Trends in Excess Mortality Among Patients With ... - Wiley Online Library

Arthritis Care & Research Vol. 67, No. 8, August 2015, pp 1047–1053 DOI 10.1002/acr.22553 C 2015, American College of Rheumatology V

ORIGINAL ARTICLE

Trends in Excess Mortality Among Patients With Rheumatoid Arthritis in Ontario, Canada JESSICA WIDDIFIELD,1 SASHA BERNATSKY,2 J. MICHAEL PATERSON,3 GEORGE TOMLINSON,4 KAREN TU,5 BINDEE KURIYA,6 J. CARTER THORNE,7 JANET E. POPE,8 SIMON HOLLANDS,9 AND CLAIRE BOMBARDIER4

Objective. To evaluate excess mortality over time, comparing rheumatoid arthritis (RA) patients with the general population. Methods. We computed all-cause mortality rates among Ontario residents age ‡15 years with RA versus without RA from 1996 to 2009. Age- and sex-standardized mortality rates were expressed as the number of deaths per 1,000 population. Excess mortality rates were calculated as the difference between death rates among RA patients and those in the general population. We estimated standardized mortality ratios (SMRs) and mortality rate ratios (MRRs) to assess relative excess mortality over time. Results. From 1996 to 2009, SMRs in RA ranged from 13.0 (95% confidence interval [95% CI] 12.2, 13.9) to 9.2 deaths per 1,000 RA patients (95% CI 8.4, 10.0); and for those without RA from 8.7 (95% CI 8.6, 8.7) to 6.0 deaths (95% CI 5.9, 6.0) per 1,000 general population. Over the study period, the excess mortality rate among RA patients was approximately 3 excess deaths per 1,000 population. Relative reductions in standardized mortality rates occurred over time for those with and without RA (221.4% versus 213.4%). The SMRs for RA patients in 1996–1997, 2000– 2001, 2004–2005, and 2008–2009 were 1.51 (95% CI 1.43, 1.59), 1.50 (95% CI 1.43, 1.57), 1.43 (95% CI 1.37, 1.50), and 1.41 (95% CI 1.35, 1.47), respectively. We did not find a significant change in the MRR by calendar time. Conclusion. Mortality for RA patients has decreased over time but remains elevated compared to the general population, with 40–50% more deaths among RA patients. The relative excess mortality over time (mortality gap) remains unchanged in our sample.

The excess mortality in patients with rheumatoid arthritis (RA) has been studied for decades, and continues to be a topic of great interest (1). Despite substantial improvements in RA management over the past decade, several factors (including autoimmune inflammatory burden, genetics, and increased comorbidity risk) confer a sub-

stantial disadvantage on survival over time in patients with RA compared with the general population (2). Recently, we found that the increasing survival of RA patients is contributing to the increasing prevalence of RA in Ontario (3). A recent meta-analysis confirmed that premature mortality has decreased among RA patients over time (4), a finding that has several plausible mechanisms. Advances in disease management and RA

The opinions, results, and conclusions are those of the authors and are independent from the funding sources. No endorsement by the Institute for Clinical Evaluative Sciences or the Ontario Ministry of Health and Long-Term Care is intended or should be inferred. Supported by the Canadian Institutes of Health Research (grant 119348) and the Institute for Clinical Evaluative Sciences, a nonprofit research corporation funded by the Ontario Ministry of Health and Long-Term Care. This study was also supported and performed in the context of the Ontario Best Practices Research Initiative, which performs real-world surveillance of rheumatoid arthritis outcomes in Ontario through provincial administrative database analyses. Dr. Widdifield’s work was supported by a Fellowship Award from The Arthritis Society. Dr. Bernatsky’s work was supported by a career award from the Fonds de la Recherche en Sant e du Qu ebec. Dr. Tu’s work was supported by a Canadian Institutes of Health

