubMed]
6.
UK Prospective Diabetes Study Group
.
Tight blood pressure control and risk of macrovascular and microvascular complications in type 2†diabetes: UKPDS 38
.
BMJ
1998
;
317
:
703
713
7.
Shah
BR
,
Hux
JE
,
Laupacis
A
,
Zinman
B
,
Zwarenstein
M
.
Deficiencies in the quality of diabetes care: comparing specialist with generalist care misses the point
.
J Gen Intern Med
2007
;
22
:
275
279
[PubMed]
8.
Christianson
J
,
Leatherman
S
,
Sutherland
K
.
Financial Incentives, Healthcare Providers and Quality Improvements: A Review of the Evidence
.
London
,
The Health Foundation
,
2007
9.
Rosenthal
MB
,
Frank
RG
.
What is the empirical basis for paying for quality in health care?
Med Care Res Rev
2006
;
63
:
135
157
[PubMed]
10.
Ontario Ministry of Health and Long Term Care
.
Changes to the Schedule of Benefits and Related Changes Effective April 1, 2002
.
Toronto
,
Ontario Ministry of Health and Long Term Care
,
2002
[OHIP bulletin no. 2038]
11.
Schedule of Benefits for Physician Services under the Health Insurance Act
.
Toronto
,
Ontario Ministry of Health and Long-Term Care
,
2009
, p.
A1
A111
12.
Canadian Diabetes Association Clinical Practice Guidelines Expert Committee
.
Canadian Diabetes Association 2003 clinical practice guidelines for the prevention and management of diabetes in Canada
.
Can J Diabetes
2003
;
27
(
Suppl. 2
)
:S22,S60,S77
13.
Hux
JE
,
Ivis
F
,
Flintoft
V
,
Bica
A
.
Diabetes in Ontario: determination of prevalence and incidence using a validated administrative data algorithm
.
Diabetes Care
2002
;
25
:
512
516
[PubMed]
14.
Kralj
B
.
Measuring “Rurality” for Purposes of Health-Care Planning: An Empirical Measure for Ontario
.
Toronto
,
Ontario Medical Association
,
2005
15.
Steele
LS
,
Glazier
RH
,
Lin
E
,
Evans
M
.
Using administrative data to measure ambulatory mental health service provision in primary care
.
Med Care
2004
;
42
:
960
965
[PubMed]
16.
Canadian Mental Health Association Ontario
.
Diabetes and Serious Mental Illness: Future Directions for Ontario
.
Toronto
,
Canadian Mental Health Association
,
2009
.
17.
Johns Hopkins University
.
Johns Hopkins University ACG Case-Mix Adjustment System
,
1997
. Available from http://www.acg.jhsph.org. Accessed 1 February 2010
18.
Hubbard
AE
,
Ahern
J
,
Fleischer
NL
, et al
.
To GEE or not to GEE: comparing population average and mixed models for estimating the associations between neighborhood risk factors and health
.
Epidemiology
2010
;
21
:
467
474
[PubMed]
19.
Rosenthal
MB
,
Frank
RG
,
Li
Z
,
Epstein
AM
.
Early experience with pay-for-performance: from concept to practice
.
JAMA
2005
;
294
:
1788
1793
[PubMed]
20.
Saunders
M
,
Schattner
P
,
Mathews
M
.
Diabetes ‘cycles of care’ in general practice: do government incentives help?
Aust Fam Physician
2008
;
37
:
781
784
[PubMed]
21.
Harris
SB
,
Stewart
M
,
Brown
JB
, et al
.
Type 2 diabetes in family practice: room for improvement
.
Can Fam Physician
2003
;
49
:
778
785
[PubMed]
22.
Russell
GM
,
Dahrouge
S
,
Hogg
W
,
Geneau
R
,
Muldoon
L
,
Tuna
M
.
Managing chronic disease in Ontario primary care: the impact of organizational factors
.
Ann Fam Med
2009
;
7
:
309
318
23.
The Commonwealth Fund
.
Performance Snapshots: International Comparisons: Effectiveness
,
2010
24.
Frayne
SM
,
Halanych
JH
,
Miller
DR
, et al
.
Disparities in diabetes care: impact of mental illness
.
Arch Intern Med
2005
;
165
:
2631
2638
25.
Goldberg
RW
,
Kreyenbuhl
JA
,
Medoff
DR
, et al
.
Quality of diabetes care among adults with serious mental illness
.
Psychiatr Serv
2007
;
58
:
536
543
26.
Jones
LE
,
Clarke
W
,
Carney
CP
.
Receipt of diabetes services by insured adults with and without claims for mental disorders
.
Med Care
2004
;
42
:
1167
1175
[PubMed]
27.
Brown
AF
,
Gregg
EW
,
Stevens
MR
, et al
.
Race, ethnicity, socioeconomic position, and quality of care for adults with diabetes enrolled in managed care: the Translating Research Into Action for Diabetes (TRIAD) study
.
Diabetes Care
2005
;
28
:
2864
2870
[PubMed]
28.
Khunti
K
,
Ganguli
S
,
Baker
R
,
Lowy
A
.
Features of primary care associated with variations in process and outcome of care of people with diabetes
.
Br J Gen Pract
2001
;
51
:
356
360
[PubMed]
29.
Dallo
FJ
,
Wilson
FA
,
Stimpson
JP
.
Quality of diabetes care for immigrants in the U.S
.
Diabetes Care
2009
;
32
:
1459
1463
[PubMed]
30.
Giuffrida
A
,
Gosden
T
,
Forland
F
, et al
.
Target payments in primary care: effects on professional practice and health care outcomes
.
Cochrane Database Syst Rev
2000
;(
3
):
CD000531
[PubMed]
31.
British Medical Association. Revisions to the GMS contract 2006/07: delivering investment in general practice[article online], 2006. Available from http://www.bma.org.uk/employmentandcontracts/independent_contractors/general_medical_services_contract/revisionnGMSFeb20062.jsp. Accessed 1 December 2010
32.
Campbell
SM
,
Reeves
D
,
Kontopantelis
E
,
Sibbald
B
,
Roland
M
.
Effects of pay for performance on the quality of primary care in England
.
N Engl J Med
2009
;
361
:
368
378
33.
Doran
T
,
Fullwood
C
,
Kontopantelis
E
,
Reeves
D
.
Effect of financial incentives on inequalities in the delivery of primary clinical care in England: analysis of clinical activity indicators for the quality and outcomes framework
.
Lancet
2008
;
372
:
728
736
[PubMed]
34.
The Information Centre
.
National Quality and Outcomes Framework Statistics for England 2006/07
,
2007
.
35.
Medicare Australia. Practice incentives program: diabetes inventive guidelines, 2011. Available from http://www.medicareaustralia.gov.au/provider/incentives/pip/forms-guides.jsp. Accessed 27 February 2012
36.
Scott
A
,
Schurer
S
,
Jensen
P
,
Sivey
P
.
The effects of financial incentives on quality of care: the case of diabetes. Health, Econometrics, and Data Group (HEDG) Working Papers. York, U.K., Department of Economics, University of York
,
2008
.