What specific adverse events (organ systems, grades) have randomized trials of ivermectin reported, and how consistent are their definitions across studies?
Executive summary
Randomized trials of ivermectin for COVID-19 and other indications report adverse events spanning neurological, gastrointestinal, dermatologic, ocular, and systemic (including multiorgan failure) categories, but serious events are rare and trials differ markedly in what they record and how they label severity [1] [2] [3]. Systematic reviews and meta-analyses generally find no clear increase in overall adverse events or in severe adverse events versus control, but they emphasize low or very low certainty driven by heterogeneous definitions, small event counts, and inconsistent reporting across RCTs [4] [5] [6].
1. What randomized trials actually reported — the raw inventory
Large collections of randomized controlled trials included in systematic reviews and meta-analyses extracted adverse event data as secondary endpoints, with many individual RCTs reporting counts of any adverse event (AE) and of severe adverse events (SAE) rather than standardized organ-by-organ lists; pooled analyses therefore typically present summary AE and SAE rates rather than harmonized symptom inventories [4] [7] [8]. Several high-profile RCTs specifically noted neurological or systemic SAEs: a JAMA randomized trial of mild COVID-19 reported multiorgan failure as the most common SAE (four events, split equally between ivermectin and placebo) [1]. Oncology and other non-COVID RCTs using ivermectin list specific graded AEs (rash, anemia, diarrhea, muscle weakness) when employing oncology trial reporting standards [9].
2. Organ systems and typical adverse events documented
Across pharmacovigilance summaries, RCTs, and reviews the recurrent organ-system categories are neurological (headache, dizziness, encephalopathy/confusion, tremor), gastrointestinal (nausea, vomiting, diarrhea), dermatologic (rash, pruritus), ocular (transient visual symptoms noted particularly in onchocerciasis/high-dose contexts), and systemic events including fever or multiorgan failure in rare cases [10] [2] [3]. Meta-analyses of COVID-19 RCTs report composite AE endpoints and find no consistent signal for increased AEs with ivermectin, and note ocular AEs rose in at least one high-dose onchocerciasis trial while otherwise severity profiles were similar across standard and higher dosing in pooled analyses [3] [5].
3. Severity grading: what systems were used and how consistent are they?
Severity grading is inconsistent: some trials and oncology studies report AEs using CTCAE-like grades (examples from a phase I/II oncology study list specific grade 1–3 events), many COVID-19 RCTs report only “any AE” versus “serious AE” without applying a common grading rubric, and pharmacovigilance databases code events by MedDRA System Organ Class and Preferred Terms rather than trial-grade numerics [9] [10] [4]. Systematic reviewers repeatedly flag this heterogeneity—different thresholds for what counts as an SAE, variable follow-up windows, and diverse adjudication processes—leading to downgraded certainty in pooled safety estimates [4] [8].
4. What pooled evidence concludes about frequency and seriousness
Multiple systematic reviews and meta-analyses conclude ivermectin does not clearly increase overall AEs or SAEs compared with control, with some analyses reporting no difference in severity between standard and higher doses and others describing no significant difference in SAE rates in COVID-19 trials, though many reviews grade that evidence as low or very low certainty because of imprecision and trial limitations [3] [6] [5]. The WHO and guideline panels therefore advise restricting ivermectin use to trials in COVID-19 contexts because benefit is unproven and safety data remain heterogeneous [4] [8].
5. Why definitions and reporting diverge — implications for interpreting harm
Divergence arises from multiple sources: RCTs were run in different settings with varying case-report forms and AE adjudication, some trials pooled AEs into composite counts while pharmacovigilance uses MedDRA coding, oncology trials apply formal grade scales, and rare but serious signals (neurologic encephalopathy or multiorgan failure) are often single-case or few-case events that are hard to attribute in small trials—this mix produces statistical imprecision and inconsistent labeling across publications [10] [2] [9]. Review authors explicitly cite imprecision and inconsistent definitions as reasons to downgrade confidence in safety estimates [4] [5].
6. Bottom line
Randomized trials report adverse events across multiple organ systems with most events mild or transient, serious events uncommon and unevenly distributed across studies, and important inconsistencies in how severity is defined and reported across RCTs and pharmacovigilance sources; pooled analyses generally do not detect a clear increase in SAE risk but classify the certainty of that conclusion as low to very low because of heterogeneity, small numbers of events, and variable grading practices [3] [4] [6].