Eligible patients are those between the ages of 21 and 64 who present to the Emergency Department (ED) with a complaint of acute extremity pain of less than seven days duration in one or more extremities and for whom the clinician plans to discharge on an oral opioid for pain management. Patients will be excluded if they have ever taken methadone; have a chronic condition requiring frequent pain management such as sickle cell disease, fibromyalgia, or any neuropathy; report a history of an adverse reaction to any of the study medications; if they have taken prescribed opioids in the past 24 hours or if they report having ever taken recreational narcotics; if they have a medical condition that might affect their metabolism of opioid analgesics or acetaminophen; or if they take any medicine that might interact with one of the study medications. Patients will be referred by the attending physician or clinician and consent as well as the initial and follow-up data collections will be obtained by our team of trained bilingual (Spanish and English) salaried research associates who staff the ED 24 hours a day and 7 days a week.
In the first study, patients will be randomized to one of two experimental groups: hydrocodone 5mg / acetaminophen 500mg or codeine 30mg / acetaminophen 300mg. Randomization will be performed in blocks of 10 and determined by a sequence generated at http://www.randomization.com. The pharmacist working in an area inaccessible to ED staff will ensure proper blinding of the study by masking the medication and inserting it into unmarked gel capsules and filling any void with small amounts of lactose. A three-day supply (18 doses) of the blinded medications will be dispensed by the ED staff to the patient in the order determined by randomization accompanied by instructions to use one tablet of the medication every 4 hours as needed for pain and to avoid use of any other analgesics.
Using a power of 80%, a significance criterion of 0.5, and an estimated delta of 1.3 NRS units, a sample size of 85 patients per group. In order to account for those that do not end up taking the medicine, it is estimated that 120 patients per group will have to be enrolled.
Data will be collected on a standardized data collection instrument, entered by a trained data clerk, and reviewed and audited for accuracy and completeness. The investigators will calculate descriptive statistics for all variables: frequencies, means and standard deviations, medians and IQR, and proportions. Chi-square tests will be used to test differences between dichotomous variables, t-tests will be used to test mean differences. Multivariate models will be used if there are background variables that are unevenly distributed between the two groups. Variables associated with group membership with probability of 0.20 or less will be included in OLS multivariate regression or logistic regression models in order to test the role of group membership while accounting for chance baseline disparities. Interaction terms will be tested and dropped from the models if they were not statically significant at the 0.05 level. SPSS version 17 (Chicago, IL.) will be used to conduct all data analyses.
The second trial will be identical to the first trial with the exception of one of the study drugs, which will be oxycodone 5mg / acetaminophen 325 mg which will be compared to codeine 30mg / acetaminophen 300mg.
The third trial will be identical to the first and second trials with the exception of study drugs. This study will use oxycodone 5 mg/acetaminophen 325mg and hydrocodone 5mg/acetaminophen 325mg.
