ORIGINAL RESEARCH Trends in Socioeconomic Status–related Differences in Mortality among People with Chronic Obstructive Pulmonary Disease Andrea S. Gershon1,2,3,4, Jeremiah Hwee2, J. Charles Victor2,4, Andrew S. Wilton2, and Teresa To2,3,5 1
Department of Medicine, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada; 2Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada; 3Child Health Evaluative Sciences, The Hospital for Sick Children, Toronto, Ontario, Canada; and 4Department of Medicine, and 5The Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada
Abstract Rationale: Previous studies have demonstrated that people of higher socioeconomic status (SES) have better chronic obstructive pulmonary disease (COPD) health outcomes than those of lower SES. Mortality of people with COPD has decreased over the last decade; however, it is not known if all individuals with COPD have benefitted equally. Objective: The objective of the current study was to examine the impact of SES on mortality trends of individuals with COPD. Methods: We conducted a population-based study using health administrative data from Ontario, Canada, between 1996/7 and 2011/12. Individuals with COPD were identified using a previously validated case definition. SES was determined using average household income of the individual’s neighborhood as per the Canadian Census. Trends in standardized COPD mortality rates among different SES quintiles were observed over time.
Measurements and Main Results: Overall, age- and sexstandardized mortality of people with COPD decreased from 5.7% (95% confidence interval [CI], 5.5–5.8) in 1996/7 to 3.7% (95% CI, 3.6–3.8) in 2011/12, representing a 35% relative decrease. The mortality difference between the lowest and highest income quintiles increased from 67 per 10,000 individuals in 1996/7 to 86 per 10,000 individuals in 2011/12, representing a 28% relative increase (P , 0.001). Conclusions: Mortality in people with COPD has decreased faster in people with the highest compared with the lowest SES, causing increased disparity between rich and poor. Further study and strategies are needed to explore and address factors responsible for this increasing disparity in the COPD population. Keywords: pulmonary disease, chronic obstructive; mortality; income; social class; trends
(Received in original form March 5, 2014; accepted in final form July 19, 2014 ) This work was supported by funding from the Government of Ontario; by the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care; by a New Investigator Award funded by team grant OTG-88591 from the Canadian Institutes of Health Research Institute of Nutrition, Metabolism and Diabetes (A.S.G.); by the Physicians Services Incorporated Foundation Fellowship for Translational Health Research (A.S.G.); and by The Dales Award in Medical Research from the University of Toronto, Toronto, Ontario, Canada (T.T.). Author Contributions: A.G. participated in the study conception and design, acquisition, analysis and interpretation of data; drafting and critical revision of the manuscript; obtaining funding; supervising the writing; and approving the final version. J.H. and J.C.V. participated in the analysis and interpretation of data, participated in statistical analysis and critical revision of the manuscript, and approved the final version. A.S.W. participated in analysis and interpretation of data, critical revision of the manuscript, and approving the final version. T.T. participated in the conception and design, analysis and interpretation of data, critical revision of the manuscript, and approving the final version. Correspondence and requests for reprints should be addressed to Andrea Gershon, M.D., Institute for Clinical Evaluative Sciences, G1 06, 2075 Bayview Avenue, Toronto, ON, M4N 3M5 Canada. E-mail: [email protected] This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org Ann Am Thorac Soc Vol 11, No 8, pp 1195–1202, Oct 2014 Copyright © 2014 by the American Thoracic Society DOI: 10.1513/AnnalsATS.201403-094OC Internet address: www.atsjournals.org
Chronic obstructive pulmonary disease (COPD) is a manageable, progressive condition that places a large burden on individuals and health care systems. COPD is the third leading cause of death in North America, affects 8 to 22% of the adult
population, and is one of the most common reasons for hospitalization (1–3). Moreover, COPD prevalence is increasing, particularly in women, and this trend is likely to worsen as the population ages (2). On the positive side, overall mortality of
Gershon, Hwee, Victor, et al.: SES Mortality Trends in COPD
people with COPD is decreasing, likely due to declining smoking rates and better treatment of comorbidities like cardiovascular disease (2, 4, 5). This progress, however, has not been experienced by all individuals with COPD. 1195
ORIGINAL RESEARCH Previous studies have demonstrated an inverse association between socioeconomic status (SES) and various COPD outcomes, such as mortality (6–8). To the best of our knowledge, however, no studies have examined how this relationship has changed over time. In Canada before 1996, mortality differences between rich and poor populations were decreasing (9), but it is not known if this trend continued or if it applies to individuals with COPD. It does not appear to apply to the general population in the United States (10) or to people with other chronic diseases (e.g., diabetes), where socioeconomic disparity in mortality has been increasing (11). Government spending cuts on health care services, public health, and basic living conditions have led many to predict that differences in health between the rich and the poor will increase (12). For people with COPD, cuts might lead to increases in smoking rates and greater exposure to working and living conditions that worsen their disease (13). We conducted the current population study to test the hypothesis that observed socioeconomic mortality differences of individuals with COPD are increasing. Some of the results of this study have been previously reported in the form of an abstract (14).
