Disability pension after coronary revascularization: a prospective nationwide register-based Swedish cohort study

Abstract

Background

Scientific knowledge on disability pension (DP) after revascularization by coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI) is scarce. The aim was to study the incidence of and risk factors for being granted DP in the 5 years following a first CABG or PCI, accounting for socio-demographic and medical factors.

Methods

This is a nationwide population-based study using Swedish registers including all patients 30–63 years of age (n = 34,643, 16.4% women) who had a first CABG (n = 14,107) or PCI (n = 20,536) during 1994–2003. All were alive and without reintervention 30 days after the procedure and were not on DP or old-age pension. Multivariable adjusted Cox proportional hazard ratios (HR) for DP were estimated with 95% confidence intervals (CI).

Results

In 5 years following revascularization, 32.4% had been granted DP and the hazard ratio (HR) was higher in women (HR 1.55, 95% CI 1.48–1.62), and in CABG patients compared with PCI patients (HR 1.35, 95% CI 1.30–1.40). Long-term sick leave in the year before intervention was the strongest predictor for DP following revascularization. After adjustments for socio-demographic factors and sick-leave days in the 12 months before revascularization, HR remained high in all patients with diabetes mellitus regardless of type of revascularization.

Conclusions

DP after coronary revascularization was common, especially among women and CABG patients. Most studied medical covariates, including mental and musculoskeletal disorders, were risk factors for future DP, especially long-term sickness absence.

Introduction

Cardiovascular disease (CVD) represents a major public health problem that causes impaired physical capacity, work disability, and mortality.¹ After mental and musculoskeletal diagnoses, CVD is also a major diagnostic group for disability pension (DP).^2,3

Two well-documented and often-used effective treatments for acute coronary syndromes as well as stable symptomatic coronary heart disease are coronary revascularization by coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI), resulting in symptom reduction, improved physical performance, and, in some patient groups, reduced mortality.⁴ Mortality, morbidity, and other medical outcomes following coronary revascularization are well documented.^4–10 However, there is hardly any scientific knowledge regarding DP following coronary revascularization in general nor in subgroups, despite major consequences of DP for patients as well as the society.^2,11–13

We only found one systematic literature review regarding studies about DP and return to work following coronary revascularization.³ Few studies were included, many of them with methodological limitations. The review established limited evidence that most patients return to work following CABG or PCI. Nevertheless, many patients are sickness absent or leave the labour market permanently due to coronary revascularization (e.g. through DP).¹⁴ To our knowledge, there are no previous studies with the specific aim to investigate DP following CABG or PCI. However, some results regarding DP following CABG are published.^15–18 All were questionnaire- or interview-based follow-up studies, with sample sizes from 141 up to 2047 individuals. In one study, 25% of the patients who worked 1 year after CABG and were younger than 60 years old at the 10-year follow up, were granted DP.¹⁶ In another study, 17% of 141 postoperatively working patients were granted DP before the age of 60.¹⁵ None of these studies presented results by gender, although it is well known that more women than men generally are granted DP.^19–27 Other reasons for conducting gender-specific analyses are that women generally require coronary intervention about 10 years later than men,^28,29 implying higher risks of comorbidity and DP^6,30,31 and that women have smaller coronary arteries than men which may be related to worse prognosis for CABG.⁶ In general, people with less education, immigrants, and people with long-term sick leave have higher risks of DP.³²

The aim was to study the incidence of and risk factors for future DP following a first coronary revascularization by CABG or PCI among women and men at working ages, taking into account socio-demographic and medical factors.

Methods

A population-based, prospective cohort study based on register data was conducted. Included were all patients in Sweden (n = 34,643, 16.4% women) 30–63 years of age, who during 1994–2003 had a first coronary revascularization by CABG or PCI. They were alive without reintervention within 30 days after the procedure and were not on DP or old-age pension.

The patients were identified from the nationwide quality register SWEDEHEART (Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies),³³ including information of all coronary revascularizations conducted in Sweden regarding date and type of procedure and patient variables.

For each patient, annual data for the years 1993–2007 was obtained from the Swedish Social Insurance Agency (date of granted DP, numbers of sick-leave days, and old-age pension), Statistics Sweden (educational level, type of living area, marital status, country of birth, emigration), and the National Board of Health and Welfare’s National Patient Register (inpatient care for cardiovascular or other major conditions) and Cause of Death Register (date of death).

The study was approved by the Regional Ethical Review Board of Stockholm (2006/661–31).

