A 20-valent pneumococcal conjugate vaccine (PCV20) was approved for adults by the US Food and Drug Administration in 2021. We conducted a data-mining signal-identification analysis to assess possible adverse events after vaccination. We included individuals aged ≥ 19 years from ten Vaccine Safety Datalink sites who received PCV20 between June 2021 and October 2024 and had continuous medical coverage from ≥ 400 days before through 84 days after vaccination. Using a self-controlled design, incident diagnoses from inpatient and emergency department encounters were assessed for statistical clustering during post-vaccination risk windows within the tree-like structure of International Classification of Diseases, 10th Revision, Clinical Modification diagnosis codes. Two analyses using tree-based scan statistics were conducted: conditional tree-temporal, with moving risk and control windows of varying lengths within an 84-day period; and conditional Bernoulli, with pre-specified risk and control windows of 1–21 and 43–84 days post-vaccination, respectively. A statistically significant cluster was defined as a p-value < 0.01. We identified 1,042,847 PCV20 vaccinees. In the tree-temporal analysis, statistically significant clusters included codes for non-specific adverse effects of a vaccine or other medical product, within 7 days post-vaccination (p-value range: 0.0001–0.0059); fever, 1–2 days post-vaccination (p-value range: 0.0001–0.0004); cellulitis of the left upper limb, days 2–5 (p-value 0.0001); and hemorrhoids or perianal venous thrombosis, days 42–63 (p-value 0.0045). We also detected statistically significant clusters for non-specific adverse effects from vaccines via the Bernoulli analysis within 21 days post-vaccination (p-value range: 0.0001–0.0069). We reviewed individual-level data from a sample of patients in the cluster for hemorrhoids; hemorrhoids appeared to be either present before vaccination or found during routine colonoscopy. This untargeted scan for potential adverse events following PCV20 vaccination did not identify clusters of diagnoses representing new, serious safety concerns, though there was low power to detect an association with rare outcomes. The identified clusters were known vaccination-associated events in pre-licensure randomized controlled trials and prior observational analyses or unlikely to be associated with the vaccine.
Maternal immunisation is an effective strategy for reducing morbidity and mortality in pregnant women and neonates, yet it requires close safety monitoring, but safety surveillance capacity in Africa has historically been limited. The coronavirus disease 2019 (COVID-19) pandemic provided an unprecedented opportunity to evaluate Africa’s capacity for maternal vaccine pharmacovigilance and assess whether pandemic-era gains could be sustained. To characterize maternal vaccine safety reporting from Africa in VigiBase® (2010–2024), compare reporting patterns for COVID-19 and non-COVID vaccines, and identify priorities for strengthening maternal immunization safety surveillance A retrospective descriptive analysis of individual case safety reports submitted by African countries to VigiBase®, the World Health Organization global pharmacovigilance database, was conducted for January 2010 to December 2024. Reports were identified using the Uppsala Monitoring Centre pregnancy algorithm and classified according to the Brighton Collaboration Global Alignment of Immunisation Safety Assessment in Pregnancy adverse events following the maternal immunisation framework. Descriptive statistics and chi-square tests were used to compare COVID-19 and non-COVID vaccine reporting patterns, and a Mann–Whitney U test was used to compare pre-pandemic and pandemic annual reporting volumes. Africa contributed 817 of 44,316 global maternal vaccine safety reports (1.84
The United States Prescribing Information (USPI), also known as the Food and Drug Administration (FDA) drug label, serves as a primary source of information for healthcare professionals to access reliable information on drug benefits and risks. Within the drug label, adverse events are listed across varying sections (e.g., Contraindications, Warnings and Precautions, Adverse Reactions). However, at present it is unclear what evidentiary thresholds FDA applies before including specific adverse event types in various sections of the drug label. To examine this question, we conducted an analysis of United States (US) regulations, guidance documents, standard operating procedures and policies published by FDA pertaining to the assessment of adverse events and drug labeling. This study examines how causality assessments shape decisions regarding the inclusion and placement of adverse events in the US drug label. We analyzed US regulations and policy documents published by FDA up to June 2024. Sources included the electronic Code of Federal Regulations, FDA policy documents (e.g., guidance documents, standard operating procedures, and policies) identified through searches of the FDA website and Westlaw legal database. Documents were analyzed using ATLAS.ti to extract and compare causality-related language across drug labeling sections. Our analysis identified heterogeneous and sometimes contradictory causality assessment evidentiary thresholds across US drug label sections. We found no predefined or standardized causality thresholds specified in FDA regulations, allowing interpretive flexibility that may contribute to variability and opacity in labeling decisions. For example, the Adverse Reactions section requires “some basis to believe” a causal relationship, while the Warnings and Precautions section states that “reasonable evidence” of causality is needed for serious events. We recommend greater standardization of causality-related terminology and public documentation of evidentiary rationales supporting labeling decisions. Improved clarity regarding evidence certainty would better support prescribers and patients in understanding drug-related harms and strengthen public trust in regulatory drug safety assessments.
Disproportionality signal detection methods can help identify potential adverse outcomes of medication exposure. Such methods scan all medication–outcome pairs within a dataset to identify possible associations requiring further investigation. Many signal detection methods use separate statistical tests for each pair, however, information sharing between groups of ‘similar’ medications or outcomes may produce more sensitive and specific detection of harmful effects. Data-defined groups identified through clustering may provide a better reflection of dependencies within the data. We present a new signal detection method that combines the information component, biclustering and Bayesian hierarchical modelling (IC–Bic–BHM) to identify groups within the data and use these groups within the analysis. We use simulations to demonstrate the importance of using groupings that accurately reflect dependency within the data; and compare sensitivity, specificity, positive predictive value and negative predictive value of the IC–Bic–BHM with other methods. Results of application to 536 medications and 61 types of congenital anomaly in babies/foetuses from 87,513 EUROmediCAT pregnancies were also compared. Sensitivity and specificity of Bayesian hierarchical models are dependent upon how accurately groupings reflect dependency within the data. The IC-Bic-BHM is the first method to use data-defined dependencies between medications and outcomes in signal detection. Simulations suggest it performs well in datasets with large numbers of medication–outcome pairs but requires adequate power to detect groups. The IC-Bic-BHM identified signals involving several known teratogenic medications within EUROmediCAT data. This study presents a novel approach to signal detection but requires further investigation before adoption.
