CONTINUING MEDICAL EDUCATION
Continuing Medical Education Activity in Academic Emergency Medicine CME Editor: Hal Thomas, MD Authors: Andrew K. Chang, MD, MS, Polly E. Bijur, PhD, Kevin G. Munjal, MD, MPH, and E. John Gallagher, MD Article Title: Randomized Clinical Trial of Hydrocodone/Acetaminophen Versus Codeine/Acetaminophen in the Treatment of Acute Extremity Pain After Emergency Department Discharge If you wish to receive free CME credit for this activity, please refer to the website: http://www.wileyhealthlearning.com/aem.
Accreditation and Designation Statement: Blackwell Futura Media Services designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit. Physicians should only claim credit commensurate with the extent of their participation in the activity. Blackwell Futura Media Services is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. Educational Objectives After completing this exercise the participant will be better able to describe the differences and similarities of two opioid analgesics. Activity Disclosures No commercial support has been accepted related to the development or publication of this activity. Faculty Disclosures: CME editor – Hal Thomas, MD: No relevant financial relationships to disclose. Authors Andrew K. Chang, MD, MS, Polly E. Bijur, PhD, Kevin G. Munjal, MD, MPH, and E. John Gallagher, MD. Have no relevant financial relationships to disclose. This manuscript underwent peer review in line with the standards of editorial integrity and publication ethics maintained by Academic Emergency Medicine. The peer reviewers have no relevant financial relationships. The peer review process for Academic Emergency Medicine is double-blinded. As such, the identities of the reviewers are not disclosed in line with
the standard accepted practices of medical journal peer review. Conflicts of interest have been identified and resolved in accordance with Blackwell Futura Media Services’s Policy on Activity Disclosure and Conflict of Interest. No relevant financial relationships exist for any individual in control of the content and therefore there were no conflicts to resolve. Instructions on Receiving Free CME Credit For information on applicability and acceptance of CME credit for this activity, please consult your professional licensing board. This activity is designed to be completed within an hour; physicians should claim only those credits that reflect the time actually spent in the activity. To successfully earn credit, participants must complete the activity during the valid credit period, which is up to two years from initial publication. Follow these steps to earn credit: • Log on to http://www.wileyhealthlearning.com • Read the target audience, educational objectives, and activity disclosures. • Read the article in print or online format. • Reflect on the article. • Access the CME Exam, and choose the best answer to each question. • Complete the required evaluation component of the activity. This activity will be available for CME credit for twelve months following its publication date. At that time, it will be reviewed and potentially updated and extended for an additional twelve months.
ORIGINAL CONTRIBUTION
Randomized Clinical Trial of Hydrocodone/Acetaminophen Versus Codeine/Acetaminophen in the Treatment of Acute Extremity Pain After Emergency Department Discharge Andrew K. Chang, MD, MS, Polly E. Bijur, PhD, Kevin G. Munjal, MD, MPH, and E. John Gallagher, MD
Abstract Objectives: The objective was to test the hypothesis that hydrocodone/acetaminophen (Vicodin [5/500]) provides more efficacious analgesia than codeine/acetaminophen (Tylenol #3 [30/300]) in patients discharged from the emergency department (ED). Both are currently Drug Enforcement Administration (DEA) Schedule III narcotics. Methods: This was a prospective, randomized, double-blind, clinical trial of patients with acute extremity pain who were discharged home from the ED, comparing a 3-day supply of oral hydrocodone/ acetaminophen (5 mg/500 mg) to oral codeine/acetaminophen (30 mg/300 mg). Pain was measured on a valid and reproducible verbal numeric rating scale (NRS) ranging from 0 to 10, and patients were contacted by telephone approximately 24 hours after being discharged. The primary outcome was the between-group difference in improvement in pain at 2 hours following the most recent ingestion of the study drug, relative to the time of phone contact after ED discharge. Secondary outcomes compared side-effect profiles and patient satisfaction. Results: The median time from ED discharge to follow-up was 26 hours (interquartile range [IQR] = 24 to 39 hours). The mean NRS pain score before the most recent dose of pain medication after ED discharge was 7.6 NRS units for both groups. The mean decrease in pain scores 2 hours after pain medications were taken were 3.9 NRS units in the hydrocodone/acetaminophen group versus 3.5 NRS units in the codeine/acetaminophen group, for a difference of 0.4 NRS units (95% confidence interval [CI] = –0.3 to 1.2 NRS units). No differences were found in side effects or patient satisfaction. Conclusions: Both medications decreased NRS pain scores by approximately 50%. However, the oral hydrocodone/acetaminophen failed to provide clinically or statistically superior pain relief compared to oral codeine/acetaminophen when prescribed to patients discharged from the ED with acute extremity pain. Similarly, there were no clinically or statistically important differences in side-effect profiles or patient satisfaction. If the DEA reclassifies hydrocodone as a Schedule II narcotic, as recently recommended by its advisory board, our data suggest that the codeine/acetaminophen may be a clinically reasonable Schedule III substitute for hydrocodone/acetaminophen at ED discharge. These findings should be regarded as tentative and require independent validation in similar and other acute pain models. ACADEMIC EMERGENCY MEDICINE 2014; 21:228–235 © 2014 by the Society for Academic Emergency Medicine
From the Department of Emergency Medicine, Albert Einstein College of Medicine, Montefiore Medical Center (AKC, PEB, EJG), Bronx, NY; and the Department of Emergency Medicine, Mount Sinai Medical Center (KGM), New York, NY. Received July 11, 2013; revisions received September 3 and September 13, 2013; accepted September 16, 2013. The authors have no relevant financial information or potential conflicts of interest to disclose. Supervising Editor: Mark B. Mycyk, MD. Address for correspondence: Andrew Chang, MD, MS; e-mail: achang@montefiore.org. Reprints will not be available.
228
ISSN 1069-6563 PII ISSN 1069-6563583
© 2014 by the Society for Academic Emergency Medicine doi: 10.1111/acem.12331
ACADEMIC EMERGENCY MEDICINE • March 2014, Vol. 21, No. 3 • www.aemj.org
O
ver 44 million U.S. emergency department (ED) visits per year are for chief complaints related to pain. Approximately 29%, or 12.8 million, of these visits result in prescriptions for an opioid analgesic at discharge.1 Three commonly prescribed oral opioids at ED discharge include hydrocodone/acetaminophen (Vicodin, Norco), oxycodone/acetaminophen (Percocet), and codeine/acetaminophen (Tylenol #3). Of these, hydrocodone/acetaminophen is the most commonly prescribed while codeine/acetaminophen is the least prescribed.2 The popularity of hydrocodone/acetaminophen postdischarge is supported by two small clinical trials,3,4 and several other studies that compare the efficacy of commonly prescribed oral opioids in different settings.5–9 There has been increasing concern about the abuse of oral opioids in patients discharged from the ED.10,11 The efficacy of hydrocodone versus codeine has not been adequately studied among discharged adult ED patients with acute pain. Using the clinical model of acute extremity pain, we designed a randomized trial comparing these two commonly prescribed analgesics to determine if the strong preference for use of hydrocodone/acetaminophen over codeine/acetaminophen was supported by evidence. This question has become particularly relevant in light of the Federal Drug Administration (FDA)’s recent advisory committee recommendation to reclassify hydrocodone from a Schedule III to a Schedule II narcotic, while leaving codeine unchanged as Schedule III. The objective of this investigation was to compare the outpatient analgesic efficacy, side-effect profile, and patient satisfaction of hydrocodone/acetaminophen versus codeine/acetaminophen for treatment of acute extremity pain in nonelderly adult patients discharged from the ED. We hypothesized that hydrocodone/acetaminophen would provide more efficacious pain relief than codeine/acetaminophen (primary hypothesis), with a lower incidence of side effects and greater patient satisfaction (secondary hypotheses). METHODS Study Design This was a prospective, randomized, double-blind, clinical trial comparing efficacy, side effects, and patient satisfaction of hydrocodone/acetaminophen (5 mg/500 mg) to those of codeine/acetaminophen (30 mg/300 mg) in patients discharged from the ED. The study was approved by the Montefiore Medical Center Institutional Review Board and registered at www.clinicaltrials.gov (NCT01402375). Each participant gave written informed consent in either English or Spanish, depending on his or her preference. Study Setting and Population The study recruited patients from an academic urban ED with an annual census of >100,000 adults, many of whom were underserved minorities. Patients were 21 to 64 years of age presenting with acute extremity pain and for whom the provider planned to discharge with a short course of oral opioids for outpatient pain management. “Acute” was defined as less than 7 days’ duration.