Methods In this longitudinal, population cohort study of people with COPD, we used health administrative data from Ontario, Canada to compare observed real-world changes in mortality by SES from fiscal years (April– March) 1996/7 to 2011/12. Data Sources
Ontario is a province of Canada with a diverse multicultural population of approximately 13 million. Residents of Ontario have universal public health insurance under the Ontario Health Insurance Plan, the single payer for all medically necessary services across the full spectrum of residents, providers, and hospitals. Service details are captured in health administrative databases, which can be linked on an individual level to provide a complete health services profile for each resident. The only exception is the provision of prescription medications, which are only provided to those 65 years of age or older and to those who are on welfare. Three 1196
Ontario population health administrative databases were used. The Registered Persons Database contains basic demographic information and date of death (as applicable). Deaths recorded in this database are based on probabilistic linkage of death certificate information and are cross referenced with hospital records of in-hospital deaths. The Canadian Institute of Health Information Discharge Abstract database contains information on all discharges from acute care hospitals. The Ontario Health Insurance Plan Physician Claims database contains information on all services provided by fee-for-service physicians and “shadow-billings” for physicians paid under alternate payment plans. These data sets used in this study were held securely in a linked, deidentified form and analyzed at the Institute for Clinical Evaluative Sciences. Study Population
We identified individuals with COPD using the Ontario COPD Database, which identifies people with COPD using the above health administrative databases and a previously validated COPD case definition of one or more COPD physician billing claims and/or one or more COPD hospital discharges as per the following codes: 491, 492, 496 (Ontario Health Insurance Plan and International Classification of Disease, 9th revision codes) or J41, J42, J43, J44 (International Classification of Disease, 10th revision codes). Individuals have to be age 35 years of age or older when their claim(s) or discharge occurs. This definition was found to have a sensitivity of 85% and specificity of 78% when compared with clinical evaluation and has been well described elsewhere (2, 15, 16). Once patients are identified as having COPD, including people who recently immigrated, they remain in the database until they die or move out of the province. Exposure
SES was attributed using the subject’s postal code to determine the average household income of their neighborhood derived from the Canadian census and categorized into quintiles, with 1 representing the lowest and 5 the highest (17). We used the 1996 census for fiscal years 1996 to 1998, the 2001 census for fiscal years 1999 to 2003, and the 2006 census for fiscal years 2004 and onward. People whose postal code could not be linked to a neighborhood
income level were considered “missing” and were excluded from the primary analysis. Outcomes
Annual overall and income quintile specific all-cause mortality rates in people with COPD, expressed as the percentage of people with COPD who died in a given year, were calculated from fiscal years 1996/7 to 2011/12. All-cause mortality, instead of COPD mortality, was used because COPD has been found to be underestimated as a cause of death on vital statistics death records by around 50%, because COPD deaths are often attributed to its comorbidities like pneumonia and cardiovascular disease and because cause of death information was not available (18, 19). Analysis
Frequencies and proportions were used to describe the COPD population. Overall mortality rates by socioeconomic status quintile and the mortality rate income gradient (i.e., the absolute difference in rate between the lowest and highest income quintiles) and the mortality rate ratio (i.e., the rate in the lowest divided by the highest income quintile) were compared over fiscal years. To adjust for differences in population distribution over time, mortality rates were standardized to the 2001 COPD population (20). We created 35 to 64 and over 65 years age groups because Ontario residents 65 years and older are eligible for the provincial medication insurance program and therefore may be less likely to be affected by income. To explore whether mortality trends differed by age and sex, we calculated sex-standardized mortality by age group (35–64 yr and over 65 yr) and age-standardized mortality by sex for each income quintile. Ninety-five percent confidence intervals (CIs) were calculated using the g distribution. Mortality rate ratios were based on the age- and sexstandardized rates. Multivariable negative binomial regression was used to compare mortality by study year, socioeconomic status, age, and sex. A model was fit relating the variables of interest to number of deaths in a given year using the corresponding COPD population for that year as the offset. Two- and threeway interactions between socioeconomic status, age, sex, and year were used to determine if mortality trends differed by income and whether such differences differed between men and women or those
AnnalsATS Volume 11 Number 8 | October 2014
ORIGINAL RESEARCH Table 1. Ontario chronic obstructive pulmonary disease population demographics, 1996/7 and 2011/12 Demographic n Age, mean yr (SD) Female sex, % Urban* (vs. rural) residence, % Socioeconomic status quintile, % 1 (lowest) 2 3 4 5 (highest)
1996/7
2011/12
428,046 65.1 (14.2) 47.7 81.5
807,996 66.3 (13.9) 50.7 84.3
24.9 22.3 19.8 17.1 16.0
23.2 21.7 19.5 18.7 16.9
*Urban communities were identified using Statistics Canada’s definition, which consisted of a community with a population of more than 10,000 people.
age 35 to 64 years and those age 65 years and older for whom prescription drug costs are covered by provincial health insurance. Predicted mortality rates for age and sex subgroups were based on the binomial regression model.
Ethics Approval
Our study received ethics approval from the institutional review board at Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada. Conduct of the research without explicit consent from the participants was approved.
Sensitivity Analysis
The main analysis was repeated including people with missing SES data. Income quintiles for these individuals were imputed based on age, sex, and year.
increased to 807,996 in 2011/12, which represents an 88.8% increase (Table 1). SES could not be determined for 0.4% of individuals who were excluded from the analysis. Between 1996/7 and 2011/12, the COPD population became older, the number of female patients with COPD increased, and SES increased slightly so that in 2011/12 the COPD population had a mean age of 66.3 years, was slightly more female (50.7%), and had 23.2% in the lowest and 16.9% in the highest SES quintiles. Mortality among People with COPD
Overall, the age- and sex-standardized allcause mortality of the COPD population decreased from 5.7% (95% CI, 5.5–5.8) in 1996 to 3.7% (95% CI, 3.6–3.8) in 2011/12, representing a 35% relative decrease. This decrease was greater in men than in women (41% compared with 27%; P , 0.001) and was greater in individuals 35 to 64 years of age than in individuals 65 years of age and older (44% compared with 34%; P , 0.001).
Results Mortality by SES Quintile
In 1996/7 there were 428,046 people with COPD living in Ontario. This number
In all study years, we observed an inverse association between income quintile and
7
Standardized mortality rate (per 100 persons)
6
5
4
3
Socioeconomic status 1 (lowest) 2 3
2
4 5 (highest)
1
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year Figure 1. All-cause mortality rate by socioeconomic status quintile among with patients with chronic obstructive pulmonary disease from 1996/7 to 2011/ 12. Rates are age and sex standardized to the 2001 chronic obstructive pulmonary disease population.
Gershon, Hwee, Victor, et al.: SES Mortality Trends in COPD
1197
ORIGINAL RESEARCH mortality. Crude mortality rates by income and by age group are presented in Table E1 in the online supplement. Age- and sexstandardized mortality in all quintiles declined over time; however, the decline was less pronounced from 1996/7 to 2011/12 in the lowest compared with the highest income quintiles (30% compared with 38%; P , 0.001) (Figure 1). This resulted in an increase of 28% (P , 0.001) in the mortality rate difference between the lowest and highest income quintiles from 67 per 10,000 individuals in 1996/7 to 86 per 10,000 individuals in 2011/12. The mortality rate ratio (i.e., the ratio of the highest compared with the lowest income quintile) increased 11% from 1.18 in 2006/7 to 1.26 in 2011/12 (P , 0.001) (Figure 2). Age and Sex
A significant interaction involving all main factors (year, income, and age group [P , 0.001]; year and income [P = 0.023]; and year and sex [P , 0.001]) was identified. Therefore, when examining differences in income gradient over time, analyses were stratified by sex and age group.
