No signs of dose escalations of potent opioids prescribed after tibial shaft fractures: a study of Swedish National Registries

Previous studies on consumption of opioids in patients with non-cancer pain were either non population-based [1, 2], limited to a specific population, such as workers with low back injuries [3‐5], or had only a short follow-up [6]. Moreover, studies dealing with concerns of abuse, side effects, and efficacy of long-term opioid therapy in these conditions have not been conclusive [7‐9].

There is a lack of studies on the pattern of opioid use after skeletal fractures. Most of the reports in the literature are concerned with chronic back pain or other non-cancer pain conditions, but not with skeletal trauma patients [2, 10, 11]. The design and results of these studies illustrate the need for more selected patient groups with specific end-point data such as skeletal injuries.

Fractures of the tibial shaft are among the most common of serious skeletal injuries [12]. They are slow to heal and frequently cause permanent sequelae [13]. We analysed the long-term pattern of opioid consumption in patients with tibial shaft fractures. We aimed to study if potential risk factors such as age, sex, type of treatment, and mechanisms of injury would predict a prolonged opioid therapy. Moreover, we wanted to assess the potential risk of dose escalations in prescribed opioids in these patients.

Sweden has a unique personal identification number for all residents, which allows linkage between healthcare and other information from different registers for research. Data on all patients with tibial shaft fractures were obtained from the Swedish National Hospital Discharge Register (SNHDR). The Register records diagnoses and designated treatment codes according to the International Classification of Diseases (ICD), covering at least 98% of all hospital admissions in Sweden. A matched control group without tibial fractures was extracted from the Total Population Register. Each patient in the fracture group was matched with five individuals by age, sex, and residential area. None in the control group had been admitted to a hospital for a tibial fracture during the study period. Data on death or emigration for both groups were retrieved by Statistics Sweden from the Total Population Register.

Since July 1, 2005, all prescriptions filled at pharmacies in Sweden are stored in the National Pharmacy Register [14]. This does not include over-the-counter sales, which include some analgesics such as paracetamol and some of the non-steroidal anti-inflammatory drugs. However, opioid analgesics can only be obtained in pharmacies with prescription and are thereby included in the Register.

We identified all admissions in the SNHDR with ICD diagnostic codes for tibial shaft fractures (S822, S8220, and S8221). Relevant surgical intervention codes were analysed accordingly (NGJ29-NGJ99). Mechanisms of injury were studied using ICD E-codes (external codes) and grouped into 6 categories: fall on the same level, fall from height, unspecified fall, transport accident, miscellaneous, and unreported cause. The study period was July 1, 2005 to December 31, 2008.

All opioid analgesics prescribed to the patients in the study and control group were extracted from the National Pharmacy Register. These data include the following: name of the drug, date of filling the prescription, drug strength, number of pills, and dosage. The morphine equivalent dose (MED) for each opioid prescription in milligrams (mg) was calculated by multiplying the number of pills prescribed by the drug strength. These doses were then converted to MED using available equianalgesic conversions [15]. The median MED per day was calculated for each month. The MED was categorized as beeing low (< 20 mg), moderate (20–180 mg), or high (>180 mg) [16, 17]. We analysed potent opioids (oxycodone, morphine, and fentanyl), whereas less potent opioids (dextropropoxyphene, codeine, and tramadol) were not included [18]. We did not want to be biased by patients with associated fractures, therefore we exluded all patients with other fractures than tibial shaft fractures. Moreover, we excluded patients who had potent opioids before the index hospitalisation, as we wanted to study new opioid use after fracture. The study was approved by the regional ethical review board located at the Karolinska Institutet (2009/837-31/3 and 2010/0125-32).

There has been a continuous increase of opioid use for pain treatment among patients with non-cancer pain conditions during the past decade [19]. We studied the long-term opioid prescriptions after tibial shaft fractures in a national Swedish study. 25% of the patients filled a prescription for opioid analgesics at some point after the fracture. However, the doses prescribed were rather low and we did not see any evidence of a major dose escalations over time.

We excluded all patients with potent opioid prescriptions prior to the fracture, as we wanted to study the occurrence of new opioid prescriptions. Moreover, we excluded patients with other fracture diagnoses as we wanted study a rather homogenous fracture cohort. We are aware that the included patients may have obtained opioids during follow-up due to other reasons than the skeletal injury such as back pain, extremity pain, and abdominal pain. Therefore, an age- and sex-matched control cohort without fracture was included for comparison. In a cross-sectional survey from 2010 based on a nationwide register in Denmark, a high overall prevalence of opioid consumption (4.5%) was found in the general population. The relevance of which was, however, unknown [20]. These findings are in accordance with our findings of an opioid use of 3% in the control cohort without tibial fracture.

We did not detect any indication of major dose escalation in our cohort during the follow-up period. The median daily MED was between 7 and 21 mg during month 1 and 12 after fracture. Furthermore, as shown in Figure 4, the median MED for patients taking opioids was predominantly moderate to low in the beginning. During follow-up, the frequency of patients on moderate and high doses reduced. This is consistent with other data concerning non-trauma related pain conditions. In a meta-analysis of efficacy and safety of long-term opioid therapy for chronic non-cancer pain, many patients discontinued the therapy and very few patients showed signs of opioid addiction or abuse [21].

Our finding of a higher risk for continued use of opioids in younger patients (< 50 years) may be explained by a more extensive injuries often sustained during transport accidents in comparison with falls on the same level which is more often seen in elderly people. Furthermore, this finding for the older patient group may be reassuring, recently published reports raised increasing concerns on the safety of opioid analgesics in elderly people [22‐24].

The shortcomings of our study include the following: opioids prescribed to patients is not always equivalent to requirement or consumption of opioids. The incidence figures in this study may present and over- or underestimation of actual opioid use. An overestimation of the use of potent opioid analgesics is due to the fact that not all prescribed drugs are consumed. Thus the actual number of consumed doses is probably lower than 100 percent. In contrast, we did not include less potent opioids such as codein which converts to morphine in the liver, resulting in an underestimation. Moreover, we did not analyse other analgesics such as COX-inhibitors which may augment the analgesic effect of opioids reducing the quantity of the consumed opioid required. A further limitation of the study is: we do not know anything about the efficacy of the analgesic treatment. Lack of analgesic effect and/or side effects of opioids are major reasons why opioid therapy is stopped [25]. Patients may also, after some time, be prescribed less potent opioids by their general practitioners, who may be reluctant to provide potent opioids for non-cancer pain. This is a register study, therefore we do not know the specific reason why the patients discontinued the use of opioid medication. For example, one reason for discontinuing opioid treatment for elderly patients could be due to a higher incidence of adverse events. We also do not know if the excluded patients, who were already taking opioids prior to their tibial fracture, had an increase in their prescribed opioids following the fracture.

Our study is based on well validated government controlled national registries, including all hospitalized patients and opioid prescriptions in Sweden. We only studied the use of strong opioids in order to get a more homogenous patient group and to guarantee that the patients’ consumption is registered in order to obtain accurate statistics regarding drug escalation problems.

We did not see any signs from registry-data of major dose escalations over time among patients who received prescriptions for potent opioid analgesics after tibial shaft fractures.