Gender, socioeconomic status and emergency department visits among cancer survivors in the USA: a population-based study

To assess patterns of emergency department visits and subsequent hospitalization in relation to gender and socioeconomic status among a cohort of cancer survivors in the USA.

The current study evaluates the impact of gender and socioeconomic status on the rate of emergency department visits and subsequent hospitalization among a cohort of cancer survivors in the USA. It suggested that female cancer survivors were more likely to visit the emergency department; whereas, they were less likely to be subsequently hospitalized. Moreover, cancer survivors with lower socioeconomic status (represented by incomplete insurance coverage, unemployment and lower income) were more likely to visit emergency departments and more likely to be hospitalized because of this. Other factors associated with emergency department visits included poor health status, African American race and unmarried status.

Results from the current analysis add to a growing list of publications highlighting the impact of socioeconomic status on outcomes of cancer survivors. Prior studies suggested that cancer survivors with lower socioeconomic status are more likely to have worse cancer-specific outcomes and a higher risk of suicide [16–18]. The current study highlights the impact of socioeconomic status on health resource utilization including emergency department visits and hospitalizations. The link between socioeconomic status and emergency department visits might lie in gaps in insurance coverage which would limit access to routine care. Higher likelihood of emergency department visits might arise from lack of routine care which would increase the need for emergency services. A prior NHIS-based study evaluated the impact of high-deductible health plans with health savings account and it showed that health savings account was associated with less delayed care among high-deductible health plans cancer survivors (p < 0.05) [19]. Cancer diagnosis and treatment themselves can be explanatory factors for low socioeconomic status and lack of proper insurance coverage (e.g., through inability to return to work post-treatment because of disability) [20].

While there is no clear explanation for the observed gender-based differences, a number of hypotheses can be provided based on previously published data in the general public as well among cancer patients. Previous studies have suggested that women report more frequently physical symptoms potentially related to emotional distress (e.g., headache and back pain) compared with men; and this might have led to more emergency department visits that did not result subsequently in hospitalizations [21]. A prior Swedish unlike-sex twins’ study has also shown that while women had more total health conditions compared with men, there was no excess in life-threatening health conditions among women compared with men [22]. Another study based on 3.8 million patients in the UK, showed higher mean number of visits to the family physician (general practitioner in the UK) in females than males (aged 10–65 years) [23]. A prior qualitative synthesis of studies evaluating delays in cancer presentation suggested a reluctance on the side of men compared with women with regards to seeking help related to cancer symptoms [24]. A previously published study also suggested a higher probability of many nonlife-threatening adverse events among women with metastatic colorectal cancer compared with men without a difference in serious or fatal adverse events [25]. Another possible explanation for this observation is the higher likelihood for atypical presentations of some serious conditions (e.g., stroke or myocardial infarction) among women compared with men [26,27]. This might have (theoretically) led to a higher possibility of missing these diagnoses and; therefore, less hospitalizations. A third possible explanation might be related to the differential rates of different cancers among women versus men. This might have affected the potential short-term and long-term adverse events of cancer treatment that patients have experienced and subsequently their clinical course with regards to the need for emergency department visits and/or hospitalizations. A fourth possible explanation for gender-based differences might be related to differences in socioeconomic status between women and men. As clarified in the current study women were more likely to have lower income and insufficient insurance coverage. Both of these factors might be linked to less access to routine primary care; and subsequently, higher probability of emergency department visits. Consistent with this hypothesis is the reported reasons for emergency department visits among women versus men (Figure 3); 36.6% of women versus 32.8% of men went to the emergency department because ‘they did not have another place to go’. An important caveat for the interpretation of the current study also is the magnitude of reported ORs of gender-based differences (of 0.86 and 1.07). Although these are statistically significant findings, the differences seem to be modest given the large sample size of the study.