Research Fellowship Award in Primary Care Research. Dr. Kuriya’s work was supported by a Canadian Institutes of Health Research Fellowship Award in Clinical Research. Dr. Bombardier’s work was supported by a Canada Research Chair in Knowledge Transfer for Musculoskeletal Care (2002–2016) and a Pfizer Research Chair in Rheumatology. 1 Jessica Widdifield, PhD: Institute for Clinical Evaluative Sciences, Toronto, Ontario, and McGill University, Montreal, Quebec, Canada; 2Sasha Bernatsky, MD, FRCPC, PhD: McGill University, Montreal, Quebec, Canada; 3J. Michael Paterson, MSc: Institute for Clinical Evaluative Sciences and University of Toronto, Toronto, Ontario, and McMaster University, Hamilton, Ontario, Canada; 4George Tomlinson, PhD, Claire Bombardier, MD, FRCPC: University of Toronto, Toronto, Ontario, Canada; 5Karen Tu, MD, MSc: Institute for Clinical Evaluative Sciences and University of Toronto, Toronto,

INTRODUCTION

1047

1048

Significance & Innovations All-cause mortality rates for patients with rheumatoid arthritis (RA) has decreased over the past decade, but remains elevated compared to the general population. We demonstrate 40–50% more deaths among RA patients compared to the general population. The relative excess mortality was most strongly elevated in the younger age groups. We found no significant change in the relative excess mortality between 1996 and 2009. Therefore, our analyses suggest that the RA mortality gap remains unchanged in our sample. treatment strategies may be responsible for prolonging survival (5–7), or the course of disease activity in RA patients may becoming milder in more recent years (8). Increasing survival may also be reflective of better overall population health. However, there is conflicting evidence on whether the mortality gap (the number of excess premature deaths) among individuals with RA is narrowing, widening, or remaining unchanged over time (1,4,9–11). Therefore, our aims were to study temporal trends in all-cause mortality among Ontarians with RA versus without RA, and to determine whether the excess mortality in RA has changed over time. This report extends and consolidates our earlier study of the epidemiology of RA in Ontario, Canada, reporting on the increasing prevalence of RA over a 15-year time span (3).

MATERIALS AND METHODS Setting and design. We performed a retrospective population-based study using health administrative databases that include records for all residents with universal health coverage under the publicly funded Ontario Health Insurance Plan (OHIP). Ontario is the largest and most ethnically diverse (12) Canadian province, with more than 11 million adults (13). Ethics approval was obtained from the institutional review board at Sunnybrook Health Sciences Centre in Toronto, Ontario.

Ontario, Canada; 6Bindee Kuriya, MD, FRCPC, MSc: Mount Sinai Hospital, Toronto, Ontario, Canada; 7J. Carter Thorne, MD, FRCPC: Southlake Regional Health Centre, Newmarket, Ontario, Canada; 8Janet E. Pope, MD, FRCPC, MPH: Western University, London, Ontario, Canada; 9Simon Hollands, MSc: Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada. Dr. Bombardier has received honoraria (less than $10,000 each) from Abbott Intern, Amgen, and UCB, and (more than $10,000 each) from Pfizer and Janssen, and has served on Advisory Boards for AstraZeneca, Pfizer, and Janssen. Address correspondence to Jessica Widdifield, PhD, 687 Pine Avenue West, V-Building, Room V2-09, Montreal, Quebec, Canada H3H 2R9. E-mail: [email protected]. Submitted for publication June 16, 2014; accepted in revised form January 20, 2015.