229
“Extremity” was defined as distal to and including the shoulder or hip joint. Exclusion criteria were as follows: history of prior or current methadone use; chronic condition requiring frequent or continuous pain management (e.g., sickle cell, fibromyalgia, or chronic neuropathy); allergy to hydrocodone, codeine, or acetaminophen; history of prior or current recreational opioid use; ingestion of an oral opioid in the past 24 hours; medical conditions that might affect opioid or acetaminophen metabolism (e.g., hepatitis, renal insufficiency, hypothyroidism, hyperthyroidism, or adrenal disease); or use of medication that might interact with opioids or acetaminophen (e.g., selective serotonin or norepinephrine reuptake inhibitors, cyclic antidepressants, antipsychotics, antimalaria medications, amiodarone, dronedarone, diphenhydramine, celecoxib, ranitidine, cimetidine ritonavir, terbinafine, or St. John’s Wort). Patients were recruited and enrolled by trained research associates (RAs), fluent in English and Spanish, 24 hours per day, 7 days per week, from January 2012 through August 2012. Study Protocol Patients were randomized to receive a 3-day course of hydrocodone/acetaminophen (5 mg/500 mg) or codeine/ acetaminophen (30 mg/300 mg). These doses were selected to reflect the most commonly prescribed formulations of each of these combination oral opioids, commonly known as Vicodin and Tylenol #3, respectively. Allocation assignments were generated by an online random number generator (www.randomization. com) and placed in sealed opaque envelopes to be opened in sequential order by RAs. Patients, treating physicians, and RAs were all blinded to study allocation. The pharmacist, working in an area distant from and inaccessible to ED staff, ensured proper allocation concealment by inserting study medications into identical unmarked opaque gel capsules, filling any void with lactose to equalize weight. After examining the large size of a single capsule, we decided to divide the dosage in half to ensure patients’ ability to swallow smaller prepared capsules. Thus, each capsule contained hydrocodone/acetaminophen (2.5 mg/250 mg) or codeine/ acetaminophen (15 mg/150 mg), and one dose of study medication equaled two capsules. The clinician ordered and dispensed a 3-day supply of the blinded medications to the patient along with verbal and written instructions. Patients received a 3-day supply (36 tablets or 18 doses) and were instructed to take two tablets (one dose) every 4 hours as needed for pain (maximum daily dose = six doses = 12 capsules). Patients were also strongly advised not to take any other analgesic medications, including over-the-counter or topical medications, during the study period, which concluded after the follow-up phone call approximately 24 hours following ED discharge. A structured patient diary was provided at discharge to assist patients in recording pain numeric rating scale (NRS) levels and side effects during the study period. Although NRS pain scores were obtained during the ED visit and prior to discharge, management of the patient while in the ED was at the ED attending physician’s discretion and was independent of the study
230
Chang et al. • HYDROCODONE/ACETAMINOPHEN VS. CODEINE/ACETAMINOPHEN AFTER ED DISCHARGE
protocol. The study medications were given to patients at the time of ED discharge, with the patient to be contacted approximately 24 hours later. Measurements Primary outcome data were obtained via follow-up phone call targeted for 24 hours after ED discharge. The first attempt to reach the patient was made at 24 hours unless the call was due between 10 p.m. and 8 a.m., during which we reasoned that many patients and their families might regard a phone call as intrusive. Thus, calls scheduled during these hours were made after 8 a.m. the following morning. Any patient who could not be reached at 24 hours or after 8 a.m. the following day received a minimum of twice-daily phone calls between 8 a.m. and 10 p.m. for up to 72 hours postdischarge, at which point the patient was classified as lost to follow-up. At least two phone numbers per patient were collected by the RAs at the time of study entry, one of which was almost invariably a cell phone, whose number was confirmed via a phone call made by the RA prior to discharge, and recorded as a required field to be completed prior to patient enrollment on the data collection instrument. The primary outcome was the between-group difference in improvement in mean (with standard deviation [SD]) NRS pain scores, measured at 2 hours following the most recent ingestion of the study drug prior to the time of phone contact. If the patient was reached less than 2 hours after the most recent dose of pain medicine, the survey was temporarily halted. The patient was then called again once the 2-hour period had elapsed and the survey resumed. Pain was measured on a validated and reproducible verbal 11-point NRS,12–14 based on patient recall. Patients were not informed of their pain scores during their ED visits. Secondary outcomes included incidence of prespecified side effects and overall patient satisfaction assessed dichotomously on two axes: 1) very satisfied/satisfied versus very unsatisfied/unsatisfied and 2) a simple yes/ no response to the question “Would you want the same pain medication again if you came to the ED with a similar degree of pain in the future?” Data Analysis Data were collected on a standardized data collection instrument. A trained data clerk entered the data in SPSS (version 17, IBM SPSS, Inc., Armonk, NY) on an ongoing weekly basis. To minimize transcription error, double entry of the entire data set was performed independently by a second trained clerk. All differences were reconciled by referral to the original data collection instrument. In addition, 10% of the data collection instruments were randomly audited to confirm data quality. Descriptive statistics were calculated for all variables. We visually examined the data to determine the appropriateness of describing continuous data with means or medians and the decision to use parametric or nonparametric tests for statistical inference. Means of variables that were symmetrically distributed and proportions are presented with 95% confidence intervals (CIs). We described variables that were clearly not symmetrically
distributed with medians and interquartile ranges (IQRs). t-tests and chi-square or Fisher’s exact test were used to assess differences between means and dichotomous variables, respectively. The Wilcoxon rank-sum test was used to compare medians. SPSS was also used for all analyses. To test the primary hypothesis that hydrocodone/ acetaminophen would provide superior pain relief compared to codeine/acetaminophen, we used the previously validated minimum clinically significant difference of 1.3 NRS units as the smallest improvement in pain score clinically worth detecting.12–14 With a standard two-tailed alpha of 0.05, power of 80% to detect a difference of 1.3 NRS units or greater, and a SD of 3.0, we estimated that 170 patients who took at least one dose of the prescribed medication (85 in each group) would be required. Based on a prior study with similar design, we estimated that approximately 30% of patients would fail to take their medication as prescribed.6 Thus we estimated that as many as 120 patients in each group would be needed to achieve the sample size of at least 85 patients per group. RESULTS The CONSORT diagram summarizing the flow of patients through the trial is shown in Figure 1. The RAs evaluated 849 patients for eligibility. The most common reasons for exclusion were pain duration greater than 7 days and physician judgment that the patient did not need opioid analgesics. A total of 240 patients were randomized. Data from the first 35 patients were not included in the analyses because we made changes to the data collection instrument in response to the RAs’ initial experience with it. We treated these data as a pilot to estimate the proportion of patients who were expected to take the prescribed medication and the ability of patients to recall pain before and after taking the medication. Twenty-four patients (12 in each group) did not take the prescribed discharge medicine prior to the followup phone call and therefore could not provide a 2-hour postingestion change in NRS pain score. Thus, the final sample consisted of 88 patients who received hydrocodone/acetaminophen and 93 who received codeine/acetaminophen. No patients were lost to follow-up. As shown in Table 1, the two comparison groups had similar demographic characteristics, diagnoses, use of analgesics before the visit, time to follow-up, and NRS pain scores while in the ED. The median time from discharge to follow-up was 26 hours (IQR = 24 to 39 hours) and this did not differ significantly between the two groups. Table 2 shows the main outcomes. The mean NRS pain scores for both groups before the most recent dose of pain medication were 7.6 NRS units. This intensity of pain score was expected as a high level of pain would prompt the patient to take the prescribed study analgesic pain medicine. The mean change in NRS pain scores 2 hours after the most recent dose of study medication was 3.9 NRS units for the hydrocodone/acetaminophen arm versus 3.5 NRS units for the codeine/acetaminophen arm, for a difference of 0.4 NRS units (95% CI = –0.3 to 1.2 NRS units). Although both
ACADEMIC EMERGENCY MEDICINE • March 2014, Vol. 21, No. 3 • www.aemj.org
231
Assessed for eligibility (N=849)
Excluded (n=609) Pain present more than 7 days (n=150) Opioid analgesia not warranted (n=123) Nontraumatic pain (n=70) Kidney, liver, thyroid, or adrenal disease (n=66) Took pain medication within 24 hours (n=62) Concurrent medication warranted (n=61) Allergic to study medication (n=39) Refused consent (n=32) Admitted to hospital (n=6)
Randomized (n=240)
Codeine/acetaminophen (n=120)
Hydrocodone/acetaminophen (n=120)
Analyzed for primary outcome (n=93)
Analyzed for primary outcome (n=88)
Pilot data (n=15)
Pilot data (n=20) Did not take study medication (n=12)
Did not take study medication (n=12)
Figure 1. CONSORT diagram.
Table 1 Baseline characteristics Characteristic
Hydrocodone/Acetaminophen (n = 88)
Female sex, n (%) Race/ethnicity, n (%) Hispanic African American Other Age (yr), mean (SD) NRS pain intensity on ED arrival, median (IQR) NRS pain intensity on discharge, median (IQR) Time (hours) to follow-up, median (IQR) Diagnosis, n (%) Sprain/strain Extremity fracture Other Took analgesic prior to presentation, n (%) No Yes
Codeine/Acetaminophen (n = 93)
38 (43)
48 (52)
56 27 5 34 9 7 27
69 18 6 37 9 7 25
(64) (31) (6) (12) (8 to 10) (5 to 8) (24 to 42)
(74) (19) (7) (11) (8 to 10) (5 to 8) (24 to 36)
40 (46) 20 (23) 28 (32)
54 (58) 17 (18) 22 (24)
63 (72) 24 (28)
63 (70) 27 (30)
IQR = interquartile range; NRS = numeric rating scale.
Table 2 Outcomes by Group
Outcome NRS before most recent dose, mean (SD) NRS 2 hours after most recent dose, mean (SD) Change in NRS before–after most recent dose, mean (SD) Satisfied with analgesic, n/N* (%) Want same analgesic again, n/N* (%)
Hydrocodone/Acetaminophen (n = 88)
Codeine/Acetaminophen (n = 93)
7.6 (1.7) 3.6 (2.8) 3.9 (2.5)
7.6 (1.8) 4.1 (2.9) 3.5 (2.6)
0.0 (–0.5 to 0.5) –0.5 (–1.3 to 0.4) 0.4 (–0.3 to 1.2)
72/87 (83) 59/85 (69)
66/93 (71) 66/93 (71)
11.8% (–0.7 to 24) –1.5% (–15 to 12)
Denominators are different due to missing data from unanswered questions.
