Effectiveness of Oxycodone Hydrochloride (Strong Opioid) vs Combination Acetaminophen and Codeine (Mild Opioid) for Subacute Pain After Fractures Managed Surgically

This randomized clinical trial compares the effectiveness of a strong opioid (oxycodone hydrochloride) vs a mild opioid (acetaminophen codeine combination) when taken postdischarge for subacute pain among patients who underwent orthopedic surgical treatment for fracture.

each individual's particular physical, psychological, historical, social and cultural 23 experiences and circumstances. Therefore, pain is often difficult to manage, and 24 adequate pain relief is a major concern and area of focus in Australia today [1]. 25 26 Postoperative pain is a major challenge in patients following orthopaedic surgery [2]. 27 Serious risks for adverse events are evident with both opioid and nonsteroidal anti-28 inflammatory drugs (NSAIDs). Even though opioids are considered as the primary 29 analgesic therapy in moderate to severe postoperative pain, they are associated with 30 dose-related adverse effects such as sedation, respiratory depression, postoperative 31 nausea and vomiting, pruritus, constipation and urinary retention [3,4]. Furthermore 32 the potential for misuse, abuse or addiction is also of concern [5, 6]. NSAIDs are 33 associated with adverse effects such as gastrointestinal injury, increased operative 34 site bleeding, renal toxicity, and bronchoconstriction [7,8]. In addition, NSAIDs have 35 been shown to interfere with fracture healing, bone-tendon healing, and spinal fusion 36 [9, 10]. Paracetamol with its high safety profile in recommended dosage, lack of 37 allergic potential and absence of contraindications in peptic ulcer diseases, 38 haemostatic disorders, or pulmonary dysfunction has gained popularity as a 39 complementary analgesic [11,12]. Postoperative intravenous paracetamol has been 40 found to reduce postoperative opioid consumption after orthopaedic surgery [2], but 41 at present there is insufficient data to decide whether paracetamol reduces opioid- importance and usefulness. In terms of post discharge analgesics there is a lack of 49 evidence to inform practice. There are no randomised control trials directly 50 comparing commonly prescribed analgesics and yet strong opioid prescription is 51 common practice. Surprisingly, a recent randomised controlled study involving 52 children with non-surgically managed fractures, showed that morphine (a strong 53 opioid, taken orally) or ibuprofen improved pain scores with no significant difference 54 in analgesic efficacy between the two approaches [14]. Morphine was associated 55 with a greater frequency of adverse effects. If the same applies in adult populations, 56 it is possible that we are overestimating the benefits of strong opioids and exposing 57 patients to higher risk of adverse events. The current study will fill this gap in 58 evidence and has the ability to change current practice or provide the evidence 59 required to validate current practice. If a patient is considered eligible, a member of the research team will gain informed 95 consent. The research team will provide education to the participant on how to take 96 the study treatment, including providing the participant with a sealed medication 97 Page 3 of 20 pack. A researcher (blinded) will collect baseline data from the participant prior to 98 discharge. Following baseline data collection and prior to discharge from hospital the 99 researcher will inform the participant to break the seal on the medication pack and 100 commence the study medicine once discharged. At this point the participant is 101 considered to be randomised into the study.

103
Randomisation and blinding 104 A researcher not involved in participant recruitment or data collection will generate a 105 randomisation schedule a priori using a computer-derived random number 106 sequence. Study medication will be over-encapsulated and prepared according to Investigator office in the Research Institute until transfer to pharmacy occurs. All 111 study medication will be accounted for and recorded on study pharmacy logs. The 112 randomisation schedule will be kept concealed from other researchers and all 113 participants. Following recruitment and upon discharge home from hospital, a 114 hospital doctor will complete a study script which will be provided to pharmacy. prior to discharge, thus randomising the participant to one of two groups: oxycodone 119 hydrochloride or paracetamol with codeine. The capsules for each group will be 120 identical in appearance. The randomisation process will ensure concealed allocation 121 and blinding of the researchers, participants and outcome assessors.

123
Study treatment 124 The dose of the study medication for the active group, Oxycodone hydrochloride 125 immediate release, will be 5mg or 10mg four times daily or 1-2 tablets, providing a 126 maximum dose of 40 mg daily or 8 tablets. The dose of the study medication for the 127 control group, Paracetamol and Codeine phosphate, will be 500mg/8mg or 128 1000mg/14mg four times daily or 1-2 tablets, providing a maximum dose of 4000mg of 129 Paracetamol and 64mg of codeine daily or 8 tablets.