Widdifield et al Patients and data sources. Patients with RA were identified using the Ontario RA Administrative Database (ORAD), which is generated from health administrative data using a validated RA case definition. To generate the database, hospital discharge abstracts from the Canadian Institute for Health Information (14) were used to identify patients admitted to the hospital with a diagnosis of RA, and physician service claims from the OHIP Claims History Database (15) were used to identify physician services for RA. Ontario residents having at least 1 hospital admission or 3 physicians’ claims bearing an RA diagnosis (with at least 1 provided by a musculoskeletal specialist) within 2 years are included in the database. This case definition has been validated within Ontario, based on medical record reviews of rheumatology and primary care medical records and has high sensitivity (78%), specificity (100%), positive predictive value (PPV; 78%), and negative predictive value (100%) (16,17). Once patients meet the criteria for this case definition, disease onset is defined as the first health care encounter with RA as the recorded diagnosis. Data are available from April 1, 1991, until March 31, 2011. A 5-year run-in period (from 1991 to 1995) was used to identify preexisting RA patients (those diagnosed before the onset of administrative data). This run-in period is required as all preexisting RA patients identified in the earliest years of observation would otherwise appear as incident. All patients are followed until outmigration, death, or end of study. Results are reported up until 2009 to allow for a 2-year “look forward” period to meet the terms of the case definition, allowing the proportionate number of incident and prevalent RA cases to be consistent over each year for the outcome to be assessed. For our purposes, all patients not included in ORAD comprised the general population of Ontario. As ORAD excludes patients age ,15 years, we applied this exclusion to the general population as well. Demographic information (age, sex, and place of residence), vital status and dates of death, and health insurance status were obtained from the OHIP Registered Persons Database (18). Census population estimates were obtained from Statistics Canada (13). These data sets are held securely in a linked coded form and analyzed at the Institute for Clinical Evaluative Sciences (ICES; www.ices.on.ca).

Analysis. Observed all-cause mortality rates, excess mortality rates, standardized mortality ratios (SMRs), and mortality rate ratios (MRRs) were calculated by calendar year. Crude observed mortality rates were calculated as the number of deaths per 1,000 population. To adjust for differences in population distribution over time, direct age and sex standardization were undertaken using 1991 Ontario Census population estimates (18). Observed RA mortality rates (with corresponding 95% confidence intervals [95% CIs]) were determined by dividing the number of deaths among individuals with RA each year by the number of individuals with RA in each year. The denominator includes all RA patients (prevalent and new cases identified each year). Similar analyses were done to obtain observed mortality rates in the general population without RA. Age-specific mortality rates were also determined

Excess Mortality in RA

1049

using 10-year age bins, except for patients ages 15–54 years, which were combined due to few deaths occurring within each year. We compared changes in mortality rates over time in the RA population to standardized estimates of mortality in the general population without RA, over the overall period, by computing the relative changes between 1996 and 2009 within each population. Excess mortality rates were calculated as the difference between observed mortality rates among RA patients and the general population estimates. Age-specific excess mortality rates were also estimated within strata. We computed SMRs as the ratio of the mortality rate in RA patients versus the age- and sex-specific general population mortality estimates. Rather than compute the SMR for the entire period (as our observation period started in 1991 and both the general population and RA population increases and ages over time), we present SMRs for 4 periods: 1996–1997, 2000–2001, 2004–2005, and 2008–2009. To help simplify our analysis and reporting, we divided the 14-year study period into four 2-year intervals, each separated by 2 years. Age- and sex-specific SMRs were also determined over these 4 distinct periods. We used Poisson regression with observed stratumspecific numbers of deaths as the outcome to estimate MRRs to compare the RA mortality rates across the calendaryear intervals, accounting for the expected mortality rate based on age- and sex-specific mortality rates in the Ontario population. Analyses were performed at ICES using R, version 3.0, and SAS, version 9.2, with a 2-tailed type 1 error rate of 0.05 used to determine statistical significance.