Outcome

The outcome variable was being granted DP, within 5 years after a first CABG or PCI.

Categorization of covariates

The following covariates were categorized as follows. Age at intervention: 30–49, 50–54, 55–59, and 60–63 years. Level of education: elementary school, high school/college, and university. Type of living area: six homogeneous regions,³⁴ which in the regression analyses were merged into larger cities (Stockholm, Gothenburg, Malmö), middle-sized cities, and smaller communities. Marital status: unmarried (unmarried, divorced, or widow/widower) and married. Country of birth: Sweden, other Nordic countries, EU25 (except the Nordic countries), and other countries, which were dichotomized into Sweden vs. other countries in the regression analyses.

The indication for intervention was categorized according to International Classification of Diseases version 10 (ICD-10)³⁵ into acute coronary syndrome (ACS) with non-ST-elevation myocardial infarction or rarely ST-elevation myocardial infarction or stable angina. If data was lacking in the quality registers, the National Patient Register was used. Year of intervention was categorized into 1994–1996, 1997–2000, and 2001–2003. Number of reinterventions within 5 years were categorized as none or at least one. For diabetes mellitus, yes/no data were obtained from the Quality Registers and the Patient Register.

Inpatient care in the 5 years before CABG/PCI was categorized according to ICD-10 into angina pectoris (I20), myocardial infarction (I21–I22), heart failure (I50), cerebrovascular disease (I60–I69), transient cerebral ischaemic attacks (G45), mental disorders (F00–F99), and musculoskeletal disorders (M00–M99). In the regression analyses, the following three groups were used: CVD (angina pectoris, myocardial infarction, heart failure, cerebrovascular disease, and transient cerebral ischaemic attacks), mental, and musculoskeletal disorders.

The number of sick-leave days in the 12 months before CABG/PCI was categorized into 0, 1–89, 90–179, and 180–365 days. In the regression analyses, the first two groups were combined.

Statistical analyses

Baseline characteristics and the incidence of being granted DP within 5 years after coronary revascularization were presented as proportions. Kaplan–Meier estimate was applied to describe proportions of DP over the 5 years following coronary revascularization (Supplementary Online Appendix). Cox proportional hazard model was used to calculate hazard ratios (HR) with 95% confidence intervals (CI) for being granted DP with regards to independent variables. The selection of possible socio-demographic confounding factors was based on the crude analyses, including the factors with a 95% CI either above or below 1 (for women or men). All patients contributed with person time (days at risk) until date of being granted DP, old-age pension, emigration, death, or the date of 5 years after inclusion, if the latter occurred before 31 December 2007, which was the last day of follow up. The data was analysed separately for women and men and for CABG and PCI.

Social insurance in Sweden

In Sweden, all adults below the age of 65 can be granted DP if disease or injury has led to permanent work incapacity. All with income from work or unemployment benefits can get sickness benefits if work capacity is reduced due to disease or injury. The employment frequency was high in the studied years, among women and men and also in higher ages.

Results

In the study cohort (n = 34,643), a first intervention of PCI (n = 20,536; 18.8% women) was more common than CABG (n = 14,107; 12.9% women) (Table 1). Mean age at intervention was 54 years and a majority of the patients lived in medium-sized cities (51.6%), were married (65.0%), were born in Sweden (84.3%), and did not have university education (79.7%). Stable angina pectoris was the main indication for the coronary revascularization in 55% of both the CABG and the PCI patients. Reintervention within 5 years following the primary intervention occurred in 5.0% in CABG and 23.9% in PCI patients. Diabetes mellitus was more common among CABG (13.0%) than PCI patients (6.1%). Seventy per cent of CABG patients had been hospitalized for angina pectoris and about a third had myocardial infarction in the 5 years before the intervention; the corresponding figures in PCI patients were 37.5 and 27.6%, respectively. Among CABG patients, 15.7% had 90–179 sick-leave days in the 12 months before the intervention and 18.8% had 180–365 sick-leave days; the corresponding figures in PCI patients were 10.8 and 13.1%, respectively. During follow up, a total of 23.5% were granted old-age pension, 0.8% emigrated, and 3.7% died (results not shown).

More women (42.4%) than men (30.4%) were granted DP within 5 years after the first coronary revascularization, regardless of type of intervention (Table 2). Adjusting for age, HR for women compared to men was 1.54 (95% CI 1.47–1.61). Patients with CABG had a higher risk of DP compared to PCI patients (age-adjusted HR 1.29, 95% CI 1.24–1.34) (Table 2).