The mortality difference between the lowest and highest income quintiles increased significantly over the study period among younger women (P , 0.001) (Figure 3A) and older men (P = 0.006) (Figure 4B) but not among older women (P = 0.08) (Figure 3B) and younger men (P = 0.139) (Figure 4A). Sensitivity Analysis
When a sensitivity analysis including the 0.4% of individuals for whom income quintile was missing was conducted, mortality rate estimates varied by a relative difference of, at most, 1.6%.
Discussion We conducted a longitudinal population study of individuals with COPD and found that, although observed overall mortality rates declined between 1996/7 and 2011/12, the difference between the lowest and highest income quintiles increased by 28%. This trend was most evident in younger women and older men and suggests that, even in a jurisdiction with universal health
care insurance, improvements in care for people with COPD disproportionately benefit the affluent. To the best of our knowledge, this is the first study to examine trends in socioeconomic-related health outcomes in people with COPD. Our findings are consistent with literature showing COPD mortality and other outcomes to be inversely associated with socioeconomic status (as measured by education, occupation, and income) at one point or period of time (6). Our study expands on these findings by examining whether this association has changed over time. Our study is also consistent with literature demonstrating increases in socioeconomic mortality differences for other chronic diseases, such as diabetes and stroke (11, 21). Our results also may be, at least in part, reflective of socioeconomic mortality differences in the general population, although we are not aware of any studies that have documented this. There might be several explanations for our findings. Government funding cuts on health care services, public health, and basic living conditions have decreased
1.4
Socioeconomic status 1 (lowest) vs. 5 (highest)
Mortality rate ratio
1.3
1.2
1.1
1 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year Figure 2. Mortality rate ratio of the lowest socioeconomic status quintile compared with the highest quintile among Ontarians with chronic obstructive pulmonary disease from 1996/7 to 2011/12.
1198
AnnalsATS Volume 11 Number 8 | October 2014
ORIGINAL RESEARCH
A 1.4
1.2
Mortality rate, %
1
0.8
0.6
Socioeconomic status 1 (lowest) 2
0.4 3 4 0.2
5 (highest)
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
B
9
8
7
Mortality rate, %
6
5 Socioeconomic status 4
1 (lowest) 2
3
3 4
2 5 (highest) 1
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year Figure 3. Predicted mortality by socioeconomic status quintile among women with chronic obstructive pulmonary disease 35 to 64 years of age (A) and women 65 years of age or older (B) from 1996/7 to 2011/12.
Gershon, Hwee, Victor, et al.: SES Mortality Trends in COPD
1199
ORIGINAL RESEARCH
A 2.5
Mortality rate, %
2
1.5
1
Socioeconomic status 1 (lowest) 2 3
0.5
4 5 (highest)
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
B 12
10
Mortality rate, %
8
6
Socioeconomic status 1 (lowest) 2
4
3 4 5 (highest)
2
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year Figure 4. Predicted mortality by socioeconomic status quintile among men with chronic obstructive pulmonary disease 35 to 64 years of age (A) and men 65 years of age or older (B) from 1996/7 to 2011/12.
1200
AnnalsATS Volume 11 Number 8 | October 2014
ORIGINAL RESEARCH opportunities for education, employment, food security, and stable neighborhoods and have had adverse effects on health that disproportionately affect those of lower SES (12). For people with COPD, this might cause psychological stress and lead to higher rates of smoking and/or fewer attempts to quit (22). Funding cuts might also result in fewer smoking cessation and prevention programs. People of lower SES with COPD may also have occupations that contribute to their disease in that they have less flexibility to extricate themselves. Similarly, they may live in areas where they are exposed to more air pollution. Treatments of chronic disease, including COPD, have become more complicated, involving medications that might be difficult for those with less education, less time, and less money to follow and maintain. Even successfully quitting smoking has become complicated, with drug costs that are not always subsidized by insurance. Finally, because people with COPD often die of comorbidities related to their COPD and not of COPD itself, differences in the management of other chronic disease (e.g., cardiovascular disease) in patients of different socioeconomic strata might contribute to the results. Although these changes would be unlikely to cause dramatic year-to-year fluctuations over an entire population, it is feasible that they could lead to a steady gradual decline in health over time. Inequalities in health are a central challenge for public health. Eliminating the disparity could potentially reduce disease burden, need for health services, and costs more than newly developed medications (23). These findings, however, suggest that the disparity is increasing. This is of particular importance given the higher prevalence of COPD in lower-income
populations (6). This trend is likely more pronounced in nations without universal health care coverage, especially developing countries. Our study points to the need to address barriers to prevention and management of COPD in low socioeconomic populations. We found that increasing disparity appeared most pronounced in younger women and older men. Higher disparity in older men could be due to higher rates of cardiovascular disease, for which COPD is a risk factor, making this population more vulnerable to socioeconomic influences on their health. Higher disparity in younger women might be because they put the care of their children and family responsibilities before their own health, causing them to be more susceptible to socioeconomic pressures. Women are also more likely to be underdiagnosed compared with men, and this might be compounded by less access to physicians among populations of lower SES (24). This is an important area for future study. A strength of our study is its ability to measure trends in COPD mortality according to SES in a large, complete population. Our study also had several limitations. First, our case definition of COPD might have been subject to some misclassification error (15). Any error is likely consistent over socioeconomic strata and time and therefore unlikely to bias trends; however, one could argue that the increasing awareness of COPD that was more pronounced in people of higher socioeconomic status (SES) led to more patients with milder disease being included in this group and a perceived lower mortality rate. However, we think this scenario is unlikely because previous studies have shown that people of lower, not higher, SES are more likely to be identified with COPD (6). Second, we were
References 1 Murphy S, Xu J, Kochanek K. Deaths: preliminary data for 2010. Natl Vital Stat Rep 2012;60:1–51. 2 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;170:560–565. 3 Benady S. The human and economic burden of COPD: a leading cause of hospital admission in Canada. Ottawa, Ontario, Canada: The Canadian Thoracic Society; 2010. 4 Gilmore J. Report on smoking in Canada 1985 to 2001. 2002. Ottawa, Ontario, Canada: Minister of Industry;2002.