RESULTS During the study period, the number of individuals ages $15 years, the general population denominator, increased Table 1.

from 8,903,118 in 1996 to 11,595,951 in 2009. The number of patients with RA increased from 46,961 (0.53%) to 97,499 (0.84%) over the same time period (Table 1). An average of 5,790 incident RA patients entered the cohort each year over 1996–2009. Crude observed and age- and sex-standardized all-cause mortality rates for individuals with and without RA by year are presented in Figure 1. From 1996 to 2009, crude RA mortality rates ranged from 31.0 to 25.9 deaths per 1,000 RA population. Crude mortality rates for the general population without RA ranged from 8.6 to 7.3 deaths per 1,000 population over the same time period (Figure 1). Age- and sex-standardized mortality rates ranged from 13.0 deaths per 1,000 RA patients (95% CI 12.2, 13.9) in 1996 to 9.2 deaths per 1,000 RA patients (95% CI 8.4, 10.0) in 2009. For individuals without RA, standardized mortality rates ranged from 8.7 deaths per 1,000 population (95% CI 8.6, 8.7) in 1996 to 6.0 deaths per 1,000 population (95% CI 5.9, 6.0) in 2009 (Table 1 and Figure 1). Since 1996, allcause mortality decreased for RA by a relative 21.4%, with a smaller decrease (13.4%) in the general population. Observed age-specific mortality rates for individuals with and without RA and excess mortality rates are presented in Table 2. Age-specific mortality rates increased with age for both RA and non-RA individuals. As of 2008– 2009, all-cause mortality in RA ranged from 3.8 (95% CI 3.2, 4.6) per 1,000 RA population in those age ,55 years to 131.7 (95% CI 122.5, 141.4) per 1,000 RA population among seniors age .85 years. In the general population, lower age-specific mortality rates were observed for each age group in comparison to the RA population. Agespecific mortality among the general population also decreased among each age-specific stratum over time. Overall, the excess mortality rate ranged from 4.4 to 2.7 excess deaths for patients with RA per 1,000 population. The SMRs for RA patients in 1996–1997, 2000–2001, 2004–2005, and 2008–2009 were 1.51 (95% CI 1.43, 1.59),

Age- and sex-standardized all-cause mortality for individuals with and without RA, by year* RA population†

Year

Deaths, no.

RA population, no.

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

1,458 1,542 1,721 1,844 1,888 1,966 2,074 2,076 2,158 2,139 2,347 2,366 2,461 2,529

46,961 51,248 55,398 59,129 62,795 66,537 69,997 73,575 77,330 81,614 85,706 89,420 93,558 97,499

General population Deaths per 1,000 RA (95% CI)

13.0 12.8 12.7 13.0 12.3 11.6 11.8 10.8 10.5 9.6 10.5 9.5 10.2 9.2

(12.2, 13.9) (11.8, 13.8) (11.9, 13.7) (11.9, 14.2) (11.3, 13.3) (10.9, 12.3) (10.8, 12.9) (10.1, 11.6) (9.8, 11.1) (9.0, 10.1) (9.7, 11.4) (8.8, 10.3) (9.3, 11.3) (8.4, 10.0)

* RA 5 rheumatoid arthritis; 95% CI 5 95% confidence interval. † Originally published in ref. 3.

Deaths, no.

Ontario population, no.

76,958 77,470 77,033 77,604 77,150 77,842 79,494 81,204 80,707 80,888 83,111 83,435 85,136 84,736

8,903,118 9,051,199 9,224,340 9,395,670 9,596,442 9,821,357 10,034,926 10,237,615 10,440,885 10,653,622 10,902,233 11,139,392 11,370,572 11,595,951

Deaths per 1,000 (95% CI) 8.7 8.5 8.1 7.9 7.6 7.4 7.3 7.2 7.0 6.7 6.6 6.4 6.3 6.0

(8.6, (8.4, (8.1, (7.9, (7.6, (7.4, (7.3, (7.2, (6.9, (6.7, (6.6, (6.3, (6.2, (5.9,

8.7) 8.5) 8.2) 8.0) 7.7) 7.5) 7.4) 7.3) 7.0) 6.8) 6.7) 6.4) 6.3) 6.0)

1050

Widdifield et al

Deaths/1,000 population

35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

Crude Rate: RA

31.0

30.1

31.1

31.2

30.1

29.5

29.6

28.2

27.9

26.2

27.4

26.5

26.3

25.9

Standardized Rate: RA

13.0

12.8

12.7

13.0

12.3

11.6

11.8

10.8

10.5

9.6

10.5

9.5

10.2

9.2

Crude Rate: General Population

8.6

8.6

8.4

8.3

8.0

7.9

7.9

7.9

7.7

7.6

7.6

7.5

7.5

7.3

Standardized Rate: General Population

8.7

8.5

8.1

7.9

7.6

7.4

7.3

7.2

7.0

6.7

6.6

6.4

6.3

6.0

Figure 1. Crude and standardized all-cause mortality rates for individuals with and without rheumatoid arthritis (RA), 1996 to 2009.

