Socioeconomic gradients in admission to coronary or intensive care units among Australians presenting with non-traumatic chest pain in emergency departments

Socioeconomic inequalities in cardiovascular morbidity and mortality have been reported in many regions including the US [1], the UK [2], Australia [3] and other Organisation for Economic Cooperation and Development (OECD) countries [4]. Cardiovascular morbidity measures such as admission rates for cardiac related conditions have generally dropped over the past two decades [5]. However, relative inequality in cardiovascular emergency admissions and cardiovascular related mortality actually increased in the most disadvantaged compared with the least disadvantaged [6]. These disparities have been attributed to a range of socioeconomic determinants of health and health behaviours, rooted in social rank as determined by education, occupational hierarchy and income [1, 4, 7, 8]. The relationship between socioeconomic status and general health, and, in particular, cardiovascular health, has been demonstrated within different races and ethnic groups [9], suggesting that cultural and ethnic dissimilarities do not explain the socioeconomic differences. These associations were also observed in countries with universal access to health care [2, 3], and when comparing those who have similar rates of smoking, obesity and alcohol use [9].

Despite the mounting evidence of direct associations between socioeconomic disadvantage and poorer health outcomes, disagreement remains regarding the strength of the evidence to support causality [10‐14] with reverse causation proposed as a contributor to the relationship. For example, illness may lead to lower academic achievement or loss in income [10]. Studies investigating these relationships have predominantly followed a cross-sectional design and have not been population-based limiting their generalisability [13]. Furthermore, no study found dose-response associations between lower SES and increased cardiovascular morbidity.

Chest pain is a frequently seen symptom in emergency departments, being the most common reason for presenting in the ED among Americans over the age of 65 and contributing to approximately 6 million visits per annum under the US Medicare system [15]. Clinical management of chest pain is highly variable, often depending on the underlying causes and is considered a medical emergency until all life-threatening causes have been ruled out. Potentially life-threatening causes of chest pain include acute myocardial infarction and other acute coronary syndromes as well as aortic dissection, pneumothorax, pneumonia and pulmonary embolism. Patients presenting with chest pain who require critical care are often more likely to be severely ill and / or are at risk of imminent death [16, 17]. This study used emergency admissions following repeated non-traumatic chest pain presentations to emergency departments (ED) to explore the associations between sex, age, geographic-based socioeconomic disadvantage score and cardiovascular morbidity, expressed as admission to specialised care units.

In this large population-based panel analysis that investigated emergency non-traumatic chest pain admissions, geographic-based socioeconomic disadvantage scores were independently associated with increased admissions to a coronary or intensive care unit, showing a dose-response. These associations were independent of the patient’s sex, age, ethnicity, main presenting symptoms, hospital setting, urgency of presentation, length of stay in the ED, and main acute admission diagnosis. Females compared to males of the same age were less likely to be admitted to such units.

The findings of this study suggest increased chest pain severity or increased general and, in particular, cardiovascular morbidity among those coming from low SES locations. However, although this study reports dose-response effect between socioeconomic disadvantage and admission to intensive or coronary care units, and despite the mounting and consistent findings of direct associations [4, 8, 9, 11], this study was observational, and a causal relationship between socioeconomic disadvantage and worse health outcomes cannot be inferred. The relationship between socioeconomic disadvantage and health is complex and often indirect involving multiple, temporally-evolving relationships between biological, behavioural, psychological, sociological or environmental factors [22]. Residual confounding by many factors not accounted for in this study cannot be excluded [23]. The found inverse associations could possibly be explained by unmeasured factors un-accounted for in this current study. Some of such factors relate to obesity and to smoking. Individuals coming from more disadvantaged backgrounds often smoke more and are more likely to have tobacco-related comorbidities that impact cardiovascular morbidity [24]. Similarly, the relationship of obesity with heart disease, hypertension, some cancers, type 2 diabetes and stroke are well established [25]. However, the association of obesity with socioeconomic disadvantage is multifaceted. Although obesity has been rising across all social classes, research indicates that some groups are more vulnerable than others. This variation of obesity by socioeconomic classes is complex as highlighted by a study that measured prevalence of obesity by different socioeconomic groups in 67 countries [26]. These authors show that in low-income and developing economies the affluent are more likely to be obese; however, a reverse relationship is noted in rich and developed economies where obesity is more prevalent among those coming from lower socioeconomic classes.

