Education-related variation in coronary procedure rates and the contribution of private health care in Australia: a prospective cohort study

Ischaemic heart disease (IHD) is the leading cause of mortality and morbidity globally and in Australia [1, 2] and the burden of IHD is higher in more socioeconomically disadvantaged groups [3‐5]. The prevalence of biomedical and behavioural risk factors, and high absolute cardiovascular disease (CVD) risk is disproportionately high in people of lower socioeconomic position [6], as is incidence of the disease [7]. People of lower socioeconomic position are also more likely to have a major secondary cardiovascular disease (CVD) event [7] and are more likely to die from the disease [8]. Understanding the reasons for these differences is crucial to improving outcomes for more disadvantaged groups and lowering the overall burden of the disease.

Socioeconomic gradients in CVD events and mortality are likely to reflect, at least in part, differences in uptake and continuation with secondary prevention strategies, including coronary procedures [9, 10]. Angiography and revascularisation procedures, specifically percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG), are recommended for the management of IHD, including acute myocardial infarction (AMI) and unstable angina [11]. Angiography is performed as a diagnostic procedure and, where clinically indicated, revascularisation procedures are used to improve blood flow to the heart [11, 12].

International and Australian studies have shown lower rates of coronary procedures among more disadvantaged populations [10, 13‐15]. Recent findings suggest that these socioeconomic disparities may be disappearing over time as the overall use of coronary procedures has increased [16‐20]. However, these studies have relied on area-level measures to assess socioeconomic position [13, 15, 16, 21, 22], which generally underestimate variation compared to individual-level measures [10, 23, 24]. Furthermore, many studies use data from public hospitals only, also potentially underestimating variation within the population, and providing no insights into the role of private health care in explaining inequalities in care. Australia has a hybrid health care model in which care is provided under the public system but people can opt to pay for additional private insurance [25]. The public component of the health care system, Medicare, covers the cost of public hospital admissions and some other health costs. Optional private health insurance provides access to private hospitals and covers costs of other health services not covered by Medicare. Private health insurance can reduce wait times for procedures; median wait times for private patients are around half that of public patients [26]. Around 43% of the Australian population had private health insurance in 2006 [26]. In addition, a small proportion of the population have access to private health care covered under the Department of Veterans’ Affairs Scheme. Contemporary Australian data using individual-level socioeconomic position and procedure data from both public and private hospitals are needed to provide evidence to inform clinical practice and the equitable use of coronary procedures following IHD events.

The aim of this study was to quantify the relation of education, a measure of individual-level socioeconomic position, to receipt of coronary procedures among patients admitted to public and private hospitals for AMI or angina. We also examined whether private health insurance explained any observed associations.

After applying study exclusions (Fig. 1), the AMI and angina samples consisted of 4454 and 4348 participants, respectively. Sample characteristics are in Table 1 (AMI patients) and Table 2 (angina patents). Mean age was 73.2 years (standard deviation [SD] = 11.6) for the AMI patients and 70.3 years (SD = 10.3) for the angina patients. In the AMI sample, 17.4% had low, 67.5% had intermediate and 15.1% had high education; corresponding proportions for the angina sample were 14.5%, 67.6% and 17.9%. Participants with lower levels of education were older, more likely to live outside major cities, had more functional limitations and comorbidities and were less likely to have private health insurance.

In this large-scale Australian prospective cohort study, there was a socioeconomic gradient in coronary procedure rates following hospital admission for AMI or angina. The likelihood of receiving a procedure decreased with increasing disadvantage, such that those with the lowest compared to the highest education were around 30% less likely to receive an angiogram, and 30% less likely to receive a revascularisation procedure if admitted with AMI, and 50% less likely if admitted with angina. Broad consideration of health (BMI, physical functioning and hospital-recorded comorbidities) did not explain these differences, except for the differences in revascularisation rates among AMI patients. Private health insurance accounted for some but not all of the education-related variation in procedure rates. Our study examined socioeconomic differences in procedure rates averaged over the study period. Although we did not find evidence of inequality in revascularisation among AMI patients, inequalities may have decreased over time, in line with diffusion of innovation theory [16]. However, our study did not aim to look at changes in inequalities over time and was insufficiently powered to do this.