1.50 (95% CI 1.43, 1.57), 1.43 (95% CI 1.37, 1.50), and 1.41 (95% CI 1.35, 1.47), respectively. The age- and sexspecific SMRs were strongly elevated in the younger age groups, although the SMR declines by the time patients reach the final age stratum ($85 years) (Figure 2). There were no significant differences across sex-specific SMRs within each age category over the years studied. Overall, SMR point estimates were generally higher among women than men within each age category. While there was potentially a slight trend in decreasing SMR point estimates over time (with overlapping CIs), when we assessed relative differences in mortality compared to 1996 (referent) after taking into account changes within the background population, this regression model showed no significant change in the MRRs over time (Figure 3). The relative excess mortality can also be seen to be lower for the older age groups compared to the youngest age stratum (Figure 3).

DISCUSSION In the universal public health insurance system of Ontario, we studied temporal trends from 1996 to 2009 in all-cause mortality for individuals with and without RA, identified that RA patients have excess mortality compared with the general population, and observed that mortality rates have decreased over time. While we observed a slight trend toward decreasing SMRs over time, when we assessed the relative change in excess mortality among RA patients using 1996 as the reference, we found no significant change in the MRR by calendar time. Published mortality rates in RA patients vary considerably, and comparing trends in excess mortality over time across settings has challenges. Dissimilar patient populations, different reference mortality rates, and temporal effects can contribute to different mortality estimates across settings. In our sample, we observed no increase in the mortality gap over time. As there have been dramatic declines in the overall mortality rates for the general pop-

ulation worldwide, for a widening mortality gap to occur over time, RA mortality rates must remain stable or increase over time. Previous studies defining the mortality gap between RA patients and the general population that found an increasing temporal mortality gap in RA primarily focused on inception cohorts established before the year 2000 to enable prospective followup over time (10,11). We opted to include all RA patients in our population in 1991–2009 to also reflect changes in the paradigm of RA care over time, as treatment strategies have shown a positive impact on the survival of RA patients (7). Still, the relative excess mortality in RA remained unchanged over time despite substantial improvements in RA management and growing knowledge of the determinants of increased mortality (2). In Ontario, we have observed substantial increases in timely access to rheumatologists (19) and access to disease-modifying treatment for patients with RA in recent years (20). Therefore, we can only speculate that more recent improvements in RA care will show a positive influence on the relative excess mortality in RA in the future. In general, lower SMRs (1.2–1.3) are reported in studies of inception or early RA cohorts compared to prevalent cohorts with higher SMRs (1.6–1.7) (1). Prevalent RA cohorts are more likely to include patients with established disease who have experienced long-term chronic inflammation and may have accrued higher comorbidity over time. This may be important in interpreting the findings of some inception RA cohorts who have not clearly shown declining mortality rates over time (4,9,11), although a recent meta-analysis concluded that RA mortality rates have decreased over the past 50 years (4). In our study, we included all RA patients in the population (both incident and prevalent cases) and observed an SMR (1.4–1.5) that agrees with the spectrum of excess mortality in RA. While we observed declining mortality rates in our RA population, there remained an average of 3 extra deaths per 1,000 patients with RA. This translates to over 34,000 lives shortened due to RA among a population of 11

3.6 (2.8, 4.6) 15.4 (13.0, 18.1) 35.3 (32.0, 39.0) 85.2 (78.4, 92.3) 166.8 (147.2, 188.2) 13.0 (12.2, 13.9)

