Background: The goal of the present cohort study was to review outcomes of patients exposed to interferon beta-1b during pregnancy. Methods: Pregnancy cases with exposure to interferon beta-1b reported to Bayer’s pharmacovigilance (PV) database from worldwide sources from January 1995 through February 2018 were retrieved for evaluation. Only cases where pregnancy outcomes were unknown at the time of reporting (i.e. prospective cases) were included in the analysis of this retrospective cohort study. Results: As of February 2018, 2581 prospective pregnancies exposed to interferon beta-1b were retrieved from the database; 1348 pregnancies had documented outcomes. The majority of outcomes [1106 cases (82.0%)] were live births. Health status was known for 981 live births (no known health status for 125). Most of the prospective pregnancies with known outcomes corresponded to live births with no congenital anomalies [896 cases (91.3%)]. Spontaneous abortion occurred in 160 cases (11.9%). Congenital birth defects were observed in 14/981 live births with known health status [1.4%, 95% confidence interval (CI) 0.78–2.38]. No consistent pattern in the type of birth defect was identified. Rates of both spontaneous abortion and birth defects were not higher than the general population. Conclusions: These PV data, the largest sample of interferon beta-1b-exposed patients reported to date, suggest no increase in risk of spontaneous abortion or congenital anomalies in women exposed during pregnancy.

Introduction Signal validation in pharmacovigilance is the process of evaluating data to decide whether evidence is sufficient to justify further assessment of a detected signal. During the signal validation process, safety experts in our organization are required to review signals of disproportionate reporting (SDRs) and classify them into one of six predefined categories. Objective This experiment explored the extent to which predictive machine learning (ML) models can support the decision making of safety experts by accurately identifying the most appropriate predefined signal validation category. Methods We extracted cumulative data for six medicinal products, consisting of historic SDR validations and Individual Case Safety Reports, from the company’s safety database for training and testing of the ML model. We implemented a decision tree-based supervised multiclass classifier model termed Gradient Boosted Trees followed by a SHapley Additive exPlanations (SHAP) analysis to mitigate the “black box” effect of the ensemble model by identifying the key predicting features in the model. Following a retrospective analysis, a prospective experiment was conducted to test the model accuracy and user acceptance in a real-life setting. Results The prediction accuracy of our ML model ranged from 83 to 86% over 3 months for the six medicinal products. The applicability of the model was confirmed by the company’s safety experts. Additionally, the systematic predictions provided valuable information to the safety experts and assisted them in reviewing the SDRs efficiently and consistently. Conclusions This experiment demonstrated that it is possible to train a multiclass classification model to accurately predict signal validation categories for SDRs. More importantly, the transparency of the predictions provided by the SHAP analysis led to high acceptance by the safety experts.

The aim of this study was to analyze the risk of hypersensitivity reactions (HSRs) to iopromide in children and elderly patients in comparison to adults. Materials and Methods: Four observational studies were pooled and analyzed (analysis I). In addition, spontaneous reports from 1985 to 2020 from the pharmacovigilance database were evaluated (analysis II). All patients received iopromide for angiographic procedures or contrast-enhanced computed tomography in various indications. In analysis I, a nested case-control analysis, including a multivariable logistic regression model, based on pooled observational study data, was performed. Cases were defined as patients with a typical and unequivocal HSR; controls were patients without any recorded reaction. In analysis II, all spontaneous reports on HSRs after iopromide administration recorded in the pharmacovigilance database were descriptively analyzed. Exposure estimates on the size of the exposed age groups were derived from sales data and data from market research. The primary target variable was the risk of HSR to iopromide in children (<18 years) and elderly patients (≥65 years) compared with adults (≥18 to <65 years). Results: In analysis I, a total of 132,850 patients were included (2978 children, 43,209 elderly, and 86,663 adults). Hypersensitivity reactions were significantly less frequent in children (0.47%) and elderly (0.38%) compared with adults (0.74%). The adjusted odds ratio (vs adults) for children was 0.58 (95% confidence interval, 0.34-0.98; P < 0.043), and that for the elderly was 0.51 (95% confidence interval, 0.43-0.61; P < 0.001), indicating a lower risk for both subpopulations as compared with adults. In analysis II, of the overall >288 million iopromide administrations, 5.87, 114.18, and 167.97 million administrations were administered to children, elderly, and adults, respectively. The reporting rate for HSRs in children (0.0114%) and elderly (0.0071%) was significantly lower as compared with adults (0.0143%) ( P < 0.0001). Conclusions: Hypersensitivity reactions to iopromide were significantly less frequent in children and elderly compared with adults.