Reintervention, diabetes mellitus, and a history of inpatient care for CVD (in men) for mental or for musculoskeletal disorders, were associated with a higher crude HR of DP. Moreover, a history of ≥180 sick-leave days in the year before revascularization increased the risk of DP by up to 6-fold (men, PCI HR 6.25, 95% CI 5.84–6.68) (Supplementary Online Appendix).

When adjusting for age, education, country of birth, and year of intervention (model I), the HR almost remained unchanged compared to the crude analyses (Table 3).

After final adjustments for sick-leave days in the 12 months before revascularization (model II), the high risk of DP remained in patients with diabetes mellitus and was the highest in men with PCI (HR 1.61, 95% CI 1.45–1.78). The high HR also remained in patients (except women with CABG) with one or more reinterventions within 5 years, with inpatient care for mental disorders, and in men with inpatient care for musculoskeletal disorders. The HR of DP decreased in patients with an indication of ACS and in PCI patients with a history of CVD.

Discussion

In this nationwide cohort study of all patients at working age (n = 34,643) who had a first coronary revascularization by either CABG or PCI in 1994–2003, about one-third were granted DP within the following 5 years.

Regardless of type of intervention, women had a 55% increased risk of DP compared to men. To our knowledge, this is the first gender-specific study of DP following coronary revascularization. However, several other studies of gender differences confirm the higher incidence of DP in women, both generally and regarding DP following musculoskeletal and mental disorders.^{11,19,21–27,36,37} We also conducted an intervention-specific analyses, as patients referred to CABG often have more extensive coronary heart disease than PCI patients,^4,8,9,27 to investigate why the risk of DP might be higher. This was confirmed by our results: DP was more common in patients with CABG than with PCI. The strongest risk factor for future DP was sickness absence ≥6 of the 12 months before the intervention. Men with PCI with such sickness absence had a 6-fold increased HR of DP. This is in agreement with previous studies.^17,20 However, even in those with very long sick leave before intervention, more than 20% were not granted DP. The risk of DP was also higher among those with low level of education, which is in line with previous findings where higher education had a protective effect on DP.^23,38 Men born outside Sweden had a significant higher risk of DP, which was confirmed by Beckman et al.,³⁹ claiming that the birth country is important for understanding the socioeconomic differences that relates to DP. Thus, we treated these three covariates as confounders in the adjusted models. After final adjustments (Table 3, model II), diabetes mellitus showed a higher risk for DP regardless of gender and type of intervention. This was also the case for patients with a history of recent inpatient care for mental or musculoskeletal disorders, which is in line with previous studies.^25,27,40,41 Our results also show that the indication of ACS for intervention, as well as previous inpatient care for cardiovascular disease, did not increase the risk of DP. These findings indicate the important role of comorbidity in relation to DP after coronary revascularization, needing further investigation.

Strengths and limitations

The main strengths are the population-based and prospective study design, the large cohort (n = 34,643), that all patients at risk of DP, not a sample, were included, that there was no drop-out or loss to follow up, the high quality of data from several different registers including extensive information on medical and socio-demographic factors,⁴² and the long follow-up period. The risk of DP might have been underestimated in the patients who had their first revascularization in 2003, as they had a somewhat shorter follow-up time. Another possible reason for underestimation might be that in the late 1990s and early 2000s, it could take long time before the Social Insurance Agency assessed whether long-term sickness absentees should be granted DP. This means that some were sickness absent for years before granted DP.²⁷ However, the 5-year follow up should have captured those DPs.

Conclusions

This is the first nationwide register-based cohort study investigating DP after coronary revascularization by CABG or PCI. We found a high incidence of DP within 5 years following coronary revascularization; higher among women and patients with CABG. A history of long-term sickness absence prior to the intervention, as well as reintervention, diabetes mellitus, and inpatient care for mental and musculoskeletal disorders, independently increased the risk of DP. This indicates that the patients’ medical history, beyond CVD, is of importance for DP after coronary revascularization.

Funding

This work was supported by the Research Council of Working Life and Social Research (2007–0728 + 1762) and the National School of Research in Caring Science, Karolinska Institutet.

Conflict of interest

The authors declare that there is no conflict of interest.

Previous presentations

Some results were presented orally at the 4th European Public Health Conference, Copenhagen, Denmark, 10–12 November 2011.

References