Difference (95% CI)
232
Chang et al. • HYDROCODONE/ACETAMINOPHEN VS. CODEINE/ACETAMINOPHEN AFTER ED DISCHARGE
groups of analgesic medicine reduced the NRS pain intensity score by approximately 50%, there was neither a clinically nor a statistically significant difference between the two study medicines when compared to an evidence-based, prespecified change in NRS of at least 1.3.12–14 Three patients in the hydrocodone/acetaminophen group and five patients in the codeine/acetaminophen group took additional pain medication beyond what was prescribed to them for this study. This occurred most commonly because the study medication did not adequately relieve the patient’s pain. There were no differences in satisfaction and percentage who would want the same analgesic in the future, although the study was not powered based on a satisfaction metric. Table 3 shows adverse events by group. There were no clinical or statistically important between-group differences in any adverse event category. DISCUSSION There has been significant recent controversy over the prescribing of oral opioids at ED discharge driven by assertions that they are overprescribed, diverted, and abused.10 As a result, the U.S. Drug Enforcement Administration (DEA) recently made a request that hydrocodone combination products be rescheduled.15 A 2-day meeting was held in January 2013 by the FDA’s Drug Safety and Risk Management Advisory Committee, in which members voted 19 to 10 that these products should be rescheduled from Schedule III to Schedule II under the Controlled Substances Act.16,17 Although this has not yet occurred, historically the FDA usually follows the recommendations of its advisory panels. Some states are already making these changes on their own. New York’s I-STOP legislation was signed in August 2012 and took effect in February 2013, which changed hydrocodone from Schedule III to Schedule II under the controlled substance schedules in Section 3306 of the New York State Public Health Law.18 In addition, many EDs have subsequently been asked to
follow voluntary ED discharge opioid prescribing guidelines.19 Because Schedule III medications may have less potential for abuse and lead to lower physical and psychological dependence compared to Schedule II drugs,20 Schedule III drugs can be prescribed by faxing a prescription or calling a pharmacist and are eligible for up to five refills within 6 months. In contrast, Schedule II drugs require handwritten prescriptions that cannot be phoned or faxed and may not be refilled. Thus, the reclassification of hydrocodone from Schedule III to Schedule II may have a substantial effect on outpatient oral opioid prescribing patterns simply by increasing legal impediments to the use of hydrocodone and its combinations, while leaving current restrictions governing prescription of codeine and codeine-containing combinations unaltered. Commonly prescribed oral opioids after ED discharge include several acetaminophen-opioid combinations: hydrocodone/acetaminophen (Vicodin and Norco), oxycodone/acetaminophen (Percocet), and codeine/acetaminophen (Tylenol with codeine). Hydrocodone/ acetaminophen and oxycodone/acetaminophen are generally thought to be stronger opioids than codeine/acetaminophen. This is why we selected a traditional superiority randomized clinical trial design, hypothesizing that the hydrocodone combination would provide meaningfully better pain relief than the codeine combination. However, our data indicate that these two agents appear to be clinically and statistically similar with respect to analgesic efficacy, side-effect profiles, and patient satisfaction. Hydrocodone/acetaminophen is the most common opioid combination prescribed on discharge from U.S. EDs and is used approximately three times more often than oxycodone/acetaminophen,2 which is a Schedule II drug. The popularity of hydrocodone over oxycodone may be related to the fact that hydrocodone is currently a Schedule III medication while oxycodone is a Schedule II medication, thus making the former substantially easier for physicians to prescribe. Codeine/acetaminophen, a Schedule III drug, is presently as easy to prescribe as
Table 3 Adverse Side Effects by Group Side Effect Nausea Vomiting Constipation Diarrhea Pruritus Rash Dizziness Drowsiness Confusion Number of side effects 0 1 2+
Hydrocodone/Acetaminophen 9/87 2/87 0/88 0/87 3/87 1/87 12/87 25/88 2/86
Codeine/Acetaminophen
(10) (2) (0) (0) (3) (1) (14) (28) (2)
51/86 (59) 24/86 (28) 11/86 (13)
Data are reported as n/N (%) Denominators are different due to missing data from unanswered questions.
9/92 1/92 0/93 5/93 3/93 0/92 5/93 27/93 1/93
Difference% (95% CI)
(10) (1) (0) (5) (3) (0) (5) (29) (1)
0.6 (–8 to 10) 1.2 (–4 to 6) 0 (–3 to 3) –5.4 (–11 to 0.1) 0.2 (–6 to 6) 1.1 (–3 to 5) 8.4 (–0.5 to 17) –0.6 (–14 to 13) 1.3 (–3 to 6)
52/91 (57) 30/91 (33) 9/91 (10)
2.2 (–12 to 17) –5.1 (–18 to 9) 2.9 (–7 to 12)
ACADEMIC EMERGENCY MEDICINE • March 2014, Vol. 21, No. 3 • www.aemj.org
hydrocodone/acetaminophen, but for unclear reasons, is not used as frequently. Although the Cochrane Collaboration has produced several reviews describing the analgesic effectiveness of codeine, both with and without acetaminophen,21,22 there are no Cochrane reviews or meta-analyses targeting the efficacy of hydrocodone. We could find only a single paper supporting the preferential prescribing of hydrocodone/acetaminophen over codeine/acetaminophen. Forbes et al.6 compared a single dose of hydrocodone/acetaminophen (7.5 mg/500 mg) to a single dose of codeine/acetaminophen (30 mg/300 mg) for postoperative oral surgery pain and found significantly better pain relief among the patients treated with hydrocodone, with no significant difference in side effects. Although opioid equivalencies are difficult to estimate, particularly with codeine because of the interindividual variation in metabolism of this opioid to its active metabolite, morphine, consensus data indicate that 5 mg of hydrocodone given orally is roughly the analgesic equivalent of 30 mg of oral codeine.23 If this estimate is correct, then the validity of Forbes’ single-dose study may have been compromised by performance bias due to administration of 50% more opioid analgesic-equivalent plus 200 mg more of acetaminophen in the hydrocodone arm than was administered to patients in to the codeine arm. Thus, because the analgesic “exposure” in the two groups may have been imbalanced—thus producing an inherently “unfair comparison”—one cannot confidently conclude superiority, noninferiority, nor equivalence due to a methodologic limitation. The only other publication we could find comparing hydrocodone and codeine was a well-conducted clinical trial from the University of Pittsburgh by Turturro et al.,4 which found no significant difference in the mean or median pain scores between the two groups at various intervals over the 48 hours following ED discharge for acute musculoskeletal pain. This was consistent with our 24-hour follow-up data. In contrast to the study by Forbes et al., Turturro et al. used hydrocodone/acetaminophen (5 mg/500 mg) versus codeine/ acetaminophen (30 mg/500 mg), which is consistent with an unbiased comparison of the two combination analgesics because each arm of the trial received analgesic-equivalent doses of both opioids and acetaminophen. We chose not to do this because the combination of codeine/acetaminophen (30 mg/500 mg) is not available in the United States. Two other randomized controlled trials were found comparing hydrocodone versus codeine for chronic pain.9,24 Rodriguez et al.,24 similar to Turturro et al.,4 found no significant difference in either analgesia or incidence of side effects in a study of 121 patients comparing equianalgesic doses of 5 mg of hydrocodone versus 30 mg of codeine, both in combination with 500 mg of acetaminophen. Palangio et al.8 in a three-armed (hydrocodone/ibuprofen [7.5 mg/200 mg]; hydrocodone/ibuprofen [15 mg/400 mg]; and codeine/ acetaminophen [60 mg/600 mg]) study design with 469 patients found greater analgesia in the hydrocodone/ibuprofen (15 mg/400 mg) group versus the codeine/acetaminophen (60 mg/600 mg) group. As in the only other paper suggesting analgesic superiority of
233
hydrocodone over codeine by Forbes et al.6 in the setting of acute pain, discussed above, validity is likely to be compromised by the insinuation of performance bias. In the study by Palangio et al., this bias appears in the form of allocation of nonequianalgesic doses of drugs to the groups under comparison (e.g., 50% higher dose of hydrocodone than would be considered equianalgesic to 60 mg codeine, further confounded by the presence of ibuprofen in one arm). LIMITATIONS Because 24 enrolled patients (12 in each group) did not take the study medication as directed, no data could be obtained on these patients during the 24-hour followup. Thus, we were not able to perform an intention-totreat analysis and instead performed a per-protocol analysis. The intention-to-treat design has the distinct advantage of diminishing alpha error, i.e., the probability of obtaining a false-positive finding, by driving results toward the null. In the case of a negative study result obtained via a per-protocol analysis, however, the issue of alpha error is moot. Furthermore, application of an intention-to-treat analysis can only make the findings “more negative” in a traditional superiority trial such as this one, which has failed to identify a meaningful difference between the comparison groups in the face of a sufficiently large sample size to provide traditionally adequate power of 80%. Thus, although use of a per-protocol analysis might be viewed by some as a limitation of any randomized controlled trial that employs this method, in the setting of a negative study such as this one, this would only serve to reinforce our findings. Both hydrocodone/acetaminophen and codeine/acetaminophen come in several different concentrations. Although Norco (5 mg hydrocodone/300 mg acetaminophen) is frequently prescribed in the United States, we chose to use what we believed to be the most commonly prescribed combinations, which were 5 mg hydrocodone/500 mg acetaminophen and 30 mg codeine/ 300 mg acetaminophen. Although improbable, we recognized that the additional 200 mg of acetaminophen might constitute a limitation in the form of performance bias favoring hydrocodone/acetaminophen (5 mg/ 500 mg). However, our findings indicate that, even if this design feature did contribute to the slight advantage associated with hydrocodone/acetaminophen (5 mg/500 mg), the delta that we identified in NRS score is only 0.4 (95% CI = –0.3 to 1.2). This is not only neither statistically nor clinically significant, but is also substantially closer to the numerical null of 0 than to the minimum clinically significant difference, or experimental null, of 1.3 NRS units. Furthermore, the fact that the upper limit of the 95% CI of 1.2 NRS units falls short of the minimum a priori criteria of 1.3 makes it highly unlikely that the “true” difference between the study arms is as large or larger than this threshold value required for declaring a clinically meaningful difference between the two drug combinations under comparison. Although the chosen doses of hydrocodone/acetaminophen and codeine/acetaminophen typically come in a single pill, we decided to split the dose into two capsules. We did this out of concern that the size of the
