Background

Pharmacovigilance (PV) is an essential component of health systems. Functional PV systems protect and promote public health by supporting the safe and effective use of medicinal products through the prevention and mitigation of harm. With increased simultaneous introduction of innovative products in high-, middle- and low-income countries, e.g., COVID-19 vaccines, or solely in low- and middle-income countries (LMIC), e.g., malaria vaccines, PV systems in LMIC must be able to detect safety signals and ensure adequate safety surveillance. This research aims to analyse the development of PV systems across high-, middle- and low-income countries and to carve out essential elements for implementing functional and sustainable PV systems in LMIC.

Methods

A convergent parallel mixed-methods design, consisting of qualitative and quantitative methods was used. Qualitative research consisted of semi-structured interviews. To expand the breadth and range of the study, a quantitative survey was conducted, focusing on the same thematic questions as the semi-structured interviews.

Results

Twelve key informants from nine countries were interviewed and 52 respondents from 36 countries completed an online questionnaire. Four major themes consisting of 12 categories emerged from the data. Based on these, the following elements essential for building functional and sustainable pharmacovigilance systems in LMIC were identified: understanding the drivers of PV development; adequately resolving core system challenges; implementing an efficient organisational structure and procedures for PV; investing in activities beyond reporting of adverse drug reactions; identifying alternate sources of financing; having a national strategy with a vision and mission for PV; adequately leveraging the health system; and effectively integrating the pharmaceutical sector in the national PV system.

Conclusions

Findings from this research revealed progress in PV systems in LMIC in the last decade, though significant efforts are still needed to develop these systems to meet global standards. Developing the different areas emerging from this research, which necessitates implementing functional PV structures and processes, adequately leveraging the health system and effectively engaging the pharmaceutical sector, through the mechanisms proposed, would enable a comprehensive progression from basic to stable, functional and thus sustainable PV systems in LMIC.

Study design

This research aims to analyse stakeholders’ perspectives on the development of PV systems across high-, middle- and low-income countries and to carve out essential mechanisms necessary to establish and ensure functionality and sustainability of PV systems in LMIC. To achieve this, a convergent parallel mixed-methods design, consisting of qualitative and quantitative methods was used (see Figure 1). Qualitative research consisted of semi-structured interviews. To expand the breadth and range of the study, a quantitative survey was conducted, focusing on the same thematic questions as the semi-structured interviews.

Setting and study population

Sampling of key informants was purposive. Recruitment criteria were based on informants’ knowledge and expertise in PV, active participation in national or global PV, and/ or decision-making position within national and global PV organisations. To allow a global perspective on building robust PV systems, PV stakeholders from high-, middle-, and low-income countries, from different regions of the world, according to World Bank Group country classifications (35) were recruited. To ensure adequate representation of national and global PV stakeholders, representatives from the NRA, Public health programmes (PHP) such as National Immunization Programmes (NIP), Non-Governmental Organisations (NGO) and donor agencies who provide technical and financial support (hereafter referred to as Technical and Financial Partners – TPF), were included in the selection. For the interviews, the definite sample size was determined by willingness to participate in the research and subsequently by the level of data saturation.

Data collection

Qualitative data were collected through semi-structured interviews built around the topic ‘Strategies to build functional and sustainable pharmacovigilance systems – an analysis of pharmacovigilance implementation in high-, middle- and low-income countries’. An interview guide (Additional file 1) was developed based on the study’s objective and on the previously performed scoping review (21). The interview guide was divided into five sections with a total of 30 questions: 1) key informant’s role in the national PV system; 2) organisation of the national PV system; 3) PV and health system development; 4) PV and pharmaceutical development; and 5) ensuring functional and sustainable PV systems. The interview guide was reviewed by the co-authors for clarity and validity of questions, then pilot-tested by a PV expert not directly involved in the research. Sixteen key informants were invited by email to participate in the interviews and were provided with an overview of the research. Verbal informed consent to record the interview and to use of the information provided by the key informant in the research was obtained at the start of each interview. The interviews were conducted by the first author using Zoom videoconferencing (Zoom Video Communications Inc.) over a nine-month period (November 2021 to July 2022). The duration of the interviews ranged from 60 to minutes to 90 minutes.

Quantitative data were collected using a standardised questionnaire (Additional file 2), informed by the interview questions. The questionnaire was set up in ODK (https://opendatakit.org) and pilot-tested by a PV expert not directly involved in the research. The questionnaire, in both English and French, was distributed to 80 persons by email and WhatsApp Messenger (WhatsApp Inc.) between November 2022 and February 2023; three reminders were sent to non-respondents.

Data analysis and interpretation

The Framework Method was used for the thematic analysis of qualitative data (36). The respondents’ statements from the interviews were verbatim transcribed by the first author and deductive coding was used to code the transcript. For this purpose, a codebook was developed with the predefined codes organized into corresponding categories based on the research objectives. The codes were assigned to the transcribed data, line by line, and related sub-codes were created to improve the accuracy of the analysis. The codes were then clustered into categories. Using Microsoft Excel (Microsoft Corporation, 2016) each transcript was summarized and the data were ‘charted’ into a matrix consisting of sub-codes, codes and categories. The data were interpreted, with the identification of patterns, relationships, differences and similarities leading to new thematic groups.

Quantitative data collected in ODK were exported into Microsoft Excel 2016 for analysis using a PivotTable. Data analysis consisted of descriptive statistics, primarily frequencies and percentages for categorical variables.