Nearly Half of HIV Patients in Nigerian Cohort Hit First Drug Reaction Within 13 Months

Adults living with HIV who had already spent years on antiretroviral therapy did not stop running into adverse drug reactions once the risky early months of treatment were behind them. In a retrospective cohort study of 332 treatment-experienced patients cared for at the State Specialist Hospital Gombe in north-eastern Nigeria between January 2018 and December 2023, nearly half of the participants experienced at least one adverse drug reaction during follow-up, and the median time to a first ART-related reaction was 13 months, with a marked drop in reaction-free survival between 10 and 15 months of observation. By 20 months, the Kaplan–Meier curve showed, most participants had experienced at least one ART-related adverse drug reaction.

The study, published in Discover Pharmaceutical Sciences, set out to fill a specific gap. The Gombe team, led by Erick Wesley Hedima and Adamu Umar Abubakar, instead examined treatment-experienced adults, defined as people living with HIV aged 18 or older who had been enrolled in care at the hospital, had received ART for at least one year, and had at least one documented viral load measurement. By including reactions attributed to any medication, not only antiretrovirals, the investigators hoped to capture the fuller pharmacological burden carried by patients who often take co-trimoxazole, isoniazid preventive therapy, and drugs for opportunistic infections alongside their HIV regimen.

The numbers came from medical folders selected by systematic random sampling. From a total of 3,134 people living with HIV in the centre’s records over the study period, every ninth folder was reviewed until the calculated sample size of 332 was reached. Data collectors used a structured checklist covering sociodemographic details, virological parameters, and the history of adverse drug effects, with physicians’ diagnoses recorded on patient cards and folders serving as the basis for identifying reactions and dating their onset. The analysis relied on Kaplan–Meier survival curves, log-rank tests, and univariate and multivariable Cox proportional hazards regression, with statistical significance set at a p-value below 0.05.

The cohort itself reflected the demographics of the clinic. Nearly half of the participants were aged 25 to under 35 years at ART initiation, about half had more than three years of ART experience, roughly six in ten had tertiary education, about 64 percent were female, and nearly eight in ten lived in semi-urban areas. Clinically, 65.4 percent were classified as WHO stage 1 at initiation, 82.2 percent were on the tenofovir disoproxil fumarate, lamivudine and dolutegravir regimen, 98.8 percent used co-trimoxazole, and 70.8 percent took isoniazid for tuberculosis preventive therapy. Among the 119 patients with comorbidities, 10.8 percent had hypertension and 24 patients, or 7.2 percent of the cohort, had tuberculosis co-infection.

The adverse reaction burden was substantial. Across the cohort, 133 participants, reported as 40.1 percent in the abstract and 39.9 percent in the results section, experienced at least one adverse drug reaction. The most commonly reported reactions were skin rashes, pruritus and chest pain, each cluster accounting for 25 patients or 7.5 percent of the cohort, followed by insomnia and headache at 6.9 percent. Switching to an alternative ART regimen or brand was the most frequently implemented management strategy, used in 26.2 percent of cases. The total observation time across the cohort was 2,766 person-months, yielding an overall incidence density of adverse drug reactions of 4.8 per 100 person-months and an ART-related incidence of 3.4 per 100 person-months.

Attribution, however, was not straightforward, and the paper reports the split inconsistently. The abstract states that 71.4 percent of the reactions, 95 of 133, were attributable to antiretroviral therapy, while the results section reports that only 28.8 percent, given as 96 patients, were attributed to ART. The incidence calculations in the paper use the 95-patient figure for the ART-related rate. The authors themselves acknowledge the difficulty, noting that the substantial burden of reactions combined with challenges in clearly attributing them to ART underscores the need to strengthen active pharmacovigilance within HIV programs, with particular attention to non-ART co-medications such as tuberculosis preventive therapies.

