Logistic regression is a statistical method that is widely used in cross-sectional and cohort studies to identify and estimate the effects of risk factors. Imperfect diagnostic tests may, however, bias the outcome. A negative test can be interpreted as lack of infection and a positive test can be as infection presence. As long as the covariates do not influence the sensitivity and specificity of the diagnostic tests, the bias is towards null, but if the sensitivity and specificity are influenced by covariates, the direction of the bias cannot be easily predicted.

Valle et al. performed a systematic review. PubMed was searched using different combinations of the search terms ‘malaria’, ‘logistic’, ‘models’, ‘regression’, ‘diagnosis’, and ‘diagnostic’. Of 36 studies that satisfied the criteria, 70 % did not address the issue of the imperfect detection in malaria outcome.

The authors interprets their results as suggesting that malaria epidemiologists are generally unaware of the consequences that  imperfect detection can have on parameter estimates from logistic regression. The authors also recommends using Bayesian models instead of logistic regression.

Reference

Valle D, Tucker Lima JM, Millar J, Amratia P, Haque U. Bias in logistic regression due to imperfect diagnostic test results and practical correction approaches. Malaria Journal 2015;14:434