Situated in West Africa, Liberia has experienced back-to-back shocks due to decades of civil conflict, followed by the unprecedented Ebola outbreak ¹ and ongoing public health threats, including COVID-19, mpox, measles, and Lassa fever. ^{2 – 4} Despite efforts to enhance resilience across sectors, including with the adoption of a pro-poor agenda, ⁵ the country presently ranks 177 (out of 191) on the Human Development Index. ⁶ Largely external investments in capacity-building have sought to fill gaps. ⁷ For the health sector, institutional developments have improved systems to better prepare the country for detecting, notifying, and responding to threats in the post-Ebola period; however, investments in improving data quality, analysis, and use have been less prioritized than the generation or reporting of data, and systems remain vulnerable. ⁸

Despite Liberia’s experiences with the 2014-2015 Ebola outbreak and more recent COVID-19 pandemic and the extensive quantitative research undertaken with external collaborators to understand the potential trajectories of these health crises, ^{9 – 11} few efforts have been made in Liberia to institutionalize local capacity strengthening opportunities in quantitative data analysis for public health researchers and practitioners. To date, there is no university degree being administered in biostatistics, and dynamic mathematical modeling is not taught even as a course in existing academic programs. Short-term training often facilitated by visiting collaborators is being conducted at institutional levels. However, the limited literature being produced in Liberia using quantitative analysis ¹² and the reliance on international consultants for work that is being done ¹³ provide evidence that the few existing models of teaching coding and statistics may not be generating independent capacity. That gap has implications for overall development as well as resilience in the face of health and other threats. ^{14 – 16}

To provide a local opportunity for health sector personnel to undertake advanced and sustained analytical learning, Quantitative-Data for Decision-Making (Q4D) Lab was developed by a team of local and international public health scientists. The NGO envisions training Liberians who are working in, or aim to work in, the health sector in relevant quantitative skills. The existing course series introduces foundational coding skills in R Statistical Software ( https://www.r-project.org/), followed by bivariable analysis and visualization, and ultimately multivariable regression with generalized linear models. An alternative route provides training in dynamic transmission modeling after foundational coding and statistics skills are developed. Most students engage with Q4D after learning about the program from peers who have attended classes. The student body reflects people working at the Ministry of Health, the National Public Health Institute of Liberia, international non-governmental organizations, as well as students in undergraduate and graduate public health programs or in health professions schools. The classes have nominal fees ( e.g., $50USD for Beginner R class which is less than the fee of one credit hour at the local public university), making them accessible but also attaching some value to hold students accountable to attending sessions. Moreover, class sessions focus on practice versus theory to keep the content and skills grounded in actual applications in the Liberian health sector context—both the health issues facing the sector and the constraints facing those attempting to conduct research everyday. ^{15 , 17}

Based on an understanding of the limited foundation for coding and statistics generated by current educational programs in Liberia, ¹⁷ the Q4D Beginner R class assumes no working knowledge of R or Biostatistics and builds confidence and skills from the ground up. Programmatic data offers insight into students’ baseline, with students reporting ahead of the Beginner R program that they were not too confident that they could develop coding skills in R software (35% rated themselves 1 or ‘not at all confident’ out of 5 on a confidence scale, with a median ranking of 3) ( Figure 1A) and that they anticipated their lack of previous analysis and coding experience would be a challenge to successfully completing the program. Thus, despite often being already employed in the health sector and having roles related to monitoring and evaluation, most students enter the Q4D program with limited computer skills and little or no prior experience with coding or statistical software (When asked if they had previously encountered a problem that needed to be solved using a spreadsheet or data analysis, nine out of ten indicated that they had and that they had been unable to solve it.). While the Q4D program was developed with this contextual understanding, no study to date has been undertaken to evaluate whether the skills taught in the program are fulfilling the goals of Q4D Lab in terms of their relevance to data and analytics needs in the health sector and their utilization outside of the classroom. Specifically, to inform programmatic growth and larger scale evaluations, it is important to gauge whether the capacity being built is being used by students in their personal and professional lives and is perceived as having impact.

This paper thus aims to highlight contributions that the Q4D program has made towards quantitative analysis capacity in Liberia, specifically in the health sector. The findings offer insight into the utility and relevance of the course content for current workforce needs and reflect how the skills gained have contributed to personal or professional advancement of students to date.

The findings presented here highlight how the R coding and statistics skills being taught through the Q4D program have relevance and utility to Liberians employed in health-related sectors or in the pipeline for future employment. Most of the surveyed Q4D students are confidently applying the skills they learned, and several have leveraged their new skill sets to obtain scholarships, consultancies, and promotions at work. While the present study was not a direct assessment of the skills taught (although skills evaluation is conducted as part of the courses and ahead of certification), these opportunities reflect that the capacity being built is recognized and valued by employers and/or selection committees. Moreover, our study demonstrates that students have acquired interest in quantitative research careers and that they perceive the value of quantitative research evidence for informing decision-making in the health sector. While our study is small, it offers insight into possible outcome and impact measures for assessing capacity-strengthening efforts in the quantitative sciences and lays the groundwork for larger scale evaluation as the program grows.

Increasing local buy-in for evidence-driven decision-making in health serves as a compelling case for investment in initiatives to address health workforce capacity gaps effectively. ^{18 , 19} Such initiatives must be two-fold—addressing the gaps to promote evidence generation and addressing the gaps to promote evidence utilization. As ongoing efforts are being undertaken to encourage research utilization in the health workforce of Liberia, ²⁰ Q4D is focused on the former. Capacity-building to produce cadres of well-trained statisticians with awareness of and adherence to quality standards will not only enhance the available evidence base ²¹ but also reduce the risk of erroneous research findings that have been attributed to lack of appropriately trained statisticians. ²² Specifically, with increased capacity for quantitative research embedded in the health sector, analyses to assess health trends, understand factors associated with health risks, and evaluate the effectiveness of interventions will increasingly empower policymakers with rigorous evidence to make informed decisions for improved health outcomes. This result contributes to broader development objectives under the Sustainable Development Goals. ²³ However, it is important to note that building in-depth analytical capacity in Liberia and other sub-Saharan countries is complex and warrants ongoing evaluation of context-sensitive strategies, as we have aimed to do with our tracer study.

