Data Collection Methods, Training, and Perceived Health Information Quality Among Professionals in Three Ghanaian Hospitals: A Cross-Sectional Study.

Background and Aims: Reliable health information underpins clinical care, disease surveillance, and health-system planning. However, routine data in many low- and middle-income countries, including Ghana, are often incomplete, inaccurate, or delayed. Most prior evidence concerns objectively measured data quality in single facilities. Less is known about how healthcare professionals perceive the adequacy of their data-collection methods and collected-data quality. Also, how these perceptions relate to the method used and to prior training is unknown. This cross-sectional study assessed those perceptions and their associations across three Ghanaian hospitals. Methods: Of 574 professionals sampled, 564 responded (98.3%). A pretested questionnaire measured perceptions on five-point scales, dichotomized at a two-thirds cut-off. Associations were quantified with crude and adjusted binary logistic regression (odds ratios with 95% CIs). Results: Methods were perceived adequate by 67.2% (379/564) and data quality by 60.1% (339/564). Training showed the strongest association with perceived method adequacy (adjusted OR: 5.54, 95% CI: 3.66-8.39). Electronic methods showed no significant advantage over manual methods in either crude or adjusted models. Conclusion: In this study, training demonstrated a stronger adjusted association with favorable perceptions than the type of data-collection method used. Because the design is cross-sectional and outcomes self-reported, causation cannot be inferred. Longitudinal studies with objective data-quality verification are needed.