A Randomized Control Trial of Nitrous Oxide Use in the Emergency Department

This will be a randomized control trial. All patients for whom an emergency provider determines that abscess drainage or laceration repair are necessary will be considered for enrollment. The provider must be willing to participate in the study and be certified to use Nitrous oxide. The investigators providers completed a training course to be certified to use Nitrous oxide. The patient must be age 18 or greater, consent to participate in the study, and be able to complete a satisfaction survey. If the patient states there is a possibility of pregnancy, the investigators will regress to our standard of care and perform a urine pregnancy test. Patients with a first trimester pregnancy will be excluded. Patients who agree to enroll in the study will be randomized to Nitrous oxide versus oxygen (to be administered through the same system), and will complete a survey after the procedure, as will the provider. There will be a separate block randomization for each procedure. Providers may select other agents of choice for analgesia/sedation, such as local anesthesia, and pain medications, based on their clinical judgment. If the patient appears to have inadequate analgesia during the procedure, the provider can elect to treat as is typically done (pain medications or local anesthetics) at their discretion. The survey will ask patients to give a self-reported pain score before and during their procedure (both to be completed after the procedure). It will also ask the patient/guardian and provider for a satisfaction level with the analgesia during the procedure. In addition, side effects and procedure start and stop time will also be documented. In addition to the survey, the investigators will document clinical characteristics including diagnosis, size of laceration or abscess, additional medications required and dosages, comorbidities, age, and gender. Other data that will be gathered will include which adjunctive analgesics/anesthetics were used.