In resource-constrained settings, quantitative research applied to routine data has the potential to extend its value and provide meaningful evidence without requiring additional cost-intensive field- or laboratory-based work. As one barrier to regular and rigorous analysis of such data, education and training programs throughout sub-Saharan Africa lack resources to develop the human capital needed for science, technology, engineering and math (STEM) fields. ²⁴ In parallel, quantitative capacity-building efforts by development partners have tended to be directed towards technical assistance or short-term workshops not always sensitive to local constraints ^{25 , 26} or aimed at contextually relevant outcomes. ²⁷

To achieve sustainable, sufficient, and well-grounded capacity in quantitative sciences, there is a clear need for improvements in fundamental STEM education across all levels. ²⁸ Accomplishing this is a paradoxical situation, as building capacity where there is a lack of it requires that there are well-equipped cadres of teachers and professors across quantitative fields. ²⁴ Until systemic changes can be affected, however, in-service training—such as that offered by Q4D programs—can boost confidence and capacity for coding and statistical analysis when it gauges, recognizes, and builds on the existing foundation of quantitative skills, regardless of how strong it is. Such medium-term, intensive in-service training that builds basic through advanced skills thus offers a model for demand-driven, locally based training in settings that may have a weak baseline for quantitative skills learning. Through the approach, initial progress can be made to contribute to a skilled health workforce that can address current and emerging health challenges effectively, as recommended by the World Health Organization. ²⁹

While the present study suggests that the Q4D program is providing skills training in areas that are of relevance and use in students’ workplaces and/or academic programs, it was not an evaluation of the skills gained. Regular assessment is undertaken throughout the Q4D courses to ensure that students can independently demonstrate sets of pre-defined skills. One next step will be such an outcome evaluation comparing coding and analysis skills among Q4D graduates to a control group, ideally in a workplace setting. While continuing to evaluate existing efforts will be important, the present study highlights areas where the program could improve for greater impact. To promote continued opportunities for Q4D students, increased acceptance or recognition of local capacity and demand for it by health sector leadership are necessary. It has been observed that health ministries in low- and middle-income countries fail to realize their potential for knowledge transfer as learning organizations that are supportive of developing people and processes within them. ³⁰ Therefore, activities by which Q4D engages health sector leadership and increases the visibility of students’ skills applications should be undertaken. To accomplish this, Q4D can learn from other local training programs, such as the Liberia Field Epidemiology Training Program, with strong ties across the health sector. Moreover, it is important that those who have completed the program and are applying the skills receive continued support to evaluate their applications for rigor (such as by adherence to statistical guidelines and standards) and to ensure ongoing exposure to additional methods. This need for in-service mentorship to extend the impact of intensive training has been noted with other programs. ^{31 , 32}

In addition, it is recognized that the current programs offered at Q4D may be limited in scope, as they are focused on R Statistical Software and examples that largely reflect concepts from epidemiology. These design components were intentional—as R is free, open-source software, and it does not require a significant amount of space on personal computers, and as many of the initial cohorts of students were personally and professionally interested in epidemiology. However, future programs should reflect other skills and concepts to address the broader needs of the health sector. For instance, with the ongoing shift towards more transdisciplinary planetary health, ³³ the Q4D program has experienced a higher proportion of students in or aspiring to be in the environmental health, environmental science, and/or One Health sectors. As the needs and interests of the student population evolve, project topics, datasets, and examples used in class can draw from student experiences to enhance the relevance of materials and allow for more seamless application of skills in their jobs and studies. In parallel, Q4D can align itself with other organizations that provide capacity-building in One Health or complementary areas and partner with experts from local and international universities.

In conclusion, the Q4D program in Liberia offers quantitative skills training to a diverse student body, including a high proportion who are employed within the health sector. While students reported personal opportunities that arose with attainment of new skills, the program’s impact has potential to extend beyond individual growth, contributing to the country’s capacity for data-driven decision-making, research, and sustainable development. As Liberia continues to strengthen its capacity for quantitative analysis, the Q4D program could serve as a model for locally-relevant capacity building initiatives. Further evidence generation to assess skills acquisition and its use in the health sector with a larger sample size will be critical. Importantly, this focus on regular monitoring and evaluation of Q4D programmatic outcomes will ensure that ongoing programmatic decisions are evidence-based and nationally impactful. It will also set a precedent for other quantitative sciences capacity-strengthening initiatives in Liberia.

STB and LAS developed the initial research objectives and study protocol. MKAK, TOY provided substantive feedback on the objectives and protocol. Data collection was managed by STB and GBD and monitored by TOY and MKAK. The data analysis plan was developed by STB, TOY, MKAK, and LAS and implemented by STB and LAS. STB, GDB, and LAS developed the first draft. TOY, MKAK provided substantive feedback on the draft. All authors read and approved the final manuscript.

The study protocol was approved by the University of Liberia IRB (ULIRB IORG-IRB Number: IRB00013730) on May 16, 2023. All study procedures were carried out according to the Declaration of Helsinki. Participants were asked to provide written consent via an online form ahead of gaining access to the questionnaire. The consent form text is included in the Extended data. All results have been presented using aggregate statistics rather than any individual participant’s response. For direct quotes on specific examples of skills applications, potentially identifying information—such as on health district where one works—was removed.