Psychology Ph.D. Dissertations

Using the Health Belief Model to Predict Injecting Drug Users' Use of Harm Reduction

Date of Award


Document Type


Degree Name

Doctor of Philosophy (Ph.D.)



First Advisor

Harold Rosenberg (Committee Chair)

Second Advisor

William O'Brien (Committee Member)

Third Advisor

Anne Gordon (Committee Member)

Fourth Advisor

Molly Laflin (Committee Member)


Based on the Health Belief Model (Rosenstock, 1966), the current study was designed to examine whether injecting drug users' (IDUs) beliefs about two injecting-related health conditions (i.e., non-fatal overdose and bacterial infections) and two harm reduction behaviors that prevent these conditions (i.e., injecting test shots and pre-injection skin cleaning) predicted their short-term intentions to engage in those two health behaviors. Ninety-one current IDUs recruited from needle exchange programs in Ohio and Michigan completed a series of questionnaires. Specifically, participants answered questions about their perceived susceptibility to and the perceived severity of the two health conditions, the perceived benefits and barriers of engaging in these two harm reduction behaviors, their self-efficacy to use these harm reduction behaviors, perceived social network norms regarding use of the harm reduction behaviors, and their recent use of and short-term intentions to use the behaviors in four drug-use situations (i.e., in withdrawal, not in withdrawal, alone, with others). Participants also completed a brief interview about their health promotion behaviors and perceived barriers to using the two harm reduction behaviors. Results indicated that recent past use of these two harm reduction behaviors consistently and positively predicted short-term intentions across all four situations. Only two constructs of the Health Belief Model also predicted intentions to engage in harm reduction, depending on the drug-use situation. Specifically, perceived susceptibility to non-fatal overdose and perceived benefits of test shots were significant positive predictors of intentions to engage in test shots, but only if injecting when not in withdrawal; perceived susceptibility was a significant positive predictor of intentions to do test shots, but only when injecting if alone. Participants' rating of how often other injectors in their network use test shots was also a significant and positive predictor of doing test shots, but only if injecting with others or injecting alone. Analysis of open-ended interview responses also revealed that participants engage in behaviors to preserve their health in general and specific to injecting, and that they identify a variety of barriers to engaging in harm reduction behaviors. Based on these results, future research could evaluate whether discussing susceptibility to non-fatal overdose, listing the benefits of test shots, and encouraging IDUs to inject with others who engage in harm reduction behaviors may be valuable additions to interventions designed to increase these two harm reduction behaviors. Future research should explore whether the HBM predicts actual use of harm reduction interventions in addition to behavioral intentions and what other non-cognitive factors predict the use of harm reduction.