A hybrid data-driven and qualitative approach for critical factor identification in pharmaceutical supplier selection: safeguarding the continuity of pharmaceutical in disruption-prone settings

Purpose This study aims to develop a comprehensive and integrated approach for identifying and prioritizing critical factors influencing pharmaceutical supplier selection and supply chain performance, with direct implications for health system resilience and access to essential medicines. By positioning supplier selection as an upstream determinant of preparedness, the study addresses the concerns of humanitarian and health supply chain actors that depend on commercial pharmaceutical suppliers for the uninterrupted availability of essential medicines during disruptions and emergencies. Design/methodology/approach A hybrid methodology is adopted, combining Principal Component Analysis (PCA), a machine learning-based dimensionality reduction technique and qualitative analysis using NVivo. Initially, 67 factors were identified through literature and expert input, reduced to 20 using PCA and subsequently validated through qualitative thematic analysis. Findings The analysis identified seven key themes encompassing operational, risk, supplier capability, financial, governance, technological and ethical dimensions of pharmaceutical supply chains. Operational factors such as delivery accuracy, availability of supplies and cold chain capabilities are fundamental to supply continuity and patient safety. Risk-related factors, including political instability and labour disruptions, significantly impact supply chain resilience. Supplier capabilities, governance factors, technological innovation and ethical considerations jointly drive resilience, sustainability and competitiveness. Research limitations/implications The current study contributes to supplier selection theory by combining PCA with qualitative thematic analysis, with a focus on operational, technological, governance and ethical aspects. The hybrid method of dimensionality reduction using machine learning and qualitative depth increases the reliability and interpretability of the data and understanding of the assessment processes for the healthcare supply chain. The framework also extends the humanitarian logistics literature by connecting commercial upstream supplier decisions to the preparedness and resilience of downstream health supply chains serving populations affected by disruptions and crises. Practical implications Healthcare organizations should adopt a multi-criteria approach to supplier evaluation, considering delivery accuracy, cold chain, financial sustainability and regulatory compliance, all of which affect patient safety and the continuity of essential medicines. Building resilient, health-system-aligned supply chains requires investment in digital technologies like AI, along with proactive measures such as multi-sourcing and geographic diversification. For humanitarian organizations, national procurement agencies and global health programmes, the prioritized factors provide a screening template for evaluating pharmaceutical suppliers whose reliability ultimately determines whether essential medicines reach crisis-affected and underserved populations. Originality/value This study contributes to the literature by providing a holistic, multi-dimensional framework that integrates PCA-based machine learning and NVivo-based qualitative insights for supplier selection – an approach not previously applied in the pharmaceutical supply chain context. It further contributes to the humanitarian logistics and health supply chain literature by demonstrating how upstream supplier selection functions as a preparedness mechanism for maintaining access to essential medicines under conditions of political instability, labour disruption and demand shocks.