131
In the standard study dosing regimen, the maximum tolerated dose for each 132 participant will be maintained for 2 weeks before the study medication is titrated 133 down to cessation in the final (3rd) treatment week. If adequate pain relief is 134 achieved, defined as a pain rating of 0, 1 or 2 out of 10, for a minimum of 48 hours, 135 before the 3-week standard treatment regimen is completed, early titration down to 136 cessation is possible. The maximum treatment period is 3 weeks.

138
In addition to the study medicines, both groups may receive usual care during the 3- Data collection will be conducted by the study researchers in person at baseline and 147 via telephone at day 3 (3 days post discharge), and weeks 1, 2 and 3 post discharge.

148
Data will be collected regardless of participants' compliance to the study treatment.

149
Data collection by trained study researchers will increase data collection quality and 150 aid participant retention.

152
Primary outcome 153 The primary outcome is average pain during week 1 of treatment measured using 154 the Numerical Pain Rating Scale (NRS). The participants will be asked to rate their 155 overall pain daily over the last 24 hours out of 10, with zero representing 'no pain', 156 and 10 representing the 'worst pain imaginable'.

158
Secondary outcomes 159 The key secondary outcome is the EuroQol EQ-5D-5L measured at day 3 post 160 discharge and weeks 1, 2 and 3 post discharge to assess quality of life. In addition, 161 question 1 (mobility) and question 3 (usual activities) will be used to measure 162 function. Pain will be measured daily using the Numerical Pain Rating Scale. The 163 participants will be asked to rate their overall pain and worst pain over the last 24 164 hours out of 10, with zero representing 'no pain', and 10 representing the 'worst pain 165 imaginable'. Medication adverse events will be measured daily using a list of 166 common side effects experienced where participants answer yes or no. The Global 167 Perceived Effect will be measured at weeks 1, 2, and 3, this asks the participant to 168 compare their pain to when their injury first occurred. It is measured on a Likert 169 scale, from a score of −5 (vastly worse) to 0 (unchanged), to +5 (completely 170 recovered). Work and health utilisation questions will be asked at weeks 1, 2 and 3.

171
These will report the use of health services and the return to paid employment 172 following their injury.

174
Other data collected 175 Initial data (baseline) will be obtained directly from the patient after consent to the 176 trial. This data will include age, gender, employment status, diagnoses, type of 177 surgery, mechanism of injury, insurance status, pain and quality of life.

179
Details of adverse events will be collected at day 3 and weeks 1, 2 and 3 post-180 discharge. Any serious adverse events, defined as an event that is life threatening, 181 results in death, hospitalisation, or significant disability, will be reported immediately 182 to the Principle Investigator. The PI will investigate the relationship between a 183 serious adverse event and the study medication, and if a potential relationship is 184 suspected, unblinding to treatment allocation is permissible. The ethics committee 185 will also be informed of the serious adverse event.

187
Questions will be asked regarding blinding and satisfaction at week 3. The blinding 188 question will ask the participant to guess the treatment group to which they were 189 randomised. The satisfaction question will ask the participant to rate how satisfied  Data integrity and analysis 201 The integrity of trial data will be monitored by regularly scrutinising data files for 202 omissions and errors. Verification of source data will be performed by 100% data 203 verification on all primary and secondary outcomes. The source of any inconsistencies 204 will be explored and resolved. Electronic data will be stored on a secure server and 205 paper copies located in a locked cabinet. Data will only be accessible by researchers 206 and participant confidentiality will be maintained through secure data storage, during 207 and post-trial.

209
The data will be analysed according to the intention-to-treat principle. Standard 210 statistical techniques will be used to describe characteristics of patients in both 211 groups. We will compare baseline characteristics in the two treatment groups and if 212 incomparability appears, we will adjust for differences in secondary analysis. The 213 primary outcome, average of pain over week 1, will be compared between groups 214 using independent t-test. If adjustment for possible baseline incomparability is 215 needed, analysis of covariance will be done. We will use the SAS statistical software 216 package for data analysis. We consider p values less than 0.05 to be significant.