Previous Australian studies have reported mixed findings regarding socioeconomic variation in the receipt of angiography, ranging from no evidence of an association [16, 21], to statistically non-significant findings in the same direction as those observed in our study [22]. These studies used area- rather than individual-level data and were restricted to public hospital data only, which may have underestimated socioeconomic variation in procedure rates. Our findings are similar to those reported in international studies using area- and individual-level measures, which have found that low socioeconomic position is associated with delays in receiving angiography [36] and lower angiography procedure rates [20, 37, 38]. Our finding that socioeconomic differences in revascularisation rates among those with angina were substantially attenuated after accounting for angiography is also consistent with previous studies. A Norwegian study (2001–2009) found that socioeconomic variation in revascularisation rates could be fully explained by differences in the receipt of angiography [20]. Socioeconomic differences are most apparent for diagnostic procedures and tend to diminish with subsequent IHD treatment such as revascularisation [10]. Taken together, these findings suggest that, among angina patients where coronary procedures are less protocol-driven than for AMI patients [39], socioeconomic variation is more evident during diagnostic stages (angiogram) but, once investigated, patients receive the same level of revascularisation regardless of their socioeconomic position.

Private health insurance is one possible mechanism through which socioeconomic inequalities in provision of care may arise in Australia. In our study, people with higher education were more likely to hold private health insurance than those with less education, and consistent with earlier Australian studies, [15, 40] privately insured patients were more likely to receive angiography. Remuneration for angiography to health care providers within the private health sector in Australia is fee-for-service and therefore is based on the quantity, not quality, of care provided [40], and the ratio of angiography to revascularisation is greater within the private health care system compared to the public health sector [41]. This suggests possible over-utilisation among privately insured patients, but it may reflect under-utilisation among public patients. This relative over- or under-utilisation could also reflect differing patient and clinician expectations [42‐45].

The observed socioeconomic gradient in procedure rates may also reflect differences in clinical need for angiography. Current Australian guidelines do not recommend angiography for all AMI and angina patients, rather, indication depends on disease severity, risk of future events and other clinical characteristics. For example, patients with non-ST-elevation MI are indicated for angiography only if they are at immediate or high risk of a future cardiovascular event [11] and it is generally indicated in patients with stable angina only when symptoms cannot be managed with medication [46, 47]. A number of clinical characteristics, such as the presence of major comorbidities, are contraindications for angiography. Although we adjusted for related factors, we cannot exclude the possibility that the low education group within our study sample comprised a greater proportion of patients in whom angiography was not indicated.

The use of individual-level data on education to measure patients’ socioeconomic position was a major strength of this study. Together with the inclusion of data from both public and private hospitals, and information on private health insurance, this enabled possibly more accurate estimates of socioeconomic differences in coronary procedure rates compared to previous Australian studies. Linkage of questionnaire data to hospital administrative and death records allowed for virtually complete follow-up of patients over time and provided more detailed information on relevant variables and confounders of interest than administrative data alone. However, data on certain clinical characteristics of the patients were not available, including the disease severity at time of admission and risk of a future cardiovascular event. Data did not allow distinction between subtypes of AMI and angina. As such, it was not possible to assess whether care reflected guideline recommendations. We adjusted for comorbidities, BMI and physical functioning in an attempt to account for contraindications to coronary procedures, but a degree of residual confounding is likely. Although education is just one measure of socioeconomic position, it is a reliable measure at the individual-level and is relatively stable over time, although the relative value of education varies across age cohorts. Finally, procedure rates are not necessarily generalisable because the study cohort is not wholly representative of the general NSW population, however relative measures of association from cohort studies are broadly applicable to the general population [27, 48].

Socioeconomic variation in the provision of health care can signal potential points of intervention to improve outcomes for disadvantaged groups. Using linked data from public and private hospitals and an individual-level measure of education, this study has shown that there is a socioeconomic gradient in the receipt of angiography following an AMI or angina admission, with the most disadvantaged people less likely to receive an angiogram compared to those with the least disadvantage. Some of this variation was explained by differences in private health insurance between socioeconomic groups. Other reasons for lower rates of angiography, including the extent to which the variation is clinically appropriate or reflects under treatment of those from low socioeconomic position or over treatment of those from a high socioeconomic position, remain unclear.