Gershon, Hwee, Victor, et al.: SES Mortality Trends in COPD
only able to study diagnosed COPD, and it is known that there are many people with undiagnosed COPD in the population (25). Rates of underdiagnosis may be higher in lower socioeconomic groups because of less access to physicians (26). If this led to delayed diagnosis and, as a result, more advanced disease in this group, their mortality could appear higher. Nonetheless, as physicians have become more aware of undiagnosed COPD in recent years, it is unlikely to have gotten worse over time and to account for the trends observed (27). Third, we did not have income data at the individual level. However, this ecological measure has been shown to correlate well with individual-level measures, can account for other neighborhood factors that contribute to SES, and has proven to be robust in many population-based, peerreview studies (28, 29). Most importantly, any misrepresentation there might be of SES would be unlikely to change over time. Fourth, this study took place in Canada, where universal health care insurance that covers all medically necessary services has improved health care access for poorer Canadians; therefore, this study might underestimate socioeconomic gradients and trends in other jurisdictions without such coverage. Finally, we did not have information on cause of death or smoking so were unable to compare differences in COPD-specific mortality or smoking rates due to SES over time. In summary, socioeconomic disparity in mortality in people with COPD appears to be getting worse. More study is needed to determine the factors responsible for this trend and how they can be addressed to improve the health of the population. n Author disclosures are available with the text of this article at www.atsjournals.org.
5 Tu JV, Nardi L, Fang J, Liu J, Khalid L, Johansen H; Canadian Cardiovascular Outcomes Research Team. National trends in rates of death and hospital admissions related to acute myocardial infarction, heart failure and stroke, 1994-2004. CMAJ 2009;180: E118–E125. 6 Gershon AS, Dolmage TE, Stephenson A, Jackson B. Chronic obstructive pulmonary disease and socioeconomic status: a systematic review. COPD 2012;9:216–226. 7 Tjepkema M, Wilkins R, Long A. Cause-specific mortality by education in Canada: a 16-year follow-up study. Health Rep 2012; 23:23–31. 8 Akinbami LJ, Liu X. Chronic obstructive pulmonary disease among adults aged 18 and over in the United States, 1998–2009. NCHS Data Brief 2011;63:1–8.
1201
ORIGINAL RESEARCH 9 Wilkins R, Berthelot J, Ng E. Trends in mortality by neighbourhood income in urban Canada from 1971 to 1996. Health Rep 2002;13:1–28. 10 Jemal A, Ward E, Anderson RN, Murray T, Thun MJ. Widening of socioeconomic inequalities in U.S. death rates, 1993-2001. PLoS ONE 2008;3:e2181. 11 Lipscombe LL, Austin PC, Manuel DG, Shah BR, Hux JE, Booth GL. Income-related differences in mortality among people with diabetes mellitus. CMAJ 2010;182:E1–E17. 12 Woolf SH. Public health implications of government spending reductions. JAMA 2011;305:1902–1903. 13 Levy DT, Chaloupka F, Gitchell J. The effects of tobacco control policies on smoking rates: a tobacco control scorecard. J Public Health Manag Pract 2004;10:338–353. 14 Gershon AS, Victor JC, Wilton A, To T. The impact of socioeconomic status on mortality in COPD: is the gap getting wider? [abstract]. Am J Respir Crit Care Med 2010;181:A6566. 15 Gershon AS, Wang C, Guan J, Vasilevska-Ristovska J, Cicutto L, To T. Identifying individuals with physcian diagnosed COPD in health administrative databases. COPD 2009;6:388–394. 16 Gershon AS, Warner L, Cascagnette P, Victor JC, To T. Lifetime risk of developing chronic obstructive pulmonary disease (COPD). Lancet 2011;378:991–996. 17 Wilkins R. PCCF1 version 4D user’s guide (Geocodes/PCCF). Automated geographic coding based on the Statistics Canada postal code conversion files, including postal codes to December 2003. Ottawa, Ontario, Canada: Statistics Canada; 2004. 18 Camilli AE, Robbins DR, Lebowitz MD. Death certificate reporting of confirmed airways obstructive disease. Am J Epidemiol 1991;133: 795–800. 19 Hansell AL, Walk JA, Soriano JB. What do chronic obstructive pulmonary disease patients die from? A multiple cause coding analysis. Eur Respir J 2003;22:809–814. 20 Anderson RN, Rosenberg HM. Age standardization of death rates: implementation of the year 2000 standard. Natl Vital Stat Rep 1998; 47:1–16, 20.
1202
21 Avendaño M, Kunst AE, van Lenthe F, Bos V, Costa G, Valkonen T, Cardano M, Harding S, Borgan JK, Glickman M, et al. Trends in socioeconomic disparities in stroke mortality in six european countries between 1981-1985 and 1991-1995. Am J Epidemiol 2005; 161:52–61. 22 Reid JL, Hammond D, Boudreau C, Fong GT, Siahpush M; ITC Collaboration. Socioeconomic disparities in quit intentions, quit attempts, and smoking abstinence among smokers in four western countries: findings from the International Tobacco Control Four Country Survey. Nicotine Tob Res 2010;12:S20–S33. 23 Woolf SH, Johnson RE, Phillips RL Jr, Philipsen M. Giving everyone the health of the educated: an examination of whether social change would save more lives than medical advances. Am J Public Health 2007;97:679–683. 24 Chapman KR, Tashkin DP, Pye DJ. Gender bias in the diagnosis of COPD. Chest 2001;119:1691–1695. 25 Buist AS, McBurnie MA, Vollmer WM, Gillespie S, Burney P, Mannino DM, Menezes AM, Sullivan SD, Lee TA, Weiss KB, et al.; BOLD Collaborative Research Group. International variation in the prevalence of COPD (the BOLD Study): a population-based prevalence study. Lancet 2007;370:741–750. 26 Chan BT, Austin PC. Patient, physician, and community factors affecting referrals to specialists in Ontario, Canada: a populationbased, multi-level modelling approach. Med Care 2003;41:500–511. 27 Soriano JB, Zielinski J, Price D. Screening for and early detection of chronic obstructive pulmonary disease. Lancet 2009;374:721–732. 28 Krieger N. Overcoming the absence of socioeconomic data in medical records: validation and application of a census-based methodology. Am J Public Health 1992;82:703–710. 29 Alter DA, Iron K, Austin PC, Naylor CD; SESAMI Study Group. Socioeconomic status, service patterns, and perceptions of care among survivors of acute myocardial infarction in Canada. JAMA 2004;291:1100–1107.