234
Chang et al. • HYDROCODONE/ACETAMINOPHEN VS. CODEINE/ACETAMINOPHEN AFTER ED DISCHARGE
smallest single capsule containing the full dose of medication that could be prepared by our research pharmacist was far too large for many patients to comfortably swallow and would therefore produce adherence bias (another form of performance bias), thus unbalancing randomly allocated exposure to the two analgesics under comparison. We therefore split the dose into two smaller capsules and instructed patients to take two pills at a time, just as they might take two acetaminophen, two aspirin, or two other nonsteroidals for pain. We recognized that this might also contribute to adherence bias, but we judged it to be a less likely impediment to ingestion than a single pill of intimidating size. Fortunately, the 24 patients who did not take the study medication exactly as prescribed following ED discharge were evenly distributed across the hydrocodone/acetaminophen and codeine/acetaminophen comparison groups and did not diminish our analyzable sample size below the number required to achieve a statistical power of 80% to detect the target delta of 1.3 NRS units,12–14 if a between-group difference in pain improvement of this magnitude were in fact present. Oral opioids are typically prescribed as a range of one or two tablets every 4 to 6 hours as needed. Given the large number of possible combinations such a range would entail, we decided to fix the dose at one tablet every 4 hours as needed, which is within the range of how the prescriptions are typically written. We chose to start at the lower range of dosing (one tablet) since this was our first study examining oral opioids and we wished to proceed in a stepwise manner, investigating potentially higher doses (e.g., two tablets) depending on the outcomes of the current study. Indeed, the chosen dosages resulted in an approximately 50% reduction in pain intensity for both groups of analgesics. And finally, we were cognizant of recent case reports regarding certain patients who hypermetabolize codeine, leading to toxic morphine levels and associated respiratory depression.25 The time of the pain scores varied depending on when the patient took the medication. There may also be recall bias because of the difference in time between when patients took the study medications and when they answered the follow-up questions, although it is likely that this would have affected both groups equally. Since only adults between the ages of 21 and 64 years of age were studied, our findings cannot be extrapolated to children, adolescents, or the elderly. The study was also carried out in a single center composed of a largely minority population of urban, poor, and medically underserved patients. Because there have been reports of oligoanalgesia associated with ethnicity and socioeconomic status, and because there are racial differences in drug metabolism, the results of this study may not generalize to other populations in the United States. CONCLUSIONS In a randomized, controlled, double-blind clinical trial, we found that hydrocodone/acetaminophen (Vicodin [5/500]) did not provide clinically or statistically superior outpatient pain relief compared to codeine/ acetaminophen (Tylenol #3 [30/300]), when assessed at
24 hours following discharge from the ED in patients presenting with acute extremity pain. Similarly, there were no clinically or statistically important differences in side-effect profiles or patient satisfaction. If the Drug Enforcement Administration reclassifies hydrocodone as a Schedule II narcotic, as recently recommended by its advisory board, our data suggest codeine/acetaminophen (30/300 mg) may be a clinically reasonable Schedule III substitute for the currently more frequently prescribed hydrocodone/acetaminophen (5/500 mg) at ED discharge. These findings should be viewed as preliminary and require independent validation in similar and other acute pain models. References 1. Pletcher MJ, Kertesz SG, Kohn MA, Gonzales R. Trends in opioid prescribing by race/ethnicity for patients seeking care in US emergency departments. JAMA 2008;299:70–8. 2. Niska R, Bhuiya F, Xu J. National Hospital Ambulatory Medical Care Survey: 2007 emergency department summary. Natl Health Stat Report 2010:1–31. 3. Marco CA, Plewa MC, Buderer N, Black C, Roberts A. Comparison of oxycodone and hydrocodone for the treatment of acute pain associated with fractures: a double-blind, randomized, controlled trial. Acad Emerg Med 2005;12:282–8. 4. Turturro MA, Paris PM, Yealy DM, Menegazzi JJ. Hydrocodone versus codeine in acute musculoskeletal pain. Ann Emerg Med 1991;20:1100–3. 5. Charney RL, Yan Y, Schootman M, Kennedy RM, Luhmann JD. Oxycodone versus codeine for triage pain in children with suspected forearm fracture: a randomized controlled trial. Pediatr Emerg Care 2008;24:595–600. 6. Forbes JA, Bates JA, Edquist IA, et al. Evaluation of two opioid-acetaminophen combinations and placebo in postoperative oral surgery pain. Pharmacotherapy 1994;14:139–46. 7. Litkowski LJ, Christensen SE, Adamson DN, Van Dyke T, Han SH, Newman KB. Analgesic efficacy and tolerability of oxycodone 5 mg/ibuprofen 400 mg compared with those of oxycodone 5 mg/acetaminophen 325 mg and hydrocodone 7.5 mg/acetaminophen 500 mg in patients with moderate to severe postoperative pain: a randomized, doubleblind, placebo-controlled, single-dose, parallel-group study in a dental pain model. Clin Ther 2005;27: 418–29. 8. Palangio M, Damask MJ, Morris E, et al. Combination hydrocodone and ibuprofen versus combination codeine and acetaminophen for the treatment of chronic pain. Clin Ther 2000;22:879–92. 9. Palangio M, Morris E, Doyle RT Jr, Dornseif BE, Valente TJ. Combination hydrocodone and ibuprofen versus combination oxycodone and acetaminophen in the treatment of moderate or severe acute low back pain. Clin Ther 2002;24:87–99. 10. Cantrill SV, Brown MD, Carlisle RJ, et al. Clinical policy: critical issues in the prescribing of opioids for adult patients in the emergency department. Ann Emerg Med 2012;60:499–525.
ACADEMIC EMERGENCY MEDICINE • March 2014, Vol. 21, No. 3 • www.aemj.org
11. Wilsey BL, Fishman SM, Ogden C. Prescription opioid abuse in the emergency department. J Law Med Ethics 2005;33:770–82. 12. Bijur PE, Latimer CT, Gallagher EJ. Validation of a verbally administered numerical rating scale of acute pain for use in the emergency department. Acad Emerg Med 2003;10:390–2. 13. Holdgate A, Asha S, Craig J, Thompson J. Comparison of a verbal numeric rating scale with the visual analogue scale for the measurement of acute pain. Emerg Med (Fremantle) 2003;15:441–6. 14. Todd KH. Clinical versus statistical significance in the assessment of pain relief. Ann Emerg Med 1996;27:439–41. 15. Kuehn BM. FDA committee: more restrictions needed on hydrocodone combination products. JAMA 2013;309:862. 16. Kartzinel R. Guidelines for scheduling drugs under the controlled substances act [proceedings]. Psychopharmacol Bull 1981;17:40–2. 17. McCarthy M. FDA panel recommends tighter rules for prescribing opioids. BMJ 2013;346:f1012. 18. State of New York Department of Health. Frequently asked questions Part C Chapter 447 Laws of 2012 (Controlled Substance Schedule Changes). Available at: http://www.health.ny.gov/professionals/narcotic/ laws_and_regulations/part_c-chapter_447-laws_of_ 2012-faq.htm. Accessed Dec 2, 2013.
235
19. Juurlink DN, Dhalla IA, Nelson LS. Improving opioid prescribing: the New York City recommendations. JAMA 2013;309:879–80. 20. Drug Enforcement Administration. Department of Justice. Dispensing controlled substances for the treatment of pain issuance of multiple prescriptions for schedule II controlled substances. J Pain Palliat Care Pharmacother 2007;21:75–88. 21. Derry S, Moore RA, McQuay HJ. Single dose oral codeine, as a single agent, for acute postoperative pain in adults. Cochrane Database Syst Rev 2010: CD008099. 22. Toms L, Derry S, Moore RA, McQuay HJ. Single dose oral paracetamol (acetaminophen) with codeine for postoperative pain in adults. Cochrane Database Syst Rev 2009:CD001547. 23. Drugs for pain. Med Letter (pay to access) 2004;2:47–54. 24. Rodriguez RF, Castillo JM, Del Pilar Castillo M, et al. Codeine/acetaminophen and hydrocodone/acetaminophen combination tablets for the management of chronic cancer pain in adults: a 23-day, prospective, double-blind, randomized, parallel-group study. Clin Ther 2007;29:581–7. 25. Friedrichsdorf SJ, Nugent AP, Strobl AQ. Codeineassociated pediatric deaths despite using recommended dosing guidelines: three case reports. J Opioid Manag 2013;9:151–5.