The multivariable Cox model identified regimen and disease stage as independent predictors of time to first ART-related reaction. Compared with patients on the widely used TDF/3TC/DTG regimen, those receiving TDF/3TC/EFV had a 50 percent lower risk of experiencing an adverse drug reaction, with an adjusted hazard ratio of 0.50 and a 95 percent confidence interval of 0.26 to 0.96, while patients on abacavir-based regimens showed a 60 percent reduced risk, with an adjusted hazard ratio of 0.40 and a confidence interval of 0.20 to 0.80. WHO clinical stage II at ART initiation more than doubled the risk relative to stage I, with an adjusted hazard ratio of 2.19 and a confidence interval of 1.34 to 3.58. On unadjusted log-rank analysis, patients on first-line therapy and those in WHO stages II and III at initiation reached their first reaction sooner, with median times of 12 and 11 months respectively.

Some of these associations ran against the authors’ expectations. The finding that the dolutegravir-based first-line regimen carried higher risk was, the paper concedes, counterintuitive, since that regimen is generally regarded as well tolerated, and the authors suggest the result could reflect a type 1 error. The protective association with abacavir also contrasts with other literature cited in the discussion, where abacavir-based regimens showed the highest adverse event rates, 52.7 per 100 person-years, and tenofovir-based regimens the lowest, 26.4 per 100 person-years, in second-line cohorts. The authors note that low local prevalence of the HLA-B*5701 hypersensitivity marker may reduce abacavir reactions in this population, and that their time-to-event design among treatment-experienced adults may yield different risk estimates than cumulative incidence approaches.

The discussion places the incidence figure in regional context. A Nigerian study reported 4.6 adverse drug reactions per 100 person-years, close to the Gombe rate, while Ethiopian studies reported lower rates of 0.67 and 3.5 per 1,000 person-months, differences the authors attribute to follow-up duration and regimen composition. Skin reactions emerged as the most consistently reported category across the cited literature, with cutaneous reactions affecting 16.08 percent of 3,532 patients in one study and skin rash in 28 percent of patients in another, although chest pain and pruritus were not consistently highlighted as leading reactions elsewhere. The authors also report that virological status predicted reactions in their cohort, with virally suppressed patients at lower risk, plausibly because suppression reduces polypharmacy for opportunistic infections.

For clinical practice, the central message is temporal. Because the median time to first reaction was 13 months and the steepest decline in reaction-free survival occurred between 10 and 15 months, the authors argue that intensified monitoring should extend well beyond ART initiation into the first one to two years of treatment, particularly for treatment-experienced patients, those on isoniazid preventive therapy, people in WHO stage II, and individuals with unsuppressed viral loads. They recommend closer follow-up, more frequent symptom screening, structured adverse reaction assessment integrated into routine HIV care, standardized reporting tools, and improved training of health workers in recognition, documentation and causality assessment. The retrospective single-centre design carries real limitations the authors acknowledge: missing or incomplete records may have led the study to underestimate the true incidence and spectrum of adverse drug reactions, and physician-diagnosed reactions in paper folders are an imperfect gold standard. Even so, the study offers a concrete, dated window into when reactions arrive for patients long past their first year of therapy.

Subject of Research: Time to first adverse drug reaction among treatment-experienced adults living with HIV on antiretroviral therapy in Nigeria

Article Title: Evaluation of time to first adverse drug reaction among treatment-experienced adult persons living with HIV/AIDS: a retrospective cohort study

Article References: Hedima, E. W., & Abubakar, A. U. (2026). Evaluation of time to first adverse drug reaction among treatment-experienced adult persons living with HIV/AIDS: a retrospective cohort study. Discover Pharmaceutical Sciences, 2(1), Article 9. https://doi.org/10.1007/s44395-026-00044-6

Image Credits: AI Generated

DOI: 10.1007/s44395-026-00044-6

Keywords: HIV, antiretroviral therapy, adverse drug reactions, pharmacovigilance, retrospective cohort, Nigeria, Kaplan-Meier, Cox regression, WHO clinical staging, treatment-experienced patients, drug safety, viral load suppression