How can an opioid prescribing rate change when the unit is a visit instead of a patient?
Two denominators in the same shoulder-injury study
A retrospective study of acute rotator cuff tear encounters in a large health system identified 9,376 eligible visits involving 5,310 patients. At 1,559 visits, or 16.6%, clinicians issued at least one opioid or benzodiazepine prescription. At the patient level, 1,096 people, or 20.6%, received at least one of those prescriptions during the study period.
Why the percentages differ
A person could have several eligible visits. In the visit calculation, every encounter counts separately. In the patient calculation, a person who received a qualifying prescription at any visit counts once. The investigators also reported the total number of prescriptions, which can exceed the number of visits involving prescribing because a visit may generate more than one order.
What an injury record can establish
A charted prescription provides evidence of an order issued during care. To reconstruct exposure before a particular event, distinguish the prescription date from any dispensing record and from evidence of actual use. The study’s percentages describe prescribing within its participating system. They do not measure how many patients took the drug, and the combined endpoint should be quoted as opioid or benzodiazepine prescribing.
Source study
Read the full paper in PubMed Central (PMC8129054).