AnnalsATS Volume 11 Number 8 | October 2014
Department of Medicine, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada; 2Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada; 3Child Health Evaluative Sciences, The Hospital for Sick Children, Toronto, Ontario, Canada; and 4Department of Medicine, and 5The Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada
Abstract Rationale: Previous studies have demonstrated that people of higher socioeconomic status (SES) have better chronic obstructive pulmonary disease (COPD) health outcomes than those of lower SES. Mortality of people with COPD has decreased over the last decade; however, it is not known if all individuals with COPD have benefitted equally. Objective: The objective of the current study was to examine the impact of SES on mortality trends of individuals with COPD. Methods: We conducted a population-based study using health administrative data from Ontario, Canada, between 1996/7 and 2011/12. Individuals with COPD were identified using a previously validated case definition. SES was determined using average household income of the individual’s neighborhood as per the Canadian Census. Trends in standardized COPD mortality rates among different SES quintiles were observed over time.
Measurements and Main Results: Overall, age- and sexstandardized mortality of people with COPD decreased from 5.7% (95% confidence interval [CI], 5.5–5.8) in 1996/7 to 3.7% (95% CI, 3.6–3.8) in 2011/12, representing a 35% relative decrease. The mortality difference between the lowest and highest income quintiles increased from 67 per 10,000 individuals in 1996/7 to 86 per 10,000 individuals in 2011/12, representing a 28% relative increase (P , 0.001). Conclusions: Mortality in people with COPD has decreased faster in people with the highest compared with the lowest SES, causing increased disparity between rich and poor. Further study and strategies are needed to explore and address factors responsible for this increasing disparity in the COPD population. Keywords: pulmonary disease, chronic obstructive; mortality; income; social class; trends
(Received in original form March 5, 2014; accepted in final form July 19, 2014 ) This work was supported by funding from the Government of Ontario; by the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care; by a New Investigator Award funded by team grant OTG-88591 from the Canadian Institutes of Health Research Institute of Nutrition, Metabolism and Diabetes (A.S.G.); by the Physicians Services Incorporated Foundation Fellowship for Translational Health Research (A.S.G.); and by The Dales Award in Medical Research from the University of Toronto, Toronto, Ontario, Canada (T.T.). Author Contributions: A.G. participated in the study conception and design, acquisition, analysis and interpretation of data; drafting and critical revision of the manuscript; obtaining funding; supervising the writing; and approving the final version. J.H. and J.C.V. participated in the analysis and interpretation of data, participated in statistical analysis and critical revision of the manuscript, and approved the final version. A.S.W. participated in analysis and interpretation of data, critical revision of the manuscript, and approving the final version. T.T. participated in the conception and design, analysis and interpretation of data, critical revision of the manuscript, and approving the final version. Correspondence and requests for reprints should be addressed to Andrea Gershon, M.D., Institute for Clinical Evaluative Sciences, G1 06, 2075 Bayview Avenue, Toronto, ON, M4N 3M5 Canada. E-mail: [email protected] This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org Ann Am Thorac Soc Vol 11, No 8, pp 1195–1202, Oct 2014 Copyright © 2014 by the American Thoracic Society DOI: 10.1513/AnnalsATS.201403-094OC Internet address: www.atsjournals.org
Chronic obstructive pulmonary disease (COPD) is a manageable, progressive condition that places a large burden on individuals and health care systems. COPD is the third leading cause of death in North America, affects 8 to 22% of the adult
population, and is one of the most common reasons for hospitalization (1–3). Moreover, COPD prevalence is increasing, particularly in women, and this trend is likely to worsen as the population ages (2). On the positive side, overall mortality of
Gershon, Hwee, Victor, et al.: SES Mortality Trends in COPD
people with COPD is decreasing, likely due to declining smoking rates and better treatment of comorbidities like cardiovascular disease (2, 4, 5). This progress, however, has not been experienced by all individuals with COPD. 1195
ORIGINAL RESEARCH Previous studies have demonstrated an inverse association between socioeconomic status (SES) and various COPD outcomes, such as mortality (6–8). To the best of our knowledge, however, no studies have examined how this relationship has changed over time. In Canada before 1996, mortality differences between rich and poor populations were decreasing (9), but it is not known if this trend continued or if it applies to individuals with COPD. It does not appear to apply to the general population in the United States (10) or to people with other chronic diseases (e.g., diabetes), where socioeconomic disparity in mortality has been increasing (11). Government spending cuts on health care services, public health, and basic living conditions have led many to predict that differences in health between the rich and the poor will increase (12). For people with COPD, cuts might lead to increases in smoking rates and greater exposure to working and living conditions that worsen their disease (13). We conducted the current population study to test the hypothesis that observed socioeconomic mortality differences of individuals with COPD are increasing. Some of the results of this study have been previously reported in the form of an abstract (14).
