Changes in Surgical Volume and Outcomes During the Coronavirus Disease 2019 Pandemic at Two Tertiary Hospitals in Ethiopia: A Retrospective Cohort Study

BACKGROUND: Limited data exist concerning how the coronavirus disease 2019 (COVID-19) pandemic has affected surgical care in low-resource settings. We sought to describe associations between the COVID-19 pandemic and surgical care and outcomes at 2 tertiary hospitals in Ethiopia. METHODS: We conducted a retrospective observational cohort study analyzing perioperative data collected electronically from Ayder Comprehensive Specialized Hospital (ACSH) in Mekelle, Ethiopia, and Tibebe Ghion Specialized Hospital (TGSH) in Bahir Dar, Ethiopia. We categorized COVID-19 exposure as time periods: "phase 0" before the pandemic (November 1-December 31, 2019, at ACSH and August 1-September 30, 2019, at TGSH), "phase 1" starting when elective surgeries were canceled (April 1-August 3, 2020, at ACSH and March 28-April 12, 2020, at TGSH), and "phase 2" starting when elective surgeries resumed (August 4-August 31, 2020, at ACSH and April 13-August 31, 2020, at TGSH). Outcomes included 28-day perioperative mortality, case volume, and patient district of origin. Incidence rates of case volume and patient district of origin (outside district yes or no) were modeled with segmented Poisson regression and logistic regression, respectively. Association of the exposure with 28-day mortality was assessed using logistic regression models, adjusting for confounders. RESULTS: Data from 3231 surgeries were captured. There was a decrease in case volume compared to phase 0, with adjusted incidence rate ratio (IRR) of 0.73 (95% confidence interval [CI], 0.66-0.81) in phase 1 and 0.90 (95% CI, 0.83-0.97) in phase 2. Compared to phase 0, there were more patients from an outside district during phase 1 lockdown at ACSH (adjusted odds ratio [aOR], 1.63 [95% CI, 1.24-2.15]) and fewer patients from outside districts at TGSH (aOR, 0.44 [95% CI, 0.21-0.87]). The observed 28-day mortality rates for phases 0, 1, and 2 were 1.8% (95% CI, 1.1-2.8), 3.7% (95% CI, 2.3-5.8), and 2.9% (95% CI, 2.1-3.9), respectively. A confounder-adjusted logistic regression model did not show a significant increase in 28-day perioperative mortality during phases 1 and 2 compared to phase 0, with aOR 1.36 (95% CI, 0.62-2.98) and 1.54 (95% CI, 0.80-2.95), respectively. CONCLUSIONS: Analysis at 2 low-resource referral hospitals in Ethiopia during the COVID-19 pandemic showed a reduction in surgical case volume during and after lockdown. At ACSH, more patients were from outside districts during lockdown where the opposite was true at TGSH. These findings suggest that during the pandemic patients may experience delays in seeking or obtaining surgical care. However, for patients who underwent surgery, prepandemic and postpandemic perioperative mortalities did not show significant difference. These results may inform surgical plans during future public health crises.