A decade after the empirical testing of the Health Belief Model by Becker et al. in 1974, Janz and Becker published a review in 1984 in the Health Education Quarterly Journal.
What studies they reviewed
They have reviewed 46 Health Belief Model (HBM) – related studies: 18 prospective and 28 retrospectives. It included both preventive health behaviours and sick role behaviours.
Studies on healthy role behaviours
It included 3 studies of vaccination behaviour against Swine Flu, one vaccination study against influenza, one study related to screening for genetic disease (Tay-Sachs disease), four studies related to breast self-examination, and one study on high blood pressure screening.
Studies on drug compliance behaviours
Not only studies on health promotion behaviours, they also looked at medicine compliance behaviour studies too. It included anti-high blood pressure compliance, adherence to dietary changes among individuals with diabetes and end-stage renal disease, mother’s compliance with physician advice on child’s health, and clinic attendance.
Overall, one important finding they mentioned (p.41) that findings from the prospective study designs (which is superior design than retrospective ones), were either similar or better than the findings from retrospective study designs. This is great news.
According to their comparison of pre-1974 and post-1974 findings, “perceived susceptibility” topped as the most powerful variable in pre-1974 research whereas “perceived barriers” became the most powerful variable in post-1974 studies irrespective of the study design either prospective or retrospective and in the preventive and sick role behaviours. However, “perceived susceptibility” remains more important in preventive health behaviours than sick role behaviours. “Perceived severity” earned much lower significant score (36% in the significant ratio) in preventive health behaviours whereas it rose up to 85% in sick role behaviours such as drug compliance.
A major limitation
The reviewers have correctly identified that the absence of standardised measuring tools for each HBM variable renders significant difficulties in the model appraisal process.