Methods In this longitudinal, population cohort study of people with COPD, we used health administrative data from Ontario, Canada to compare observed real-world changes in mortality by SES from fiscal years (April– March) 1996/7 to 2011/12. Data Sources
Ontario is a province of Canada with a diverse multicultural population of approximately 13 million. Residents of Ontario have universal public health insurance under the Ontario Health Insurance Plan, the single payer for all medically necessary services across the full spectrum of residents, providers, and hospitals. Service details are captured in health administrative databases, which can be linked on an individual level to provide a complete health services profile for each resident. The only exception is the provision of prescription medications, which are only provided to those 65 years of age or older and to those who are on welfare. Three 1196
Ontario population health administrative databases were used. The Registered Persons Database contains basic demographic information and date of death (as applicable). Deaths recorded in this database are based on probabilistic linkage of death certificate information and are cross referenced with hospital records of in-hospital deaths. The Canadian Institute of Health Information Discharge Abstract database contains information on all discharges from acute care hospitals. The Ontario Health Insurance Plan Physician Claims database contains information on all services provided by fee-for-service physicians and “shadow-billings” for physicians paid under alternate payment plans. These data sets used in this study were held securely in a linked, deidentified form and analyzed at the Institute for Clinical Evaluative Sciences. Study Population
We identified individuals with COPD using the Ontario COPD Database, which identifies people with COPD using the above health administrative databases and a previously validated COPD case definition of one or more COPD physician billing claims and/or one or more COPD hospital discharges as per the following codes: 491, 492, 496 (Ontario Health Insurance Plan and International Classification of Disease, 9th revision codes) or J41, J42, J43, J44 (International Classification of Disease, 10th revision codes). Individuals have to be age 35 years of age or older when their claim(s) or discharge occurs. This definition was found to have a sensitivity of 85% and specificity of 78% when compared with clinical evaluation and has been well described elsewhere (2, 15, 16). Once patients are identified as having COPD, including people who recently immigrated, they remain in the database until they die or move out of the province. Exposure
SES was attributed using the subject’s postal code to determine the average household income of their neighborhood derived from the Canadian census and categorized into quintiles, with 1 representing the lowest and 5 the highest (17). We used the 1996 census for fiscal years 1996 to 1998, the 2001 census for fiscal years 1999 to 2003, and the 2006 census for fiscal years 2004 and onward. People whose postal code could not be linked to a neighborhood
income level were considered “missing” and were excluded from the primary analysis. Outcomes
Annual overall and income quintile specific all-cause mortality rates in people with COPD, expressed as the percentage of people with COPD who died in a given year, were calculated from fiscal years 1996/7 to 2011/12. All-cause mortality, instead of COPD mortality, was used because COPD has been found to be underestimated as a cause of death on vital statistics death records by around 50%, because COPD deaths are often attributed to its comorbidities like pneumonia and cardiovascular disease and because cause of death information was not available (18, 19). Analysis
Frequencies and proportions were used to describe the COPD population. Overall mortality rates by socioeconomic status quintile and the mortality rate income gradient (i.e., the absolute difference in rate between the lowest and highest income quintiles) and the mortality rate ratio (i.e., the rate in the lowest divided by the highest income quintile) were compared over fiscal years. To adjust for differences in population distribution over time, mortality rates were standardized to the 2001 COPD population (20). We created 35 to 64 and over 65 years age groups because Ontario residents 65 years and older are eligible for the provincial medication insurance program and therefore may be less likely to be affected by income. To explore whether mortality trends differed by age and sex, we calculated sex-standardized mortality by age group (35–64 yr and over 65 yr) and age-standardized mortality by sex for each income quintile. Ninety-five percent confidence intervals (CIs) were calculated using the g distribution. Mortality rate ratios were based on the age- and sexstandardized rates. Multivariable negative binomial regression was used to compare mortality by study year, socioeconomic status, age, and sex. A model was fit relating the variables of interest to number of deaths in a given year using the corresponding COPD population for that year as the offset. Two- and threeway interactions between socioeconomic status, age, sex, and year were used to determine if mortality trends differed by income and whether such differences differed between men and women or those
AnnalsATS Volume 11 Number 8 | October 2014
ORIGINAL RESEARCH Table 1. Ontario chronic obstructive pulmonary disease population demographics, 1996/7 and 2011/12 Demographic n Age, mean yr (SD) Female sex, % Urban* (vs. rural) residence, % Socioeconomic status quintile, % 1 (lowest) 2 3 4 5 (highest)
1996/7
2011/12
428,046 65.1 (14.2) 47.7 81.5
807,996 66.3 (13.9) 50.7 84.3
24.9 22.3 19.8 17.1 16.0
23.2 21.7 19.5 18.7 16.9
*Urban communities were identified using Statistics Canada’s definition, which consisted of a community with a population of more than 10,000 people.
age 35 to 64 years and those age 65 years and older for whom prescription drug costs are covered by provincial health insurance. Predicted mortality rates for age and sex subgroups were based on the binomial regression model.
Ethics Approval
Our study received ethics approval from the institutional review board at Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada. Conduct of the research without explicit consent from the participants was approved.
Sensitivity Analysis
The main analysis was repeated including people with missing SES data. Income quintiles for these individuals were imputed based on age, sex, and year.
increased to 807,996 in 2011/12, which represents an 88.8% increase (Table 1). SES could not be determined for 0.4% of individuals who were excluded from the analysis. Between 1996/7 and 2011/12, the COPD population became older, the number of female patients with COPD increased, and SES increased slightly so that in 2011/12 the COPD population had a mean age of 66.3 years, was slightly more female (50.7%), and had 23.2% in the lowest and 16.9% in the highest SES quintiles. Mortality among People with COPD
Overall, the age- and sex-standardized allcause mortality of the COPD population decreased from 5.7% (95% CI, 5.5–5.8) in 1996 to 3.7% (95% CI, 3.6–3.8) in 2011/12, representing a 35% relative decrease. This decrease was greater in men than in women (41% compared with 27%; P , 0.001) and was greater in individuals 35 to 64 years of age than in individuals 65 years of age and older (44% compared with 34%; P , 0.001).
Results Mortality by SES Quintile
In 1996/7 there were 428,046 people with COPD living in Ontario. This number
In all study years, we observed an inverse association between income quintile and
7
Standardized mortality rate (per 100 persons)
6
5
4
3
Socioeconomic status 1 (lowest) 2 3
2
4 5 (highest)
1
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year Figure 1. All-cause mortality rate by socioeconomic status quintile among with patients with chronic obstructive pulmonary disease from 1996/7 to 2011/ 12. Rates are age and sex standardized to the 2001 chronic obstructive pulmonary disease population.
Gershon, Hwee, Victor, et al.: SES Mortality Trends in COPD
1197
ORIGINAL RESEARCH mortality. Crude mortality rates by income and by age group are presented in Table E1 in the online supplement. Age- and sexstandardized mortality in all quintiles declined over time; however, the decline was less pronounced from 1996/7 to 2011/12 in the lowest compared with the highest income quintiles (30% compared with 38%; P , 0.001) (Figure 1). This resulted in an increase of 28% (P , 0.001) in the mortality rate difference between the lowest and highest income quintiles from 67 per 10,000 individuals in 1996/7 to 86 per 10,000 individuals in 2011/12. The mortality rate ratio (i.e., the ratio of the highest compared with the lowest income quintile) increased 11% from 1.18 in 2006/7 to 1.26 in 2011/12 (P , 0.001) (Figure 2). Age and Sex
A significant interaction involving all main factors (year, income, and age group [P , 0.001]; year and income [P = 0.023]; and year and sex [P , 0.001]) was identified. Therefore, when examining differences in income gradient over time, analyses were stratified by sex and age group.
