Individuals may have different access to diagnostic tests, due to cultural, geographic, economic or other reasons, and these factors affect the detection of disease. This leads to diagnostic access bias, which may cause a study’s results to be biased, or unrepresentative of the broader population of interest and under or overestimate the true incidence of disease in the population depending on the level of access.
Often sarcoidosis (an inflammatory disease usually affecting the lungs) does not cause symptoms and so whether an individual has a diagnosis of sarcoidosis depends in part on access to medical methods of detection. In an observational study, 46% of health professional were diagnosed via routine exam compared to 19% of the general population. Health professionals were therefore 2.4 times more likely to be diagnosed with sarcoidosis (95% CI: 1.34 to 4.33) due to the presence of diagnostic access bias.
Several observational studies have demonstrated low incidence and prevalence rates of Atrial Fibrillation despite higher rates of relevant cardiovascular risk factors in African Americans compared with Caucasians. A potential explanation is their lower access to healthcare and the diagnostic tests for Atrial Fibrillation.
Diagnostic access bias would over-estimate the prevalence of disease in some groups health when they have greater access to testing. Whereas, in some group lower access leads to an underestimation in specific populations.
When researchers design an observational study that examines the association between particular patient groups and disease they should consider inequalities in access to diagnostic tests, and healthcare more broadly. If inequalities are anticipated, a priori sensitivity analysis should be conducted to determine if results are confounded by these inequalities.