Continuing Medical Education Activity in Academic Emergency Medicine CME Editor: Hal Thomas, MD Authors: Andrew K. Chang, MD, MS, Polly E. Bijur, PhD, Kevin G. Munjal, MD, MPH, and E. John Gallagher, MD Article Title: Randomized Clinical Trial of Hydrocodone/Acetaminophen Versus Codeine/Acetaminophen in the Treatment of Acute Extremity Pain After Emergency Department Discharge If you wish to receive free CME credit for this activity, please refer to the website: http://www.wileyhealthlearning.com/aem.
Accreditation and Designation Statement: Blackwell Futura Media Services designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit. Physicians should only claim credit commensurate with the extent of their participation in the activity. Blackwell Futura Media Services is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. Educational Objectives After completing this exercise the participant will be better able to describe the differences and similarities of two opioid analgesics. Activity Disclosures No commercial support has been accepted related to the development or publication of this activity. Faculty Disclosures: CME editor – Hal Thomas, MD: No relevant financial relationships to disclose. Authors Andrew K. Chang, MD, MS, Polly E. Bijur, PhD, Kevin G. Munjal, MD, MPH, and E. John Gallagher, MD. Have no relevant financial relationships to disclose. This manuscript underwent peer review in line with the standards of editorial integrity and publication ethics maintained by Academic Emergency Medicine. The peer reviewers have no relevant financial relationships. The peer review process for Academic Emergency Medicine is double-blinded. As such, the identities of the reviewers are not disclosed in line with
the standard accepted practices of medical journal peer review. Conflicts of interest have been identified and resolved in accordance with Blackwell Futura Media Services’s Policy on Activity Disclosure and Conflict of Interest. No relevant financial relationships exist for any individual in control of the content and therefore there were no conflicts to resolve. Instructions on Receiving Free CME Credit For information on applicability and acceptance of CME credit for this activity, please consult your professional licensing board. This activity is designed to be completed within an hour; physicians should claim only those credits that reflect the time actually spent in the activity. To successfully earn credit, participants must complete the activity during the valid credit period, which is up to two years from initial publication. Follow these steps to earn credit: • Log on to http://www.wileyhealthlearning.com • Read the target audience, educational objectives, and activity disclosures. • Read the article in print or online format. • Reflect on the article. • Access the CME Exam, and choose the best answer to each question. • Complete the required evaluation component of the activity. This activity will be available for CME credit for twelve months following its publication date. At that time, it will be reviewed and potentially updated and extended for an additional twelve months.
ORIGINAL CONTRIBUTION
Randomized Clinical Trial of Hydrocodone/Acetaminophen Versus Codeine/Acetaminophen in the Treatment of Acute Extremity Pain After Emergency Department Discharge Andrew K. Chang, MD, MS, Polly E. Bijur, PhD, Kevin G. Munjal, MD, MPH, and E. John Gallagher, MD
Abstract Objectives: The objective was to test the hypothesis that hydrocodone/acetaminophen (Vicodin [5/500]) provides more efficacious analgesia than codeine/acetaminophen (Tylenol #3 [30/300]) in patients discharged from the emergency department (ED). Both are currently Drug Enforcement Administration (DEA) Schedule III narcotics. Methods: This was a prospective, randomized, double-blind, clinical trial of patients with acute extremity pain who were discharged home from the ED, comparing a 3-day supply of oral hydrocodone/ acetaminophen (5 mg/500 mg) to oral codeine/acetaminophen (30 mg/300 mg). Pain was measured on a valid and reproducible verbal numeric rating scale (NRS) ranging from 0 to 10, and patients were contacted by telephone approximately 24 hours after being discharged. The primary outcome was the between-group difference in improvement in pain at 2 hours following the most recent ingestion of the study drug, relative to the time of phone contact after ED discharge. Secondary outcomes compared side-effect profiles and patient satisfaction. Results: The median time from ED discharge to follow-up was 26 hours (interquartile range [IQR] = 24 to 39 hours). The mean NRS pain score before the most recent dose of pain medication after ED discharge was 7.6 NRS units for both groups. The mean decrease in pain scores 2 hours after pain medications were taken were 3.9 NRS units in the hydrocodone/acetaminophen group versus 3.5 NRS units in the codeine/acetaminophen group, for a difference of 0.4 NRS units (95% confidence interval [CI] = –0.3 to 1.2 NRS units). No differences were found in side effects or patient satisfaction. Conclusions: Both medications decreased NRS pain scores by approximately 50%. However, the oral hydrocodone/acetaminophen failed to provide clinically or statistically superior pain relief compared to oral codeine/acetaminophen when prescribed to patients discharged from the ED with acute extremity pain. Similarly, there were no clinically or statistically important differences in side-effect profiles or patient satisfaction. If the DEA reclassifies hydrocodone as a Schedule II narcotic, as recently recommended by its advisory board, our data suggest that the codeine/acetaminophen may be a clinically reasonable Schedule III substitute for hydrocodone/acetaminophen at ED discharge. These findings should be regarded as tentative and require independent validation in similar and other acute pain models. ACADEMIC EMERGENCY MEDICINE 2014; 21:228–235 © 2014 by the Society for Academic Emergency Medicine
From the Department of Emergency Medicine, Albert Einstein College of Medicine, Montefiore Medical Center (AKC, PEB, EJG), Bronx, NY; and the Department of Emergency Medicine, Mount Sinai Medical Center (KGM), New York, NY. Received July 11, 2013; revisions received September 3 and September 13, 2013; accepted September 16, 2013. The authors have no relevant financial information or potential conflicts of interest to disclose. Supervising Editor: Mark B. Mycyk, MD. Address for correspondence: Andrew Chang, MD, MS; e-mail: achang@montefiore.org. Reprints will not be available.
228
ISSN 1069-6563 PII ISSN 1069-6563583
© 2014 by the Society for Academic Emergency Medicine doi: 10.1111/acem.12331
ACADEMIC EMERGENCY MEDICINE • March 2014, Vol. 21, No. 3 • www.aemj.org
O
ver 44 million U.S. emergency department (ED) visits per year are for chief complaints related to pain. Approximately 29%, or 12.8 million, of these visits result in prescriptions for an opioid analgesic at discharge.1 Three commonly prescribed oral opioids at ED discharge include hydrocodone/acetaminophen (Vicodin, Norco), oxycodone/acetaminophen (Percocet), and codeine/acetaminophen (Tylenol #3). Of these, hydrocodone/acetaminophen is the most commonly prescribed while codeine/acetaminophen is the least prescribed.2 The popularity of hydrocodone/acetaminophen postdischarge is supported by two small clinical trials,3,4 and several other studies that compare the efficacy of commonly prescribed oral opioids in different settings.5–9 There has been increasing concern about the abuse of oral opioids in patients discharged from the ED.10,11 The efficacy of hydrocodone versus codeine has not been adequately studied among discharged adult ED patients with acute pain. Using the clinical model of acute extremity pain, we designed a randomized trial comparing these two commonly prescribed analgesics to determine if the strong preference for use of hydrocodone/acetaminophen over codeine/acetaminophen was supported by evidence. This question has become particularly relevant in light of the Federal Drug Administration (FDA)’s recent advisory committee recommendation to reclassify hydrocodone from a Schedule III to a Schedule II narcotic, while leaving codeine unchanged as Schedule III. The objective of this investigation was to compare the outpatient analgesic efficacy, side-effect profile, and patient satisfaction of hydrocodone/acetaminophen versus codeine/acetaminophen for treatment of acute extremity pain in nonelderly adult patients discharged from the ED. We hypothesized that hydrocodone/acetaminophen would provide more efficacious pain relief than codeine/acetaminophen (primary hypothesis), with a lower incidence of side effects and greater patient satisfaction (secondary hypotheses). METHODS Study Design This was a prospective, randomized, double-blind, clinical trial comparing efficacy, side effects, and patient satisfaction of hydrocodone/acetaminophen (5 mg/500 mg) to those of codeine/acetaminophen (30 mg/300 mg) in patients discharged from the ED. The study was approved by the Montefiore Medical Center Institutional Review Board and registered at www.clinicaltrials.gov (NCT01402375). Each participant gave written informed consent in either English or Spanish, depending on his or her preference. Study Setting and Population The study recruited patients from an academic urban ED with an annual census of >100,000 adults, many of whom were underserved minorities. Patients were 21 to 64 years of age presenting with acute extremity pain and for whom the provider planned to discharge with a short course of oral opioids for outpatient pain management. “Acute” was defined as less than 7 days’ duration.