The current study is also in line with prior studies conducted in the noncancer population which suggested that incomplete health insurance coverage is associated with a higher probability of emergency department visits/hospitalizations [28,29]. Possible reasons for this observation would include the fact that incomplete health coverage is usually associated with delayed disease presentation, more advanced disease, less access to regular care and hence a higher probability of complications necessitating emergency management [30,31]. Individuals with no or limited coverage are also less likely to have a regular primary care provider. Looking at Figure 3 (which describes patient-attributed reasons for emergency department visits), ‘not having another place to go’ was the reason in a third of participants, this also highlights the potential link between lack of access to a regular care provider and frequent emergency department visits. Another potential reason is the prior observation that poorer Americans are more likely to have poorer health [32]. This means a higher probability for emergency department visits (even for reasons unrelated to cancer).

For uninsured US residents, part of the costs of emergency department visits/hospitalizations is paid as an out-of-pocket expense; but also part of the costs is being absorbed by other entities including some public and private institutions (that serve the uninsured at no/reduced charge); different levels of government (federal and state governments, that support hospitals’ operations); and donations from charities or philanthropic entities [33]. Thus, improving access of cancer survivors to elective insured health services (and possible subsequent decrease in the number of emergency department visits) is likely to be a cost-saving intervention as it will decrease the costs incurred by inpatient hospital care for cancer survivors. Taken together, these findings suggest that inadequate health coverage (and subsequent lack of access to regular providers) would not only worsen health outcomes but also economic outcomes (as hospital/inpatient care is generally more expensive than outpatient care). It is interesting also to note that individuals with private health insurance plans were less likely to visit the emergency department; while individuals with Medicaid coverage plans were more likely to visit the emergency department. This might indicate that not only the mere presence of coverage but also the type/quality of the coverage might be associated with emergency department visits.

It is important here also to highlight the long-term adverse effects, multiple different cancer treatments might have on cancer survivors (e.g., treatment-related cardiotoxicity) [34]. This may have played a role in the increased risk of all-cause hospitalization and/or emergency department visits. There is thus an important role for post-treatment survivorship care and transition back to primary care providers to monitor long-term cancer treatment adverse events and subsequent hospitalizations or emergency department visits.

The current study has several limitations and strengths that need to be recognized. First, information about cancer diagnosis as well as emergency department visits and subsequent hospitalizations was provided through self-reporting of included participants rather than through objective review of administrative records. This is coupled with the fact that NHIS datasets lack information about cancer stage or treatment(s) received. Although this is the nature of population-based surveys like NHIS, this would weaken the veracity of association analyses between socioeconomic status, gender and emergency department visits/hospitalizations. One also is not sure about the sensitivity/ specificity of self-reporting of cancer diagnosis in this setting. Second, NHIS datasets do not provide a distinction between an individual with cancer undergoing active therapy versus a cancer survivor who finished active treatment. This is particularly relevant in the context of assessment of reasons of emergency department visits and subsequent hospitalizations. Third, information about the general condition of the participants is provided through a self-reported health status variable and the medical reasons for emergency department visits/hospitalizations are not mentioned but only patient-reported reasons for emergency department visits were reported. It is possible that background medical conditions other than cancer were the reason for emergency department visits and this was not captured through NHIS datasets. This is compounded by the absence of information about standardized comorbidity indices (e.g., Charlson Comorbidity Index) within NHIS datasets. These limitations highlight the fact that the results of the current analysis indicate association rather than causation. On the other hand, the potential strengths of the current study included the large sample size of the study and the rigorous quality of NHIS datasets. Because of the unique nature of the healthcare system in the US, extrapolation of the results of the current analysis to other healthcare settings (e.g., in Canada or Europe) should be done with caution.

In conclusion and within the study cohort, female cancer survivors were more likely to visit the emergency department, whereas they were less likely to be subsequently hospitalized. Cancer survivors with lower socioeconomic status were more likely to visit emergency departments and more likely to be hospitalized because of this. Further work needs to be done to minimize disparity in access to cancer care services.

This work is based on the NHIS annual datasets (2011–2017).

The author has received advisory board honoraria from Eisai Canada, Roche Canada and Lilly Canada. The author has no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

The author state that he has followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations.