HLT 205 Week 2 DQ 1: Defensive Medicine

Define it precisely and then split it, because the split is what makes the three impact questions answerable. Defensive medicine is clinical action taken primarily to reduce liability exposure rather than because it is expected to benefit the patient, and it takes two forms. Assurance behaviour is doing more — additional imaging, referral, admission, repeat testing. Avoidance behaviour is doing less — declining high-risk patients or procedures, dropping obstetrics from a practice, leaving a speciality or a state with an unfavourable liability climate. The two have opposite effects on access, which is why treating defensive medicine as a single phenomenon produces a muddled answer to the third impact question. Notice that the definition turns on motivation rather than on the act itself, which is what makes defensive medicine genuinely hard to measure: the same scan is defensive or indicated depending on why it was ordered, and only the clinician knows. Say so, because it explains why the cost estimates disagree so violently.

Take the three impacts separately. Cost is where the largest claims are made and where you should be most careful, since estimates vary by an order of magnitude depending on method, and studies using tort reform as a natural experiment generally find smaller effects than surveys of physicians' self-reported behaviour. Report a figure with its method attached. Quality is genuinely two-directional: some assurance behaviour catches real disease, while over-testing produces incidental findings, cascades of further investigation, false positives and procedural harm, so more care is not more quality. Access is where avoidance behaviour bites, and it bites unevenly — rural obstetrics is the standard example. Then the forecast, which should reason from mechanism rather than guess: coverage expansion raises volume and therefore exposure, while value-based payment penalises unnecessary utilisation, so the two push in opposite directions and the honest answer names both rather than predicting confidently.