229
“Extremity” was defined as distal to and including the shoulder or hip joint. Exclusion criteria were as follows: history of prior or current methadone use; chronic condition requiring frequent or continuous pain management (e.g., sickle cell, fibromyalgia, or chronic neuropathy); allergy to hydrocodone, codeine, or acetaminophen; history of prior or current recreational opioid use; ingestion of an oral opioid in the past 24 hours; medical conditions that might affect opioid or acetaminophen metabolism (e.g., hepatitis, renal insufficiency, hypothyroidism, hyperthyroidism, or adrenal disease); or use of medication that might interact with opioids or acetaminophen (e.g., selective serotonin or norepinephrine reuptake inhibitors, cyclic antidepressants, antipsychotics, antimalaria medications, amiodarone, dronedarone, diphenhydramine, celecoxib, ranitidine, cimetidine ritonavir, terbinafine, or St. John’s Wort). Patients were recruited and enrolled by trained research associates (RAs), fluent in English and Spanish, 24 hours per day, 7 days per week, from January 2012 through August 2012. Study Protocol Patients were randomized to receive a 3-day course of hydrocodone/acetaminophen (5 mg/500 mg) or codeine/ acetaminophen (30 mg/300 mg). These doses were selected to reflect the most commonly prescribed formulations of each of these combination oral opioids, commonly known as Vicodin and Tylenol #3, respectively. Allocation assignments were generated by an online random number generator (www.randomization. com) and placed in sealed opaque envelopes to be opened in sequential order by RAs. Patients, treating physicians, and RAs were all blinded to study allocation. The pharmacist, working in an area distant from and inaccessible to ED staff, ensured proper allocation concealment by inserting study medications into identical unmarked opaque gel capsules, filling any void with lactose to equalize weight. After examining the large size of a single capsule, we decided to divide the dosage in half to ensure patients’ ability to swallow smaller prepared capsules. Thus, each capsule contained hydrocodone/acetaminophen (2.5 mg/250 mg) or codeine/ acetaminophen (15 mg/150 mg), and one dose of study medication equaled two capsules. The clinician ordered and dispensed a 3-day supply of the blinded medications to the patient along with verbal and written instructions. Patients received a 3-day supply (36 tablets or 18 doses) and were instructed to take two tablets (one dose) every 4 hours as needed for pain (maximum daily dose = six doses = 12 capsules). Patients were also strongly advised not to take any other analgesic medications, including over-the-counter or topical medications, during the study period, which concluded after the follow-up phone call approximately 24 hours following ED discharge. A structured patient diary was provided at discharge to assist patients in recording pain numeric rating scale (NRS) levels and side effects during the study period. Although NRS pain scores were obtained during the ED visit and prior to discharge, management of the patient while in the ED was at the ED attending physician’s discretion and was independent of the study
230
Chang et al. • HYDROCODONE/ACETAMINOPHEN VS. CODEINE/ACETAMINOPHEN AFTER ED DISCHARGE
protocol. The study medications were given to patients at the time of ED discharge, with the patient to be contacted approximately 24 hours later. Measurements Primary outcome data were obtained via follow-up phone call targeted for 24 hours after ED discharge. The first attempt to reach the patient was made at 24 hours unless the call was due between 10 p.m. and 8 a.m., during which we reasoned that many patients and their families might regard a phone call as intrusive. Thus, calls scheduled during these hours were made after 8 a.m. the following morning. Any patient who could not be reached at 24 hours or after 8 a.m. the following day received a minimum of twice-daily phone calls between 8 a.m. and 10 p.m. for up to 72 hours postdischarge, at which point the patient was classified as lost to follow-up. At least two phone numbers per patient were collected by the RAs at the time of study entry, one of which was almost invariably a cell phone, whose number was confirmed via a phone call made by the RA prior to discharge, and recorded as a required field to be completed prior to patient enrollment on the data collection instrument. The primary outcome was the between-group difference in improvement in mean (with standard deviation [SD]) NRS pain scores, measured at 2 hours following the most recent ingestion of the study drug prior to the time of phone contact. If the patient was reached less than 2 hours after the most recent dose of pain medicine, the survey was temporarily halted. The patient was then called again once the 2-hour period had elapsed and the survey resumed. Pain was measured on a validated and reproducible verbal 11-point NRS,12–14 based on patient recall. Patients were not informed of their pain scores during their ED visits. Secondary outcomes included incidence of prespecified side effects and overall patient satisfaction assessed dichotomously on two axes: 1) very satisfied/satisfied versus very unsatisfied/unsatisfied and 2) a simple yes/ no response to the question “Would you want the same pain medication again if you came to the ED with a similar degree of pain in the future?” Data Analysis Data were collected on a standardized data collection instrument. A trained data clerk entered the data in SPSS (version 17, IBM SPSS, Inc., Armonk, NY) on an ongoing weekly basis. To minimize transcription error, double entry of the entire data set was performed independently by a second trained clerk. All differences were reconciled by referral to the original data collection instrument. In addition, 10% of the data collection instruments were randomly audited to confirm data quality. Descriptive statistics were calculated for all variables. We visually examined the data to determine the appropriateness of describing continuous data with means or medians and the decision to use parametric or nonparametric tests for statistical inference. Means of variables that were symmetrically distributed and proportions are presented with 95% confidence intervals (CIs). We described variables that were clearly not symmetrically
distributed with medians and interquartile ranges (IQRs). t-tests and chi-square or Fisher’s exact test were used to assess differences between means and dichotomous variables, respectively. The Wilcoxon rank-sum test was used to compare medians. SPSS was also used for all analyses. To test the primary hypothesis that hydrocodone/ acetaminophen would provide superior pain relief compared to codeine/acetaminophen, we used the previously validated minimum clinically significant difference of 1.3 NRS units as the smallest improvement in pain score clinically worth detecting.12–14 With a standard two-tailed alpha of 0.05, power of 80% to detect a difference of 1.3 NRS units or greater, and a SD of 3.0, we estimated that 170 patients who took at least one dose of the prescribed medication (85 in each group) would be required. Based on a prior study with similar design, we estimated that approximately 30% of patients would fail to take their medication as prescribed.6 Thus we estimated that as many as 120 patients in each group would be needed to achieve the sample size of at least 85 patients per group. RESULTS The CONSORT diagram summarizing the flow of patients through the trial is shown in Figure 1. The RAs evaluated 849 patients for eligibility. The most common reasons for exclusion were pain duration greater than 7 days and physician judgment that the patient did not need opioid analgesics. A total of 240 patients were randomized. Data from the first 35 patients were not included in the analyses because we made changes to the data collection instrument in response to the RAs’ initial experience with it. We treated these data as a pilot to estimate the proportion of patients who were expected to take the prescribed medication and the ability of patients to recall pain before and after taking the medication. Twenty-four patients (12 in each group) did not take the prescribed discharge medicine prior to the followup phone call and therefore could not provide a 2-hour postingestion change in NRS pain score. Thus, the final sample consisted of 88 patients who received hydrocodone/acetaminophen and 93 who received codeine/acetaminophen. No patients were lost to follow-up. As shown in Table 1, the two comparison groups had similar demographic characteristics, diagnoses, use of analgesics before the visit, time to follow-up, and NRS pain scores while in the ED. The median time from discharge to follow-up was 26 hours (IQR = 24 to 39 hours) and this did not differ significantly between the two groups. Table 2 shows the main outcomes. The mean NRS pain scores for both groups before the most recent dose of pain medication were 7.6 NRS units. This intensity of pain score was expected as a high level of pain would prompt the patient to take the prescribed study analgesic pain medicine. The mean change in NRS pain scores 2 hours after the most recent dose of study medication was 3.9 NRS units for the hydrocodone/acetaminophen arm versus 3.5 NRS units for the codeine/acetaminophen arm, for a difference of 0.4 NRS units (95% CI = –0.3 to 1.2 NRS units). Although both
ACADEMIC EMERGENCY MEDICINE • March 2014, Vol. 21, No. 3 • www.aemj.org
231
Assessed for eligibility (N=849)
Excluded (n=609) Pain present more than 7 days (n=150) Opioid analgesia not warranted (n=123) Nontraumatic pain (n=70) Kidney, liver, thyroid, or adrenal disease (n=66) Took pain medication within 24 hours (n=62) Concurrent medication warranted (n=61) Allergic to study medication (n=39) Refused consent (n=32) Admitted to hospital (n=6)
Randomized (n=240)
Codeine/acetaminophen (n=120)
Hydrocodone/acetaminophen (n=120)
Analyzed for primary outcome (n=93)
Analyzed for primary outcome (n=88)
Pilot data (n=15)
Pilot data (n=20) Did not take study medication (n=12)
Did not take study medication (n=12)
Figure 1. CONSORT diagram.
Table 1 Baseline characteristics Characteristic
Hydrocodone/Acetaminophen (n = 88)
Female sex, n (%) Race/ethnicity, n (%) Hispanic African American Other Age (yr), mean (SD) NRS pain intensity on ED arrival, median (IQR) NRS pain intensity on discharge, median (IQR) Time (hours) to follow-up, median (IQR) Diagnosis, n (%) Sprain/strain Extremity fracture Other Took analgesic prior to presentation, n (%) No Yes
Codeine/Acetaminophen (n = 93)
38 (43)
48 (52)
56 27 5 34 9 7 27
69 18 6 37 9 7 25
(64) (31) (6) (12) (8 to 10) (5 to 8) (24 to 42)
(74) (19) (7) (11) (8 to 10) (5 to 8) (24 to 36)
40 (46) 20 (23) 28 (32)
54 (58) 17 (18) 22 (24)
63 (72) 24 (28)
63 (70) 27 (30)
IQR = interquartile range; NRS = numeric rating scale.
Table 2 Outcomes by Group
Outcome NRS before most recent dose, mean (SD) NRS 2 hours after most recent dose, mean (SD) Change in NRS before–after most recent dose, mean (SD) Satisfied with analgesic, n/N* (%) Want same analgesic again, n/N* (%)
Hydrocodone/Acetaminophen (n = 88)
Codeine/Acetaminophen (n = 93)
7.6 (1.7) 3.6 (2.8) 3.9 (2.5)
7.6 (1.8) 4.1 (2.9) 3.5 (2.6)
0.0 (–0.5 to 0.5) –0.5 (–1.3 to 0.4) 0.4 (–0.3 to 1.2)
72/87 (83) 59/85 (69)
66/93 (71) 66/93 (71)
11.8% (–0.7 to 24) –1.5% (–15 to 12)
Denominators are different due to missing data from unanswered questions.