8.6 (8.6, 8.7)

1.3 (1.2, 1.3) 8.5 (8.3, 8.7) 22.4 (22.1, 22.8) 59.1 (58.4, 59.9) 169.9 (167.5, 172.3)

General population death rate per 1,000 (95% CI) 2.4 (1.5, 3.3) 6.9 (4.4, 9.4) 12.9 (9.5, 16.4) 26.0 (19.4, 32.6) 2 3.1 (2 21.7, 15.5) 4.4 (3.3, 5.4)

Excess mortality rate per 1,000 (95% CI)

12.5 (10.6, 14.6) 32.6 (29.7, 35.8) 75.4 (70.2, 80.8) 149.7 (135.5, 164.9) 12.3 (11.3, 13.3)

3.8 (3.0, 4.6)

RA death rate per 1,000 RA (95% CI)

8.0 (8.0, 8.1)

1.2 (1.2, 1.2) 7.2 (7.1, 7.4) 19.6 (19.3, 19.9) 51.1 (50.4, 51.7) 152.4 (150.3, 154.5)

General population death rate per 1,000 (95% CI)

2000–2001

2.6 (1.8, 3.4) 5.3 (3.3, 7.2) 13.1 (10.1, 16.1) 24.3 (19.2, 29.4) 2 2.7 (2 16.2, 10.8) 4.3 (3.4, 5.1)

Excess mortality rate per 1,000 (95% CI) 3.4 (2.7, 4.1) 10.4 (8.9, 12.0) 27.8 (25.2, 30.5) 68.8 (64.4, 73.4) 140.4 (128.9, 152.7) 10.5 (9.8, 11.1)


7.7 (7.7, 7.8)

1.1 (1.1, 1.2) 6.5 (6.4, 6.7) 17.4 (17.1, 17.7) 47.3 (46.8, 47.9) 139.6 (137.8, 141.5)


2004–2005

2.3 (1.6, 2.9) 3.8 (2.3, 5.4) 10.4 (7.8, 13.0) 21.4 (17.1, 25.8) 0.8 (2 10.3, 11.8) 2.8 (2.0, 3.5)


3.8 (3.2, 4.6) 9.8 (8.5, 11.2) 24.6 (22.4, 26.9) 58.4 (54.7, 62.4) 131.7 (122.5, 141.4) 10.2 (9.3, 11.3)


Excess mortality rates and age-specific mortality rates for individuals with and without RA*

* RA 5 rheumatoid arthritis; 95% CI 5 95% confidence interval. † Age and sex standardized to the 1991 Ontario population.

All ages†

$85

75–84

65–74

55–64

15–54

Age group, years


1996–1997

Table 2.

7.5 (7.4, 7.5)

1.1 (1.0, 1.1) 6.0 (5.8, 6.1) 15.3 (15.0, 15.5) 42.7 (42.2, 43.3) 124.3 (122.8, 125.8)


2008–2009

2.8 (2.1, 3.4) 3.8 (2.5, 5.1) 9.3 (7.1, 11.6) 15.7 (11.9, 19.4) 7.4 (2 1.5, 16.2) 2.7 (2.1, 3.4)


Excess Mortality in RA 1051

1052

Widdifield et al

Figure 2. Age- and sex-specific standardized mortality ratios (SMRs) for 1996, 2000, 2004, and 2008. 95% CI 5 95% confidence interval.

million residents. Our findings clearly illustrate the need to understand reasons for inequalities in mortality, particularly surrounding the youngest age groups with the highest relative excess mortality. Though we have not yet examined cause of death, it is known that cardiovascular mortality is the leading contributor to premature mortality in RA and a recent metaanalysis has found that cardiovascular mortality rates in RA have remained relatively unchanged over time (21). It

Figure 3. Mortality rate ratios.

