This study shows that CAM use is common among adults receiving TB care at Connaught Hospital in Freetown, Sierra Leone, dominated by herbal remedies and characterised by low disclosure, largely because patients reported that clinicians did not ask. The observed prevalence exceeds many regional estimates and may reflect pluralistic health-seeking behaviour. A systematic review of traditional and complementary medicine use in the region reported wide variation across countries and populations but consistently identified herbal medicines as the most commonly used modality [14]. TB care may amplify CAM demand because symptoms can be prolonged, treatment is lengthy, and stigma and uncertainty shape health-seeking behaviours[7, 8]. In pluralistic health systems, patients may move between biomedical and traditional providers without viewing these choices as mutually exclusive. Our finding that symptom relief and wellbeing were dominant motivations suggests CAM is used as an adjunct to conventional care, often addressing unmet symptomatic or psychosocial needs. Notably, substances with dual dietary-medicinal roles (e.g., ginger, garlic) were only classified as CAM when used with explicit therapeutic intent. Kola nut, despite its widespread cultural use in Sierra Leone, had low prevalence as a therapeutic agent in this cohort and did not drive overall CAM estimates.

Non-disclosure of CAM use represents an important actionable finding. Patients perceived that clinicians did not ask, suggesting a missed opportunity for communication rather than confirmed provider behaviour. This pattern mirrors evidence from diverse settings. Systematic reviews report non-disclosure rates ranging from approximately one-quarter to three-quarters, identifying provider non-inquiry, fear of negative responses, and perception that disclosure is unimportant as major barriers [2, 19]. Meta-analytic work emphasises that disclosure is strongly influenced by patient-provider communication quality and whether clinicians ask directly [19]. These data support a practical inference: routine inquiry about CAM is likely to increase disclosure, creating opportunities to identify risks, align expectations, and strengthen therapeutic relationships.

Safety implications require caution. First-line anti-TB regimens can be hepatotoxic and commonly cause gastrointestinal intolerance; concurrent herbal ingestion can complicate symptom attribution and may contribute to preventable treatment interruptions. Systematic reviews describe plausible pharmacokinetic and pharmacodynamic mechanisms for herb-drug interactions, including effects on cytochrome P450 enzymes and drug transporters, though clinical relevance varies by product and context [10, 11]. Separate reviews highlight that herbal and dietary supplements can cause drug-induced liver injury, with causality assessment challenging due to product heterogeneity [12, 20]. In our cohort, 11% of CAM users reported complications, most commonly symptom worsening or diarrhoea. While self-reported and non-specific, these findings underscore the importance of documenting CAM use when evaluating adverse symptoms during TB therapy.

Our results are relevant to adherence and the broader TB care experience. Although adherence was not directly measured, evidence from West Africa indicates that medicinal plant use may be associated with adherence difficulties during TB treatment [21]. The motivations reported here suggest patients may use CAM to manage persistent symptoms, perceived side effects, or illness-related anxiety. This presents a patient-centred opportunity: TB teams can proactively address common symptoms, optimise supportive care, and provide clear guidance on when to seek review, thereby reducing perceived need for unmonitored adjunct therapies. Importantly, counselling should avoid stigmatising language, since fear of discouragement was a reported reason for non-disclosure and is consistently identified in the literature [2, 19].

The lower use among students may reflect different health beliefs or access to biomedical information. The absence of other independent associations and modest model performance suggest unmeasured social and cultural determinants. Future studies should incorporate variables reflecting explanatory models of illness, trust in healthcare, social networks, household income, and access to traditional healers, and should consider qualitative methods to understand decision-making around CAM use during TB treatment.

These findings suggest practical actions for TB services in Sierra Leone and similar contexts. First, integrate CAM assessment into routine history-taking using standardised prompts that normalise disclosure. Second, strengthen provider communication skills for non-judgemental inquiry and culturally respectful counselling. Third, safety-focused counselling should document CAM type and timing, discourage high-frequency use of unknown preparations during intensive therapy, encourage prompt symptom reporting, and consider targeted liver function monitoring when clinically indicated. Finally, TB programmes could develop patient education materials explicitly discussing concurrent herb and TB medication use, framing this as a safety issue rather than moral judgement.