Difference (95% CI)
232
Chang et al. • HYDROCODONE/ACETAMINOPHEN VS. CODEINE/ACETAMINOPHEN AFTER ED DISCHARGE
groups of analgesic medicine reduced the NRS pain intensity score by approximately 50%, there was neither a clinically nor a statistically significant difference between the two study medicines when compared to an evidence-based, prespecified change in NRS of at least 1.3.12–14 Three patients in the hydrocodone/acetaminophen group and five patients in the codeine/acetaminophen group took additional pain medication beyond what was prescribed to them for this study. This occurred most commonly because the study medication did not adequately relieve the patient’s pain. There were no differences in satisfaction and percentage who would want the same analgesic in the future, although the study was not powered based on a satisfaction metric. Table 3 shows adverse events by group. There were no clinical or statistically important between-group differences in any adverse event category. DISCUSSION There has been significant recent controversy over the prescribing of oral opioids at ED discharge driven by assertions that they are overprescribed, diverted, and abused.10 As a result, the U.S. Drug Enforcement Administration (DEA) recently made a request that hydrocodone combination products be rescheduled.15 A 2-day meeting was held in January 2013 by the FDA’s Drug Safety and Risk Management Advisory Committee, in which members voted 19 to 10 that these products should be rescheduled from Schedule III to Schedule II under the Controlled Substances Act.16,17 Although this has not yet occurred, historically the FDA usually follows the recommendations of its advisory panels. Some states are already making these changes on their own. New York’s I-STOP legislation was signed in August 2012 and took effect in February 2013, which changed hydrocodone from Schedule III to Schedule II under the controlled substance schedules in Section 3306 of the New York State Public Health Law.18 In addition, many EDs have subsequently been asked to
follow voluntary ED discharge opioid prescribing guidelines.19 Because Schedule III medications may have less potential for abuse and lead to lower physical and psychological dependence compared to Schedule II drugs,20 Schedule III drugs can be prescribed by faxing a prescription or calling a pharmacist and are eligible for up to five refills within 6 months. In contrast, Schedule II drugs require handwritten prescriptions that cannot be phoned or faxed and may not be refilled. Thus, the reclassification of hydrocodone from Schedule III to Schedule II may have a substantial effect on outpatient oral opioid prescribing patterns simply by increasing legal impediments to the use of hydrocodone and its combinations, while leaving current restrictions governing prescription of codeine and codeine-containing combinations unaltered. Commonly prescribed oral opioids after ED discharge include several acetaminophen-opioid combinations: hydrocodone/acetaminophen (Vicodin and Norco), oxycodone/acetaminophen (Percocet), and codeine/acetaminophen (Tylenol with codeine). Hydrocodone/ acetaminophen and oxycodone/acetaminophen are generally thought to be stronger opioids than codeine/acetaminophen. This is why we selected a traditional superiority randomized clinical trial design, hypothesizing that the hydrocodone combination would provide meaningfully better pain relief than the codeine combination. However, our data indicate that these two agents appear to be clinically and statistically similar with respect to analgesic efficacy, side-effect profiles, and patient satisfaction. Hydrocodone/acetaminophen is the most common opioid combination prescribed on discharge from U.S. EDs and is used approximately three times more often than oxycodone/acetaminophen,2 which is a Schedule II drug. The popularity of hydrocodone over oxycodone may be related to the fact that hydrocodone is currently a Schedule III medication while oxycodone is a Schedule II medication, thus making the former substantially easier for physicians to prescribe. Codeine/acetaminophen, a Schedule III drug, is presently as easy to prescribe as
Table 3 Adverse Side Effects by Group Side Effect Nausea Vomiting Constipation Diarrhea Pruritus Rash Dizziness Drowsiness Confusion Number of side effects 0 1 2+
Hydrocodone/Acetaminophen 9/87 2/87 0/88 0/87 3/87 1/87 12/87 25/88 2/86
Codeine/Acetaminophen
(10) (2) (0) (0) (3) (1) (14) (28) (2)
51/86 (59) 24/86 (28) 11/86 (13)
Data are reported as n/N (%) Denominators are different due to missing data from unanswered questions.
9/92 1/92 0/93 5/93 3/93 0/92 5/93 27/93 1/93
Difference% (95% CI)
(10) (1) (0) (5) (3) (0) (5) (29) (1)
0.6 (–8 to 10) 1.2 (–4 to 6) 0 (–3 to 3) –5.4 (–11 to 0.1) 0.2 (–6 to 6) 1.1 (–3 to 5) 8.4 (–0.5 to 17) –0.6 (–14 to 13) 1.3 (–3 to 6)
52/91 (57) 30/91 (33) 9/91 (10)
2.2 (–12 to 17) –5.1 (–18 to 9) 2.9 (–7 to 12)
ACADEMIC EMERGENCY MEDICINE • March 2014, Vol. 21, No. 3 • www.aemj.org
hydrocodone/acetaminophen, but for unclear reasons, is not used as frequently. Although the Cochrane Collaboration has produced several reviews describing the analgesic effectiveness of codeine, both with and without acetaminophen,21,22 there are no Cochrane reviews or meta-analyses targeting the efficacy of hydrocodone. We could find only a single paper supporting the preferential prescribing of hydrocodone/acetaminophen over codeine/acetaminophen. Forbes et al.6 compared a single dose of hydrocodone/acetaminophen (7.5 mg/500 mg) to a single dose of codeine/acetaminophen (30 mg/300 mg) for postoperative oral surgery pain and found significantly better pain relief among the patients treated with hydrocodone, with no significant difference in side effects. Although opioid equivalencies are difficult to estimate, particularly with codeine because of the interindividual variation in metabolism of this opioid to its active metabolite, morphine, consensus data indicate that 5 mg of hydrocodone given orally is roughly the analgesic equivalent of 30 mg of oral codeine.23 If this estimate is correct, then the validity of Forbes’ single-dose study may have been compromised by performance bias due to administration of 50% more opioid analgesic-equivalent plus 200 mg more of acetaminophen in the hydrocodone arm than was administered to patients in to the codeine arm. Thus, because the analgesic “exposure” in the two groups may have been imbalanced—thus producing an inherently “unfair comparison”—one cannot confidently conclude superiority, noninferiority, nor equivalence due to a methodologic limitation. The only other publication we could find comparing hydrocodone and codeine was a well-conducted clinical trial from the University of Pittsburgh by Turturro et al.,4 which found no significant difference in the mean or median pain scores between the two groups at various intervals over the 48 hours following ED discharge for acute musculoskeletal pain. This was consistent with our 24-hour follow-up data. In contrast to the study by Forbes et al., Turturro et al. used hydrocodone/acetaminophen (5 mg/500 mg) versus codeine/ acetaminophen (30 mg/500 mg), which is consistent with an unbiased comparison of the two combination analgesics because each arm of the trial received analgesic-equivalent doses of both opioids and acetaminophen. We chose not to do this because the combination of codeine/acetaminophen (30 mg/500 mg) is not available in the United States. Two other randomized controlled trials were found comparing hydrocodone versus codeine for chronic pain.9,24 Rodriguez et al.,24 similar to Turturro et al.,4 found no significant difference in either analgesia or incidence of side effects in a study of 121 patients comparing equianalgesic doses of 5 mg of hydrocodone versus 30 mg of codeine, both in combination with 500 mg of acetaminophen. Palangio et al.8 in a three-armed (hydrocodone/ibuprofen [7.5 mg/200 mg]; hydrocodone/ibuprofen [15 mg/400 mg]; and codeine/ acetaminophen [60 mg/600 mg]) study design with 469 patients found greater analgesia in the hydrocodone/ibuprofen (15 mg/400 mg) group versus the codeine/acetaminophen (60 mg/600 mg) group. As in the only other paper suggesting analgesic superiority of
233
hydrocodone over codeine by Forbes et al.6 in the setting of acute pain, discussed above, validity is likely to be compromised by the insinuation of performance bias. In the study by Palangio et al., this bias appears in the form of allocation of nonequianalgesic doses of drugs to the groups under comparison (e.g., 50% higher dose of hydrocodone than would be considered equianalgesic to 60 mg codeine, further confounded by the presence of ibuprofen in one arm). LIMITATIONS Because 24 enrolled patients (12 in each group) did not take the study medication as directed, no data could be obtained on these patients during the 24-hour followup. Thus, we were not able to perform an intention-totreat analysis and instead performed a per-protocol analysis. The intention-to-treat design has the distinct advantage of diminishing alpha error, i.e., the probability of obtaining a false-positive finding, by driving results toward the null. In the case of a negative study result obtained via a per-protocol analysis, however, the issue of alpha error is moot. Furthermore, application of an intention-to-treat analysis can only make the findings “more negative” in a traditional superiority trial such as this one, which has failed to identify a meaningful difference between the comparison groups in the face of a sufficiently large sample size to provide traditionally adequate power of 80%. Thus, although use of a per-protocol analysis might be viewed by some as a limitation of any randomized controlled trial that employs this method, in the setting of a negative study such as this one, this would only serve to reinforce our findings. Both hydrocodone/acetaminophen and codeine/acetaminophen come in several different concentrations. Although Norco (5 mg hydrocodone/300 mg acetaminophen) is frequently prescribed in the United States, we chose to use what we believed to be the most commonly prescribed combinations, which were 5 mg hydrocodone/500 mg acetaminophen and 30 mg codeine/ 300 mg acetaminophen. Although improbable, we recognized that the additional 200 mg of acetaminophen might constitute a limitation in the form of performance bias favoring hydrocodone/acetaminophen (5 mg/ 500 mg). However, our findings indicate that, even if this design feature did contribute to the slight advantage associated with hydrocodone/acetaminophen (5 mg/500 mg), the delta that we identified in NRS score is only 0.4 (95% CI = –0.3 to 1.2). This is not only neither statistically nor clinically significant, but is also substantially closer to the numerical null of 0 than to the minimum clinically significant difference, or experimental null, of 1.3 NRS units. Furthermore, the fact that the upper limit of the 95% CI of 1.2 NRS units falls short of the minimum a priori criteria of 1.3 makes it highly unlikely that the “true” difference between the study arms is as large or larger than this threshold value required for declaring a clinically meaningful difference between the two drug combinations under comparison. Although the chosen doses of hydrocodone/acetaminophen and codeine/acetaminophen typically come in a single pill, we decided to split the dose into two capsules. We did this out of concern that the size of the