is hoped that the heightened awareness of increased cardiovascular risk in patients with RA may lead to more aggressive cardiovascular disease detection and treatment and eventually reduce excess cardiovascular mortality. The presentation of cardiovascular disease in RA is probably somewhat different from the signs, symptoms, and course that are known in the general population. Future research is required to determine the influence of cardiovascular comorbidity and mortality in our sample. Our study has both strengths and limitations. We used an RA case definition that has high PPV, sensitivity, and specificity based on validation with patient clinical records (16,17). In addition, this is one of the largest epidemiologic studies of mortality in RA in a universal health care system with long and complete followup of all subjects, including general population mortality rates throughout the entire study period. We also used methods similar to those of other studies examining trends in mortality within validated population-based chronic disease cohorts in Ontario (22–24). A limitation is that health services data can only be used to assess the care of RA patients who sought and had access to health care providers and received a physician diagnosis of RA, but this is true in any patient population. Similarly, as we were limited to administrative data from 1991 onward, our population-based cohort captures fewer seniors age .85 years in the earliest years of study (3); therefore there is a potential for misclassification

Excess Mortality in RA in the earliest years. However, the registration of deaths in Ontario is considered to be reasonably complete because of legal reporting requirements, and Statistics Canada estimates that ,1% of Ontario residents die outside the province (13,18). Finally, some of the factors influencing mortality, such as cardiovascular disease and smoking, were not explored. Therefore, the estimated excess risks of mortality might differ within each age group. In conclusion, all-cause mortality rates for patients with RA has decreased over the past decade but remains elevated compared to the general population. Our results suggest 40–50% more deaths among RA patients compared to the general population. While we observed a slight trend toward decreasing SMRs over time, we found no significant change in the relative excess mortality between 1996 and 2009. Therefore, our analyses suggest that the RA mortality gap remains unchanged in our sample.

1053

8.

9.

10.

11.

12. 13.

AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Widdifield had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Widdifield, Bernatsky, Paterson, Tomlinson, Tu, Thorne, Bombardier. Acquisition of data. Widdifield, Paterson, Thorne, Pope, Hollands, Bombardier. Analysis and interpretation of data. Widdifield, Bernatsky, Paterson, Tomlinson, Tu, Kuriya, Thorne, Pope, Hollands, Bombardier.

REFERENCES 1. Sokka T, Abelson B, Pincus T. Mortality in rheumatoid arthritis: 2008 update. Clin Exp Rheumatol 2008;26 Suppl 51:S35–61. 2. Myasoedova E, Davis JM III, Crowson CS, Gabriel SE. Epidemiology of rheumatoid arthritis: rheumatoid arthritis and mortality. Curr Rheumatol Rep 2010;12:379–85. 3. Widdifield J, Paterson JM, Bernatsky S, Tu K, Tomlinson G, Kuriya B, et al. The epidemiology of rheumatoid arthritis in Ontario, Canada. Arthritis Rheum 2014;66:786–93. 4. Dadoun S, Zeboulon-Ktorza N, Combescure C, Elhai M, Rozenberg S, Gossec L, et al. Mortality in rheumatoid arthritis over the last fifty years: systematic review and metaanalysis. Joint Bone Spine 2013;80:29–33. 5. Carmona L, Descalzo MA, Perez-Pampin E, Ruiz-Montesinos D, Erra A, Cobo T, et al. for the BIOBADASER and EMECAR Groups. All-cause and cause-specific mortality in rheumatoid arthritis are not greater than expected when treated with tumour necrosis factor antagonists. Ann Rheum Dis 2007;66:880–5. 6. Jacobsson LT, Turesson C, Nilsson J, Petersson IF, Lindqvist E, Saxne T, et al. Treatment with TNF blockers and mortality risk in patients with rheumatoid arthritis. Ann Rheum Dis 2007;66:670–5. 7. Van Nies JA, de Jong Z, van der Helm-van Mil AH, Knevel R, Le Cessie S, Huizinga TW. Improved treatment strategies

14. 15. 16.

17.

18. 19.

20.

21.

22.

