(331) Applicability of Selected Brighton Collaboration Case Definitions in Prospective Hospital-Based Safety Surveillance for Covid-19 Vaccines, Addis Ababa, Ethiopia

Background: Standardizing case definitions for priority vaccine safety conditions facilitates uniform evaluation and consolidation of data collected in various diagnostic capabilities worldwide. The Brighton Collaboration case definitions support this harmonization and enable classification from level 1 (most certain) to level 5 (least certain [not a case]) of certainty. Assessing the performance of these definitions in real-world situations is critical, particularly in low-income countries where new vaccines may first be introduced without previous monitoring in high-income countries.

Objectives: To describe the applicability of selected Brighton Collaboration case definitions in the largest referral hospital in Ethiopia.

Methods: The hospital is located in the capital, Addis Ababa. Active surveillance prospectively identified patients eligible for Covid-19 vaccination who presented with one or more of 11 pre-specified adverse events of special interest (AESIs) at the emergency outpatient department (EOPD) or inpatient wards. From 1 June 2022 to 31 May 2023, three trained physicians employed structured data collection forms with algorithms to capture patient information and determine the Brighton Collaboration level of diagnostic certainty (LOC), regardless of vaccination status for Covid-19. Data were entered into electronic forms that applied Boolean logic to automatically determine the LOC; discrepancies between the manual and automatic LOC were reviewed and verified. We conducted descriptive analysis to characterize cases and the LOCs reached for each AESI.

Results: Overall, 203 AESI cases were identified; 162 (80%) were detected from the EOPD and 117 (58%) were residents of Addis Ababa. The most commonly detected cases were thrombosis/thromboembolism (n = 100, 49%) and generalized convulsion (n = 38, 19%). Ninety-two percent (n = 187) of the cases were confirmed at levels 1–3 LOC, including all (100%) of the cases of thrombocytopenia, 98% of cases with thrombosis/thromboembolism, 97% of cases of generalized convulsion, 84% of cases of myocarditis, 75% of cases of pericarditis, and 91% of cases of Guillain-Barré syndrome (GBS). Seven (3%) cases, including 2 pericarditis cases, were nonassessable (level 4). The difficulties in classifying pericarditis cases were due to lacked chest radiograph or electrocardiogram results to achieve at least level 3 LOC, but these tests are not specialized and are routinely available..

Conclusions: The Brighton Collaboration case definitions can be successfully applied in >95% of cases in a large referral hospital in Ethiopia, with with generalized convulsion, pericarditis, and encephalomyelitis as the exceptions. We recommend further evaluation in other low-resource settings, particularly in rural or non-tertiary hospitals, to gain additional insights for revision/adaptation of the definitions.