234
Chang et al. • HYDROCODONE/ACETAMINOPHEN VS. CODEINE/ACETAMINOPHEN AFTER ED DISCHARGE
smallest single capsule containing the full dose of medication that could be prepared by our research pharmacist was far too large for many patients to comfortably swallow and would therefore produce adherence bias (another form of performance bias), thus unbalancing randomly allocated exposure to the two analgesics under comparison. We therefore split the dose into two smaller capsules and instructed patients to take two pills at a time, just as they might take two acetaminophen, two aspirin, or two other nonsteroidals for pain. We recognized that this might also contribute to adherence bias, but we judged it to be a less likely impediment to ingestion than a single pill of intimidating size. Fortunately, the 24 patients who did not take the study medication exactly as prescribed following ED discharge were evenly distributed across the hydrocodone/acetaminophen and codeine/acetaminophen comparison groups and did not diminish our analyzable sample size below the number required to achieve a statistical power of 80% to detect the target delta of 1.3 NRS units,12–14 if a between-group difference in pain improvement of this magnitude were in fact present. Oral opioids are typically prescribed as a range of one or two tablets every 4 to 6 hours as needed. Given the large number of possible combinations such a range would entail, we decided to fix the dose at one tablet every 4 hours as needed, which is within the range of how the prescriptions are typically written. We chose to start at the lower range of dosing (one tablet) since this was our first study examining oral opioids and we wished to proceed in a stepwise manner, investigating potentially higher doses (e.g., two tablets) depending on the outcomes of the current study. Indeed, the chosen dosages resulted in an approximately 50% reduction in pain intensity for both groups of analgesics. And finally, we were cognizant of recent case reports regarding certain patients who hypermetabolize codeine, leading to toxic morphine levels and associated respiratory depression.25 The time of the pain scores varied depending on when the patient took the medication. There may also be recall bias because of the difference in time between when patients took the study medications and when they answered the follow-up questions, although it is likely that this would have affected both groups equally. Since only adults between the ages of 21 and 64 years of age were studied, our findings cannot be extrapolated to children, adolescents, or the elderly. The study was also carried out in a single center composed of a largely minority population of urban, poor, and medically underserved patients. Because there have been reports of oligoanalgesia associated with ethnicity and socioeconomic status, and because there are racial differences in drug metabolism, the results of this study may not generalize to other populations in the United States. CONCLUSIONS In a randomized, controlled, double-blind clinical trial, we found that hydrocodone/acetaminophen (Vicodin [5/500]) did not provide clinically or statistically superior outpatient pain relief compared to codeine/ acetaminophen (Tylenol #3 [30/300]), when assessed at
24 hours following discharge from the ED in patients presenting with acute extremity pain. Similarly, there were no clinically or statistically important differences in side-effect profiles or patient satisfaction. If the Drug Enforcement Administration reclassifies hydrocodone as a Schedule II narcotic, as recently recommended by its advisory board, our data suggest codeine/acetaminophen (30/300 mg) may be a clinically reasonable Schedule III substitute for the currently more frequently prescribed hydrocodone/acetaminophen (5/500 mg) at ED discharge. These findings should be viewed as preliminary and require independent validation in similar and other acute pain models. References 1. Pletcher MJ, Kertesz SG, Kohn MA, Gonzales R. Trends in opioid prescribing by race/ethnicity for patients seeking care in US emergency departments. JAMA 2008;299:70–8. 2. Niska R, Bhuiya F, Xu J. National Hospital Ambulatory Medical Care Survey: 2007 emergency department summary. Natl Health Stat Report 2010:1–31. 3. Marco CA, Plewa MC, Buderer N, Black C, Roberts A. Comparison of oxycodone and hydrocodone for the treatment of acute pain associated with fractures: a double-blind, randomized, controlled trial. Acad Emerg Med 2005;12:282–8. 4. Turturro MA, Paris PM, Yealy DM, Menegazzi JJ. Hydrocodone versus codeine in acute musculoskeletal pain. Ann Emerg Med 1991;20:1100–3. 5. Charney RL, Yan Y, Schootman M, Kennedy RM, Luhmann JD. Oxycodone versus codeine for triage pain in children with suspected forearm fracture: a randomized controlled trial. Pediatr Emerg Care 2008;24:595–600. 6. Forbes JA, Bates JA, Edquist IA, et al. Evaluation of two opioid-acetaminophen combinations and placebo in postoperative oral surgery pain. Pharmacotherapy 1994;14:139–46. 7. Litkowski LJ, Christensen SE, Adamson DN, Van Dyke T, Han SH, Newman KB. Analgesic efficacy and tolerability of oxycodone 5 mg/ibuprofen 400 mg compared with those of oxycodone 5 mg/acetaminophen 325 mg and hydrocodone 7.5 mg/acetaminophen 500 mg in patients with moderate to severe postoperative pain: a randomized, doubleblind, placebo-controlled, single-dose, parallel-group study in a dental pain model. Clin Ther 2005;27: 418–29. 8. Palangio M, Damask MJ, Morris E, et al. Combination hydrocodone and ibuprofen versus combination codeine and acetaminophen for the treatment of chronic pain. Clin Ther 2000;22:879–92. 9. Palangio M, Morris E, Doyle RT Jr, Dornseif BE, Valente TJ. Combination hydrocodone and ibuprofen versus combination oxycodone and acetaminophen in the treatment of moderate or severe acute low back pain. Clin Ther 2002;24:87–99. 10. Cantrill SV, Brown MD, Carlisle RJ, et al. Clinical policy: critical issues in the prescribing of opioids for adult patients in the emergency department. Ann Emerg Med 2012;60:499–525.
ACADEMIC EMERGENCY MEDICINE • March 2014, Vol. 21, No. 3 • www.aemj.org
11. Wilsey BL, Fishman SM, Ogden C. Prescription opioid abuse in the emergency department. J Law Med Ethics 2005;33:770–82. 12. Bijur PE, Latimer CT, Gallagher EJ. Validation of a verbally administered numerical rating scale of acute pain for use in the emergency department. Acad Emerg Med 2003;10:390–2. 13. Holdgate A, Asha S, Craig J, Thompson J. Comparison of a verbal numeric rating scale with the visual analogue scale for the measurement of acute pain. Emerg Med (Fremantle) 2003;15:441–6. 14. Todd KH. Clinical versus statistical significance in the assessment of pain relief. Ann Emerg Med 1996;27:439–41. 15. Kuehn BM. FDA committee: more restrictions needed on hydrocodone combination products. JAMA 2013;309:862. 16. Kartzinel R. Guidelines for scheduling drugs under the controlled substances act [proceedings]. Psychopharmacol Bull 1981;17:40–2. 17. McCarthy M. FDA panel recommends tighter rules for prescribing opioids. BMJ 2013;346:f1012. 18. State of New York Department of Health. Frequently asked questions Part C Chapter 447 Laws of 2012 (Controlled Substance Schedule Changes). Available at: http://www.health.ny.gov/professionals/narcotic/ laws_and_regulations/part_c-chapter_447-laws_of_ 2012-faq.htm. Accessed Dec 2, 2013.
235
19. Juurlink DN, Dhalla IA, Nelson LS. Improving opioid prescribing: the New York City recommendations. JAMA 2013;309:879–80. 20. Drug Enforcement Administration. Department of Justice. Dispensing controlled substances for the treatment of pain issuance of multiple prescriptions for schedule II controlled substances. J Pain Palliat Care Pharmacother 2007;21:75–88. 21. Derry S, Moore RA, McQuay HJ. Single dose oral codeine, as a single agent, for acute postoperative pain in adults. Cochrane Database Syst Rev 2010: CD008099. 22. Toms L, Derry S, Moore RA, McQuay HJ. Single dose oral paracetamol (acetaminophen) with codeine for postoperative pain in adults. Cochrane Database Syst Rev 2009:CD001547. 23. Drugs for pain. Med Letter (pay to access) 2004;2:47–54. 24. Rodriguez RF, Castillo JM, Del Pilar Castillo M, et al. Codeine/acetaminophen and hydrocodone/acetaminophen combination tablets for the management of chronic cancer pain in adults: a 23-day, prospective, double-blind, randomized, parallel-group study. Clin Ther 2007;29:581–7. 25. Friedrichsdorf SJ, Nugent AP, Strobl AQ. Codeineassociated pediatric deaths despite using recommended dosing guidelines: three case reports. J Opioid Manag 2013;9:151–5.
