The impact of the 2014–15 Ebola virus disease epidemic on emergency care attendance and capacity at a tertiary referral hospital in Freetown, Sierra Leone: a retrospective observational study

Abstract

BackgroundThe Ebola virus disease epidemic in West Africa has infected 28 457 people and claimed more than 11 000 lives. Many more people may have died from the indirect effects of the epidemic and closure of normal health-care facilities. Unlike other facilities in West Africa, the emergency department in Connaught Hospital, Freetown, Sierre Leone, protected by an onsite Ebola holding unit, continued to provide emergency care throughout the outbreak. We aimed to assess the effect of the outbreak on emergency department attendance and presentation. We also analysed emergency care capacity across Freetown. MethodsAttendance data from the emergency department and Ebola holding unit at Connaught Hospital were collected from June 1, 2014, to June 1, 2015. Severity of presentation was derived from South African Triage Score (SATS) assigned at first presentation to the emergency department. A mean severity score was calculated by dividing the number of presentations with a SATS of 1–2 by the total number of presentations. Local prevalence of the disease was counted as RT-PCR positive cases at the Ebola holding unit. Emergency care capacity was assessed at the seven principal hospitals in Freetown in May, 2013, and in April, 2015, with a standardised tool, the Emergency Care Capacity Score (ECCS), specifically designed for the low-income setting. All data were collected in Excel (2013). Stata (version 13) was used for statistical analysis. Findings8935 patients presented to the emergency department; mean attendance was 172 patients per week (95% CI 153–191), with attendance varying from 41 patients in the week beginning July 28, 2014, to 284 patients in the week beginning May 11, 2015. Emergency department attendance had a negative correlation with local prevalence of Ebola virus disease (r=–0·640, p<0·0001) (appendix). The proportion of severe cases also varied from a peak of 26·5% in week 40 of 2014 (a period of high local prevalence, 74·6%) to 3·2% in week 28 of 2014 before any case had occurred in Freetown. The mean severity of presentation was 11·9% (95% CI 10·3–13·4). The ECCS was reduced across all domains except the systems domain (appendix). Total ECCS for all seven hospitals decreased by 10% from 2013 to 2015, and all facilities showed a decrease in their individual ECCS score. InterpretationThe reduction in attendance probably demonstrates both a change in health-seeking behaviour—ie, great public fear of hospitals because of the perceived risk of nosocomial transmission of the virus—and a reduction in access to care. The decrease in emergency care capacity was expected and reflects the closure of many health services other than those for Ebola virus disease. Overall, this is an important case study of the impact of an infectious disease outbreak on a tertiary referral hospital in a low-income setting. FundingNone.