218
Sample size 219 The primary comparison is the difference between the two groups in average pain       Methods: One hundred and twenty (120) participants will be randomly assigned to one of two 343 treatment arms: mild opioids (8mg codeine with 30mg paracetamol) or strong opioids (5mg short 344 acting oxycodone) at discharge from hospital. Participants will be evaluated over a 3-week program or 345 until recovery. Participants will be asked to fill in a daily medication and pain diary and will be 346 Although opioids are considered as the primary analgesic therapy in moderate to severe postoperative 403 pain, they are associated with dose-related adverse effects such as sedation, respiratory depression, 404 postoperative nausea and vomiting, pruritus, constipation and urinary retention. 12,13 Furthermore the 405 potential for misuse, abuse or addiction is also of concern. 14, 15 NSAIDs are commonly used post-406 operatively but are associated with adverse effects such as gastrointestinal injury, increased operative 407 site bleeding, renal toxicity, and bronchoconstriction. 16,17 In addition, NSAIDs have been shown to 408 interfere with fracture healing, bone-tendon healing, and spinal fusion. 18-20 Paracetamol with its high choice' of opioid (mild or strong) or 'best' formulation (immediateor controlled-release). 427 Furthermore, evidence-based protocols, limits on dose, maximum prescription duration and patient 428 education are deficient for this population. In considering that 13% of people receiving a strong 429 opioid took the opioid for six or more weeks, strong opioid prescription following discharge from 430 hospital should be considered carefully. 24 In addition, while adverse events and the risk of mortality 431 rise proportionally with the opioid dose, analgesic and functional benefits may not 29,30 . The mortality 432 risk rises without increased benefit from a daily opioid dose of more than 100mg of oral morphine or 433 the equivalent. 31 We propose that discharge from hospital signifies the most suitable time point to 434 transition people down the WHO analgesic ladder from step 3 (strong opioid) to step 2 (mild opioid) 435 analgesics. We hypothesise that a strong opioid will provide better analgesia than a mild opioid-436 paracetamol combination when given post-discharge in orthopaedic surgical patients. 437 In addition to the study medicines, both groups will receive usual care during the 3-week treatment 496 period. Usual care includes physical therapies. Other analgesic medications are prohibited during the 497 treatment period except NSAIDs approved by study doctors. If participants require more pain relief 498 the participant will be discontinued from the study and referred to their general practitioner or 499 orthopaedic surgeon for review of pain management. 500 501

Data collection 502
Data collection will be conducted by the study researchers in person at baseline and via telephone at 503 day 3 (3 days post discharge), and weeks 1, 2 and 3 post discharge. Baseline data will be collected on 504 the hospital ward prior to discharge. Data will be collected regardless of participants' compliance to 505 the study treatment. Data collection by trained study researchers will increase data collection quality 506 and aid participant retention. 507 508

Primary outcome 509
The primary outcome is average daily pain during week 1 of treatment measured using the Numerical 510 Pain Rating Scale (NRS). The participants will be asked to rate their average pain daily over the last 511 24 hours out of 10, with zero representing 'no pain', and 10 representing the 'worst pain imaginable'. 512 513