This current study also reports significant associations relating to age and sex – findings that may have ramifications for critical care services. As the population ages, the proportion of patients in need of critical care resources will likely increase. A six-year retrospective study that investigated 57 ICUs in Australia and New Zealand reported a significant rise in admission rates among patients aged 80 years or older of nearly 6% per year [27]. A similar rise in demand for critical care by the elderly has been reported in other countries as well [28]. In the United States, over the past ten years, the number of ICU beds per adult population mainly grew in regions with the largest elderly populations [28], with approximately 55% of all intensive care beds being occupied by patients aged 65 or more with 14% of patients 85 or more dying in the ICU setting. A Dutch study, that compared 1996 with 2006, found a 33% increase in number of patients aged 75 or more that needed critical care [29]. This rise in demand for critical care has been reported in North America, Europe, Australia and other regions, but not all patients meeting critical care admission criteria are admitted to such facilities [30, 31]. A British study that investigated ICU admissions among surgical patients in 94 National Health Service hospitals between 1999 and 2004 reported that approximately 85% of the surgical patients who died during admission were never admitted to an intensive care facility [30]. An Israeli incidence study found that 55% of all critically ill hospitalised patients were treated in medical wards that had no intensive specialised care [31]. Similar to these authors who reported that older age was associated with less admissions to an ICU, our study found that among the hospitalised population, older patients were less likely than younger patients to be admitted to a critical care unit. Given higher prevalence of comorbidities and frailty in the elderly, a less favourable risk profile for adverse outcomes and mortality may render them unfit to be admitted to a specialised care unit. This prognosis-based selection bias is commonplace when the demand for these specialised beds far exceeds their availability [31].

Sex variation in hospital admission rates among those presenting to the ED with coronary syndromes has also been previously reported by many studies [32‐35]. In general, women take longer than men to seek medical advice for chest pain [33], and when they do, compared to men diagnosed with similar conditions, women are less likely to be admitted to an acute care hospital and to undergo coronary revascularisation. Admitted women are also less likely to be treated in a specialised care unit. A two-year retrospective Canadian study found sex- and age-specific differences in ICU admission rates observed in over 24,000 consecutive adult ICU admissions [36]. A large three-year prospective European study showed that women admitted to critical care units were less likely than their male counterparts to receive invasive therapy and to have shorter ICU durations of stay, despite women having higher severity of illness scores than men needing admission to ICU [34]. The authors also reported a significantly higher risk adjusted in-hospital and ICU mortality among women. Age may be one explanation, as women are often older than men when diagnosed with a serious cardiac condition such as AMI. Compared to men, women also undergo fewer invasive procedures which might have lowered their chances of being admitted to an ICU or CCU as such admissions are higher following major procedures and surgeries [37, 38]. Our study and others [34, 36] show that age and acute diagnoses such as AMI cannot explain the sex differences in hospital care following a chest pain presentation in the ED.

To our knowledge, this study is the first to report dose-response relationships between geographic-based socioeconomic disadvantage and higher admission rates to specialised care units in a population-based representative sample of patients presenting with emergency chest pain. The relationship between socioeconomic disadvantage and cardiovascular morbidity is complex and multifactorial. Not every individual exposed to lower socioeconomic disadvantage develops disease and our results cannot infer causal relationships but our findings add to the accumulating evidence supporting a direct association between socioeconomic disadvantage and increased morbidity.