Antihypertensive Pharmacotherapy Gaps in Nigeria: A Predictive Machine Learning Analysis of Treatment Uptake Amid Macroeconomic Shock, Using NDHS 2023-24
This study utilizes machine learning on the 2023-24 Nigeria Demographic and Health Survey to reveal significant sex-specific disparities in antihypertensive treatment non-uptake driven by macroeconomic shocks and geopolitical zones, concluding that uniform national interventions are ineffective and advocating for tailored, sex- and region-specific strategies.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
High blood pressure is a silent force that strains the heart and vessels, affecting hundreds of millions of people worldwide. While the condition is manageable with medication, a significant problem remains: many people who know they have high blood pressure never start taking the pills that could save their lives. This gap between diagnosis and treatment is a critical failure point in public health, particularly in regions where healthcare systems are still developing. In Nigeria, the world's most populous African nation, this issue is compounded by a unique set of circumstances. The country recently faced a severe economic shock, with the sudden removal of fuel subsidies and a sharp drop in the value of its currency. These events sent prices soaring, including the cost of imported medicines. Researchers wanted to know if this economic turmoil was pushing people away from their treatment, and whether men and women were experiencing this pressure differently.
To answer these questions, a team of researchers turned to a massive, nationally representative survey conducted in Nigeria between late 2023 and early 2024. They focused specifically on adults who had already been told by a doctor that they had high blood pressure. The goal was to understand why some of these diagnosed individuals were not taking their medication. The study examined nearly 3,000 women and over 500 men, looking at factors like their age, wealth, location, and the specific month they were interviewed. By treating the timing of the interviews as a natural experiment, the researchers could see if the worsening economic conditions during those months changed the likelihood of people stopping their treatment. They also tested whether complex computer models could predict who would stop taking their medicine better than standard statistical methods.
The findings revealed a striking and unexpected divide between men and women. For women, the timing of the interview mattered deeply. As the months passed and the economic crisis deepened, the odds of a diagnosed woman not taking her medication increased steadily. By the end of the survey period, the likelihood of skipping treatment had roughly doubled compared to the start. This suggests that the economic shock was directly eroding women's ability to stay on their medication. However, this pattern did not hold for men. For the men in the study, the passage of time and the deepening economic crisis showed no measurable effect on whether they took their pills. Their treatment habits remained steady regardless of when they were asked, indicating that the economic pressure was hitting the two sexes in fundamentally different ways.
Geography also played a role, but again, the impact was reversed for men and women. In certain regions of the country, being from a specific area made it less likely for a woman to skip her medication, acting as a protective factor. Yet, for men living in those exact same regions, the opposite was true; those locations became strong risk factors for not taking treatment. The researchers found that while wealth generally helped women stay on treatment, it did not show a clear pattern for men. Furthermore, having diabetes alongside high blood pressure was a strong predictor that women would continue their treatment, likely because they were already engaged with the healthcare system for their other condition. This protective link was not observed for men.
When the researchers tried to use advanced computer algorithms to predict who would stop taking their medicine, the results were mixed. For the women, the different computer models performed similarly to each other, and only slightly better than a standard statistical approach, achieving a modest level of accuracy. For the men, the computer models struggled significantly, performing no better than random guessing. This difficulty likely stemmed from the smaller number of men in the study and a lack of detailed health data for them, such as body measurements, which were available for the women. The study concluded that the complex, tree-based computer models did not offer a clear advantage over simpler methods in this specific context, especially when the data was limited.
The study challenges the idea that a single, uniform national program could effectively solve the problem of untreated high blood pressure in Nigeria. The evidence suggests that the barriers to treatment are not the same for everyone; they shift based on gender, location, and the timing of economic shocks. The researchers propose that future efforts should be tailored to these differences. For women, integrating blood pressure care into existing maternal and child health services could be effective, while men might need outreach through workplaces or community health workers. Additionally, the findings highlight the need for a more resilient supply chain for essential medicines, ensuring that economic fluctuations do not cut off access to life-saving drugs. The study does not claim to have solved the problem, but it provides a clear map of where the gaps are and why they exist, showing that a one-size-fits-all approach will not work in a country as diverse and dynamic as Nigeria.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.