Secondary outcomes 514
The key secondary outcome is the EuroQol EQ-5D-5L measured at day 3 post discharge and weeks 515 1, 2 and 3 post discharge to assess quality of life. In addition, question 1 (mobility) and question 3 516 (usual activities) will be used to measure function. Worst pain will be measured daily using the 517 Numerical Pain Rating Scale. The participants will be asked to rate their worst pain over the last 24 518 hours out of 10, with zero representing 'no pain', and 10 representing the 'worst pain imaginable'. 519 Note that this outcome (worst pain over the last 24 hours) should be distinguished from the primary 520 outcome of average pain over the last 24 hours. Medication adverse events will be measured daily 521 (and recorded in the pain diary) using a list of common side effects experienced where participants 522 Page 15 of 20 answer yes or no. The Global Perceived Effect will be measured at weeks 1, 2, and 3, this asks the 523 participant to compare their pain to when their injury first occurred. It is measured on a Likert scale, 524 from a score of −5 (vastly worse) to 0 (unchanged), to +5 (completely recovered). Return to work 525 (Yes [full or light duties] or No) will be asked at weeks 1, 2 and 3. These will report the return to 526 paid employment rate, capacity and time point following their injury. 527 528 Other data collected 529 Descriptive data including age, gender, employment status, diagnoses, type of surgery, mechanism of 530 injury, insurance status, pain and quality of life, will also be collected on entry to study. 531 532 Details of adverse events will be collected at day 3 and weeks 1, 2 and 3 post-discharge. Any serious 533 adverse events, defined as an event that is life threatening, results in death, hospitalisation, or 534 significant disability, will be reported immediately to the Principal Investigator (PI). The PI will 535 investigate the relationship between a serious adverse event and the study medication, and if a 536 potential relationship is suspected, unblinding to treatment allocation is permissible. The ethics 537 committee will also be informed of the serious adverse event. 538 539 Questions will be asked regarding blinding and satisfaction at week 3. The blinding question will ask 540 the participant to guess the treatment group to which they were randomised. The satisfaction question 541 will ask the participant to rate how satisfied they felt with the study pain management on a 5-point 542 scale (very dissatisfied, dissatisfied, neutral, satisfied or very satisfied). 543 544 Adherence to study medication will be ascertained via two methods: through participant 545 documentation in a daily medication diary; and by counting the returned medicine, compared to the 546 prescribed regimen. The latter is possible as participants will be asked to return unused medicines via 547 a reply-paid post satchel at the end of the 3-week treatment period.

Data integrity and analysis 551
The integrity of trial data will be monitored by regularly scrutinising data files for omissions and 552 errors. Verification of source data will be performed by 100% data verification on all primary and 553 secondary outcomes. The source of any inconsistencies will be explored and resolved. Electronic data 554 will be stored on a secure server and paper copies located in a locked cabinet. Data will only be 555 accessible by researchers and participant confidentiality will be maintained through secure data 556 storage, during and post-trial. 557

558
The distribution of characteristics of patients included in the study will be presented in a table 559 comparing baseline characteristics between the treatment arms. Significance testing to compare 560 baseline covariates between treatment arms will not be utilised. Data will be analysed by a statistician 561 who is blinded to group status. All analyses will be by intention-to-treat and missing data will not be 562 substituted unless there is more than 5% missing data for any variable in which case we will use 563 Multiple Imputation. The primary analysis will be a multilevel model comparing daily pain scores 564 between treatment groups whilst accounting for repeated measures by patient. Secondary analyses 565 will include a pre-protocol analysis of the primary outcome and unadjusted analysis of the secondary 566 outcomes using multilevel models. For primary outcome, a p value of <0.05 will be considered 567 statistically significant. 568 569

Sample size 570
The primary comparison is the difference between the two groups in average pain over the first 1 571 week, measured on an 11-point NRS. A sample of 46 participants per group (322 individual NRS 572 pain scores) will provide greater than 90% power to show at least a 1.0-unit (on a 10-point pain scale) 573 advantage of oxycodone over paracetamol plus codeine assuming an overall SD of 3 for the daily 574 NRS score, an ICC of 0.25 (ie. we expect 25% of the variation in NRS score to be between patients) 575 and a two-sided type 1 error rate of 5%. A sample of 60 participants per group (120 total) will be 576 collected to allow for a loss to follow-up rate of 20%. 577

DISCUSSION 579
Despite the demand for effective postoperative pain control, little evidence exists for post-discharge 580 pain management. There is no evidence for 'best choice' of opioid (mild or strong) or 'best' 581 formulation (immediate-or controlled-release). Furthermore, evidence-based protocols, limits on 582 dose, maximum prescription duration and patient education are deficient for this population. The 583 results of the RCT will provide information about the superiority of strong opioids over mild opioids 584 and evaluate the suitability of dose reduction at discharge after acute care. 585

AVAILABILITY OF DATA 590
The datasets used and/or analysed during the current study are available from the corresponding 591 author on reasonable request. 592

FUNDING 593
This trial has received competitive grant funding from the New South Wales Government. Australia, 594 State Insurance Regulatory Authority (formally known as the NSW Motor Accident Authority) Acute The funder has no role in the design of the study and collection, analysis, and interpretation of data 597 and in writing of the manuscripts. 598 599 AUTHORS' CONTRIBUTIONS 600 DJ, JN and IH were major contributors to trial and protocol design, grant writing and funding. DJ, JN 601