Fig. 1.
Forest plot of adjusted odds ratios from sensitivity analyses examining association between nerve blocks and the composite of fall-related injuries, seizures, cardiac arrest. or nerve injury: (1) adjusted for filling a preoperative opioid prescription (n = 48,391 without a block vs. n = 11,269 with a block); (2) adjusted for filling a prescription (anticoagulants, antiplatelet agents, antipsychotics, benzodiazepines, opioids, or dementia medications) in people more than 65 yr old (n = 68,243 without a block vs. 14,257 with a block); (3) adjusted for lab values (creatinine, hemoglobin, sodium, or potassium; n = 75,932 without a block vs. 15,631 with a block); (4a) one block received during the hospitalization versus none (n = 75,932 without a block vs. n = 15,068 with one block); (4b) more than one block received during the hospitalization versus none (n = 75,932 without a block vs. n = 563 with more than one block); and (5) nerve block in the emergency department versus no block (n = 75,932 without a block vs. n = 171 with a block in the emergency department).