The mortality difference between the lowest and highest income quintiles increased significantly over the study period among younger women (P , 0.001) (Figure 3A) and older men (P = 0.006) (Figure 4B) but not among older women (P = 0.08) (Figure 3B) and younger men (P = 0.139) (Figure 4A). Sensitivity Analysis
When a sensitivity analysis including the 0.4% of individuals for whom income quintile was missing was conducted, mortality rate estimates varied by a relative difference of, at most, 1.6%.
Discussion We conducted a longitudinal population study of individuals with COPD and found that, although observed overall mortality rates declined between 1996/7 and 2011/12, the difference between the lowest and highest income quintiles increased by 28%. This trend was most evident in younger women and older men and suggests that, even in a jurisdiction with universal health
care insurance, improvements in care for people with COPD disproportionately benefit the affluent. To the best of our knowledge, this is the first study to examine trends in socioeconomic-related health outcomes in people with COPD. Our findings are consistent with literature showing COPD mortality and other outcomes to be inversely associated with socioeconomic status (as measured by education, occupation, and income) at one point or period of time (6). Our study expands on these findings by examining whether this association has changed over time. Our study is also consistent with literature demonstrating increases in socioeconomic mortality differences for other chronic diseases, such as diabetes and stroke (11, 21). Our results also may be, at least in part, reflective of socioeconomic mortality differences in the general population, although we are not aware of any studies that have documented this. There might be several explanations for our findings. Government funding cuts on health care services, public health, and basic living conditions have decreased
1.4
Socioeconomic status 1 (lowest) vs. 5 (highest)
Mortality rate ratio
1.3
1.2
1.1
1 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year Figure 2. Mortality rate ratio of the lowest socioeconomic status quintile compared with the highest quintile among Ontarians with chronic obstructive pulmonary disease from 1996/7 to 2011/12.
1198
AnnalsATS Volume 11 Number 8 | October 2014
ORIGINAL RESEARCH
A 1.4
1.2
Mortality rate, %
1
0.8
0.6
Socioeconomic status 1 (lowest) 2
0.4 3 4 0.2
5 (highest)
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
B
9
8
7
Mortality rate, %
6
5 Socioeconomic status 4
1 (lowest) 2
3
3 4
2 5 (highest) 1
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year Figure 3. Predicted mortality by socioeconomic status quintile among women with chronic obstructive pulmonary disease 35 to 64 years of age (A) and women 65 years of age or older (B) from 1996/7 to 2011/12.
Gershon, Hwee, Victor, et al.: SES Mortality Trends in COPD
1199
ORIGINAL RESEARCH
A 2.5
Mortality rate, %
2
1.5
1
Socioeconomic status 1 (lowest) 2 3
0.5
4 5 (highest)
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year
B 12
10
Mortality rate, %
8
6
Socioeconomic status 1 (lowest) 2
4
3 4 5 (highest)
2
0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Year Figure 4. Predicted mortality by socioeconomic status quintile among men with chronic obstructive pulmonary disease 35 to 64 years of age (A) and men 65 years of age or older (B) from 1996/7 to 2011/12.
1200
AnnalsATS Volume 11 Number 8 | October 2014
ORIGINAL RESEARCH opportunities for education, employment, food security, and stable neighborhoods and have had adverse effects on health that disproportionately affect those of lower SES (12). For people with COPD, this might cause psychological stress and lead to higher rates of smoking and/or fewer attempts to quit (22). Funding cuts might also result in fewer smoking cessation and prevention programs. People of lower SES with COPD may also have occupations that contribute to their disease in that they have less flexibility to extricate themselves. Similarly, they may live in areas where they are exposed to more air pollution. Treatments of chronic disease, including COPD, have become more complicated, involving medications that might be difficult for those with less education, less time, and less money to follow and maintain. Even successfully quitting smoking has become complicated, with drug costs that are not always subsidized by insurance. Finally, because people with COPD often die of comorbidities related to their COPD and not of COPD itself, differences in the management of other chronic disease (e.g., cardiovascular disease) in patients of different socioeconomic strata might contribute to the results. Although these changes would be unlikely to cause dramatic year-to-year fluctuations over an entire population, it is feasible that they could lead to a steady gradual decline in health over time. Inequalities in health are a central challenge for public health. Eliminating the disparity could potentially reduce disease burden, need for health services, and costs more than newly developed medications (23). These findings, however, suggest that the disparity is increasing. This is of particular importance given the higher prevalence of COPD in lower-income
populations (6). This trend is likely more pronounced in nations without universal health care coverage, especially developing countries. Our study points to the need to address barriers to prevention and management of COPD in low socioeconomic populations. We found that increasing disparity appeared most pronounced in younger women and older men. Higher disparity in older men could be due to higher rates of cardiovascular disease, for which COPD is a risk factor, making this population more vulnerable to socioeconomic influences on their health. Higher disparity in younger women might be because they put the care of their children and family responsibilities before their own health, causing them to be more susceptible to socioeconomic pressures. Women are also more likely to be underdiagnosed compared with men, and this might be compounded by less access to physicians among populations of lower SES (24). This is an important area for future study. A strength of our study is its ability to measure trends in COPD mortality according to SES in a large, complete population. Our study also had several limitations. First, our case definition of COPD might have been subject to some misclassification error (15). Any error is likely consistent over socioeconomic strata and time and therefore unlikely to bias trends; however, one could argue that the increasing awareness of COPD that was more pronounced in people of higher socioeconomic status (SES) led to more patients with milder disease being included in this group and a perceived lower mortality rate. However, we think this scenario is unlikely because previous studies have shown that people of lower, not higher, SES are more likely to be identified with COPD (6). Second, we were
References 1 Murphy S, Xu J, Kochanek K. Deaths: preliminary data for 2010. Natl Vital Stat Rep 2012;60:1–51. 2 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;170:560–565. 3 Benady S. The human and economic burden of COPD: a leading cause of hospital admission in Canada. Ottawa, Ontario, Canada: The Canadian Thoracic Society; 2010. 4 Gilmore J. Report on smoking in Canada 1985 to 2001. 2002. Ottawa, Ontario, Canada: Minister of Industry;2002.
Gershon, Hwee, Victor, et al.: SES Mortality Trends in COPD
only able to study diagnosed COPD, and it is known that there are many people with undiagnosed COPD in the population (25). Rates of underdiagnosis may be higher in lower socioeconomic groups because of less access to physicians (26). If this led to delayed diagnosis and, as a result, more advanced disease in this group, their mortality could appear higher. Nonetheless, as physicians have become more aware of undiagnosed COPD in recent years, it is unlikely to have gotten worse over time and to account for the trends observed (27). Third, we did not have income data at the individual level. However, this ecological measure has been shown to correlate well with individual-level measures, can account for other neighborhood factors that contribute to SES, and has proven to be robust in many population-based, peerreview studies (28, 29). Most importantly, any misrepresentation there might be of SES would be unlikely to change over time. Fourth, this study took place in Canada, where universal health care insurance that covers all medically necessary services has improved health care access for poorer Canadians; therefore, this study might underestimate socioeconomic gradients and trends in other jurisdictions without such coverage. Finally, we did not have information on cause of death or smoking so were unable to compare differences in COPD-specific mortality or smoking rates due to SES over time. In summary, socioeconomic disparity in mortality in people with COPD appears to be getting worse. More study is needed to determine the factors responsible for this trend and how they can be addressed to improve the health of the population. n Author disclosures are available with the text of this article at www.atsjournals.org.
5 Tu JV, Nardi L, Fang J, Liu J, Khalid L, Johansen H; Canadian Cardiovascular Outcomes Research Team. National trends in rates of death and hospital admissions related to acute myocardial infarction, heart failure and stroke, 1994-2004. CMAJ 2009;180: E118–E125. 6 Gershon AS, Dolmage TE, Stephenson A, Jackson B. Chronic obstructive pulmonary disease and socioeconomic status: a systematic review. COPD 2012;9:216–226. 7 Tjepkema M, Wilkins R, Long A. Cause-specific mortality by education in Canada: a 16-year follow-up study. Health Rep 2012; 23:23–31. 8 Akinbami LJ, Liu X. Chronic obstructive pulmonary disease among adults aged 18 and over in the United States, 1998–2009. NCHS Data Brief 2011;63:1–8.
1201
ORIGINAL RESEARCH 9 Wilkins R, Berthelot J, Ng E. Trends in mortality by neighbourhood income in urban Canada from 1971 to 1996. Health Rep 2002;13:1–28. 10 Jemal A, Ward E, Anderson RN, Murray T, Thun MJ. Widening of socioeconomic inequalities in U.S. death rates, 1993-2001. PLoS ONE 2008;3:e2181. 11 Lipscombe LL, Austin PC, Manuel DG, Shah BR, Hux JE, Booth GL. Income-related differences in mortality among people with diabetes mellitus. CMAJ 2010;182:E1–E17. 12 Woolf SH. Public health implications of government spending reductions. JAMA 2011;305:1902–1903. 13 Levy DT, Chaloupka F, Gitchell J. The effects of tobacco control policies on smoking rates: a tobacco control scorecard. J Public Health Manag Pract 2004;10:338–353. 14 Gershon AS, Victor JC, Wilton A, To T. The impact of socioeconomic status on mortality in COPD: is the gap getting wider? [abstract]. Am J Respir Crit Care Med 2010;181:A6566. 15 Gershon AS, Wang C, Guan J, Vasilevska-Ristovska J, Cicutto L, To T. Identifying individuals with physcian diagnosed COPD in health administrative databases. COPD 2009;6:388–394. 16 Gershon AS, Warner L, Cascagnette P, Victor JC, To T. Lifetime risk of developing chronic obstructive pulmonary disease (COPD). Lancet 2011;378:991–996. 17 Wilkins R. PCCF1 version 4D user’s guide (Geocodes/PCCF). Automated geographic coding based on the Statistics Canada postal code conversion files, including postal codes to December 2003. Ottawa, Ontario, Canada: Statistics Canada; 2004. 18 Camilli AE, Robbins DR, Lebowitz MD. Death certificate reporting of confirmed airways obstructive disease. Am J Epidemiol 1991;133: 795–800. 19 Hansell AL, Walk JA, Soriano JB. What do chronic obstructive pulmonary disease patients die from? A multiple cause coding analysis. Eur Respir J 2003;22:809–814. 20 Anderson RN, Rosenberg HM. Age standardization of death rates: implementation of the year 2000 standard. Natl Vital Stat Rep 1998; 47:1–16, 20.
1202
21 Avendaño M, Kunst AE, van Lenthe F, Bos V, Costa G, Valkonen T, Cardano M, Harding S, Borgan JK, Glickman M, et al. Trends in socioeconomic disparities in stroke mortality in six european countries between 1981-1985 and 1991-1995. Am J Epidemiol 2005; 161:52–61. 22 Reid JL, Hammond D, Boudreau C, Fong GT, Siahpush M; ITC Collaboration. Socioeconomic disparities in quit intentions, quit attempts, and smoking abstinence among smokers in four western countries: findings from the International Tobacco Control Four Country Survey. Nicotine Tob Res 2010;12:S20–S33. 23 Woolf SH, Johnson RE, Phillips RL Jr, Philipsen M. Giving everyone the health of the educated: an examination of whether social change would save more lives than medical advances. Am J Public Health 2007;97:679–683. 24 Chapman KR, Tashkin DP, Pye DJ. Gender bias in the diagnosis of COPD. Chest 2001;119:1691–1695. 25 Buist AS, McBurnie MA, Vollmer WM, Gillespie S, Burney P, Mannino DM, Menezes AM, Sullivan SD, Lee TA, Weiss KB, et al.; BOLD Collaborative Research Group. International variation in the prevalence of COPD (the BOLD Study): a population-based prevalence study. Lancet 2007;370:741–750. 26 Chan BT, Austin PC. Patient, physician, and community factors affecting referrals to specialists in Ontario, Canada: a populationbased, multi-level modelling approach. Med Care 2003;41:500–511. 27 Soriano JB, Zielinski J, Price D. Screening for and early detection of chronic obstructive pulmonary disease. Lancet 2009;374:721–732. 28 Krieger N. Overcoming the absence of socioeconomic data in medical records: validation and application of a census-based methodology. Am J Public Health 1992;82:703–710. 29 Alter DA, Iron K, Austin PC, Naylor CD; SESAMI Study Group. Socioeconomic status, service patterns, and perceptions of care among survivors of acute myocardial infarction in Canada. JAMA 2004;291:1100–1107.
AnnalsATS Volume 11 Number 8 | October 2014