23. 24.

reduce the increased mortality risk in early RA patients. Rheumatology (Oxford) 2010;49:2210–6. Welsing PM, Fransen J, van Riel PL. Is the disease course of rheumatoid arthritis becoming milder? Time trends since 1985 in an inception cohort of early rheumatoid arthritis. Arthritis Rheum 2005;52:2616–24. Humphreys JH, Warner A, Chipping J, Marshall T, Lunt M, Symmons DP, et al. Mortality trends in patients with early rheumatoid arthritis over 20 years: results from the Norfolk Arthritis Register. Arthritis Care Res (Hoboken) 2014;66: 1296–301. Radovits BJ, Fransen J, Al Shamma S, Eijsbouts AM, van Riel PL, Laan RF. Excess mortality emerges after 10 years in an inception cohort of early rheumatoid arthritis. Arthritis Care Res (Hoboken) 2010;62:362–70. Gonzalez A, Kremers HM, Crowson CS, Nicola PJ, Davis JM III, Therneau TM, et al. The widening mortality gap between rheumatoid arthritis patients and the general population. Arthritis Rheum 2007;56:3583–7. Statistics Canada. Canada’s ethnocultural portrait: the changing mosaic. 2003. URL: statcan.gc.ca. Statistics Canada. Population and demography. 2011. URL: http://www12.statcan.gc.ca/census-recensement/2011/ref/92135/topics-themes/pop-eng.cfm. Canadian Institute for Health Information. The CIHI data quality framework. Ottawa (Ontario, Canada): CIHI; 2009. Ontario Ministry of Health and Long-Term Care. Ontario Health Insurance Plan: the program. 2008. URL: http:// www.health.gov.on.ca/en/public/programs/ohip/. Widdifield J, Bombardier C, Bernatsky S, Paterson JM, Green D, Young J, et al. An administrative data validation study of the accuracy of algorithms for identifying rheumatoid arthritis: the influence of the reference standard on algorithm performance. BMC Musculoskelet Disord 2014;15: 216. Widdifield J, Bernatsky S, Paterson JM, Tu K, Ng R, Thorne JC, et al. Accuracy of Canadian health administrative databases in identifying patients with rheumatoid arthritis: a validation study using the medical records of rheumatologists. Arthritis Care Res (Hoboken) 2013;65: 1582–91. Ministry of Health and Long-Term Care. Health analyst’s toolkit. 2012. URL: www.health.gov.on.ca/english/providers/ pub/healthanalytics/health_toolkit/health_toolkit.pdf. Widdifield J, Paterson JM, Bernatsky S, Tu K, Thorne JC, Ivers N, et al. Access to rheumatologists among patients with newly diagnosed rheumatoid arthritis in a Canadian universal public healthcare system. BMJ Open 2014;4: e003888. Widdifield J, Bernatsky S, Paterson JM, Thorne JC, Cividino A, Pope J, et al. Quality care in seniors with new-onset rheumatoid arthritis: a Canadian perspective. Arthritis Care Res (Hoboken) 2011;63:53–7. Meune C, Touze E, Trinquart L, Allanore Y. Trends in cardiovascular mortality in patients with rheumatoid arthritis over 50 years: a systematic review and meta-analysis of cohort studies. Rheumatology (Oxford) 2009;48:1309–13. Tu K, Chen Z, Lipscombe LL, for the Canadian Hypertension Education Program Outcomes Research Taskforce. Mortality among patients with hypertension from 1995 to 2005: a population-based study. CMAJ 2008;178:1436–40. Lipscombe L, Hux J. Trends in diabetes prevalence, incidence, and mortality in Ontario, Canada 1995–2005: a population-based study. Lancet 2007;369:750–6. Gershon AS, Wang C, Wilton AS, Raut R, To T. Trends in chronic obstructive pulmonary disease prevalence, incidence, and mortality in Ontario, Canada, 1996 to 2007: a population-based study. Arch Intern Med 2010;170:560–5.