Abstract
BACKGROUND: Most methodologically rigorous, ED-based, comparative effectiveness analgesic studies completed in the last two decades failed to find a clinically important difference between the comparators. We believe that many of these comparative effectiveness studies were biased towards the null hypothesis because some ED patients with intense pain will respond to relatively mild interventions. We hypothesized that including a run-in period would alter the results of an acute pain RCT. METHODS: We conducted a sequential, multiple-assignment, randomized study. Adults with acute moderate/severe musculoskeletal pain were randomized (3:1 ratio) to run-in period or no run-in. We administered 650 mg acetaminophen to run-in participants. Those run-in patients who reported insufficient relief one-hour later were randomized (1:1 ratio) to ibuprofen 800mg PO or ketorolac 20mg PO as were all participants randomized to no run-in. The primary outcome was achieving a clinically important improvement, defined as improvement ≥1.3 on a 0-10 scale. We built a logistic regression model including run-in/no run-in, ketorolac/ibuprofen, age and sex. RESULTS: Of 307 participants who received acetaminophen, 100 (32.6%) reported inadequate relief and were randomized to an NSAID. Of the 100 patients randomized to no run-in, 84/100 (84%) achieved the primary outcome versus 246/287 (86%) run-in participants (95% CI for difference = 2%:-7,10%). Among run-in participants who received an NSAID, 82/99(83%) achieved the primary outcome versus 84/100(84%) no run-in participants (p = 0.82). Among all ibuprofen participants, 44/49(90%) randomized to run-in and 42/50(84%) randomized to no run-in achieved the primary outcome. Among all ketorolac participants, 38/50(76%) randomized to run-in and 42/50 (84%) randomized to no run-in achieved the primary outcome. We observed the following results in a multivariable analysis: OR for ketorolac versus ibuprofen:0.60 (95% CI: 0.28, 1.28); OR for run-in versus no run-in:0.91(95% CI: 0.43, 1.93). CONCLUSIONS: Among patients with acute musculoskeletal pain, using an acetaminophen first strategy did not alter pain outcomes.