Our study revealed differences in health-seeking behavior among subgroups of the initial COVID-19 cases in Singapore. Locally-acquired cases had significantly longer duration from symptom onset to hospital admission than imported cases (median 6 days vs. 4 days). Among those with at least one visit prior to admission, a higher percentage of locally-acquired cases had sought medical attention at the primary care level than imported cases (89.8% vs. 56.5%).
Imported cases were more likely to attend ED, bypassing primary care clinics completely (43.5% of imported cases vs 10.2% of locally-acquired cases attended ED directly without any clinic visits, p < 0.001), and those who attended clinics had fewer visits before being referred for further testing at the national screening centre or other EDs (6.2% of imported cases had at least three primary care visits prior to admission compared with 21.3% of locally-acquired cases) (Table 1). This observation could be due to two reasons: first, imported cases consisted mostly of Singapore international undergraduate students and foreigners working in Singapore, who were less likely to have a regular physician and with a raised perception of their infection risk, would attend ED directly or earlier in their patient journey; second, the heightened vigilance of primary care physicians towards this group. As the pandemic unfolded and more was known about the virus, Singapore Ministry of Health issued travel advisories and revised circulars circulated to doctors on suspect case definitions. Hence, the attitude of doctors would evolve accordingly with a lower threshold on testing patients deemed to be at increased risk of infection, which resulted in fewer visits by imported cases.
In contrast, doctors’ perception of lower community transmission risk at that time resulted in locally-acquired cases being referred only after lack of clinical improvement despite repeated visits. Equally crucial is the patient’s personal cognizance and health literacy [15,16,17]. Locally-acquired cases might have attributed their symptoms to a common cold or gastroenteritis, resulting in delays lasting up to a month in seeking medical attention. While seemingly innocuous under normal circumstances, this could have devastating consequences in an outbreak situation.
Overall, Singapore residents were more likely to attend primary care clinics than non-residents (84.0% vs 58.7%). Fever, cough, sore throat and diarrhoea were common presenting symptoms of COVID-19 cases , routinely managed within the community by family physicians.
Among the subgroup of locally-acquired cases with at least two primary care visits who saw the same care provider (median of 2 visits), a smaller proportion (27.6%) had three or more visits before being referred and admitted, compared with 61.4% of those who saw different providers (median of 3 visits). Having no basis for comparison from previous visits, a different healthcare provider lacks pertinent information when formulating management plan for the patient, which typically leads to later diagnosis, isolation and treatment, and consequently increases transmission risk. Our results thus highlight the risk of seeking care from multiple care providers or “doctor shopping” within the same episode of illness [19,20,21].
Doctor shopping could be attributed to various factors. One factor is healthcare accessibility; with high concentrations of primary care clinics island-wide, the convenience of attending clinics near one’s workplace and home outweighs care continuity concerns [19, 22,23,24]. Another factor is unmet expectations; patients might have misconceptions of partaking less efficacious medications as symptoms persist, or feel unsatisfied with previous consultations [19, 22,23,24]. Hence, appropriate public health communication to the public is crucial even during peacetime.
We acknowledged several limitations in our study. The observational design of our study precluded causal inference. This study was limited to the initial COVID-19 period where cases were predominantly imported. As the epidemic progressed, health-seeking behavior would evolve and as such, an in-depth study would be useful to ascertain attitudes and responses of the Singapore population at each phase of the outbreak. Our study was confined to cases diagnosed and managed in Singapore, and the findings may not be generalized to health-seeking behavior of COVID-19 cases in other countries with different health systems and financing mechanisms. As some information related to primary healthcare visits prior to hospitalization was ascertained based on self-reporting, the data collected was subjected to recall bias. Nevertheless, there were standard operating procedures in place to ensure the accuracy and consistency of information documented, such as having trained public health officers to interview the cases, and verifying movements reported by the cases from other sources.