An observational study on risk factors for prolonged opioid prescription after severe trauma

1.1 Trauma, pain and psychiatric co-morbidity

Trauma is one of the most common causes of morbidity and mortality in people of working age [1]. The prevalence of pain is high among trauma patients in emergency care [2], [3]. Following surgery, approximately 10% of patients develop persistent postsurgical pain [4]. Chronic pain is a complex phenomenon that can adversely affect quality of life and is associated with sleep disorders and other psychiatric conditions such as anxiety and depression [5]. Existing studies vary in reported prevalence rates of psychiatric co-morbidity within the trauma patient population, both before and after trauma. Posttraumatic stress disorder (PTSD) and major depressive disorder are the most frequently reported diagnoses in longitudinal studies aimed at identifying the prevalence of psychiatric co-morbidity following severe injury [6], [7], [8], [9], [10]. Depression and PTSD have been associated with greater symptom severity and lower levels of functioning [11], [12]. Specific phobia [7], anxiety disorder [9], [10] and alcohol abuse [12] are other psychiatric diagnoses recurring in studies following trauma. Although many studies have focused on psychiatric co-morbidity after trauma, psychiatric diagnosis and substance abuse have also been shown to be independent risk factors for suffering a trauma [13].

1.2 Opioid treatment and effects

Pharmacological treatment using analgesics (including opioids) is an important part of the treatment and management of acute pain following trauma. However, for chronic pain, pharmacological treatment is often insufficient to fully alleviate pain and improve functional capacity [5], [14]. Much attention has been given to understanding the side effects of opioid treatment and the appropriateness of opioid treatment in chronic non-cancer pain has been increasingly examined. Existing studies show a high prevalence of serious side effects including drug dependency and death [15], as well as associated financial costs [16]. The association between opioid analgesic prescription and increased risks for adverse outcomes has been shown in epidemiological studies, mainly from the US. One of those studies indicated that the majority of deaths (60%) among people using opioids occur following prescription by health care providers [15]. Clinically, such prescription may occur when acute pain due to tissue damage is treated, where a continuing opioid treatment was not part of the plan at that time.

Observational studies have also shown that opioid treatment for chronic pain is associated with increased health risks, such as overdose, opioid abuse, fractures, sexual dysfunction and myocardial infarction [17]. Iatrogenic opioid dependence may also be a risk factor for less successful long-term occupational work. Opioid dependent patients have been shown to be less likely to return to work and more likely to engage in healthcare utilization from new providers, even after adjusting for relevant demographic factors and psychiatric co-morbidity [18]. Risk factors associated with higher risks of prolonged opioid prescription include lower household income, specific co-morbidities and use of specific drugs preoperatively, such as benzodiazepines [19], as well as drug and alcohol abuse and depression [20].

The present study was based on clinical observations indicating that a problematic opioid use can be initiated during trauma care, suggesting a need to evaluate opioid therapy and its effects on trauma patients. Specific attention is given to patients with known psychiatric conditions which may render them more vulnerable to problematic opioid use.

1.3 Aims

The aim of this observational study was to broadly characterize patients referred to a pain specialist after severe trauma, with a focus on trauma type, psychiatric co-morbidity, and opioid prescription pattern. This was done to tentatively investigate possible risk factors for long-term opioid use following trauma.

2.1 Recruitment of study participants

The care of severe trauma cases in the Stockholm region, Sweden, is the responsibility of Karolinska Trauma Center (KTC) at the Karolinska University Hospital (KUH). The mean number of registered trauma patients per year at KUH in Solna, Stockholm, were 1,406 during the years 2013–2015 [21]. All individuals aged 16 years or older referred to the Karolinska Pain Center (KPC), from April 1, 2013 to May 3, 2015 whilst inpatients following trauma were asked to participate. The patients were referred to the KPC for treatment consultation due to complex pain. All patients (29) who were informed about the study and asked to participate provided informed consent. The 29 trauma patients represent all consecutive patients referred to the KPC during the period.

2.2 Clinical intervention

The pain specialists were employed at the KPC, and provided consultations when physicians at the KTC at KUH referred patients to them. They gave consultations to the physicians in the KTC and met the patient for a follow up visit, but had no further contact with the patient or the treating physician. The consultation consisted of a pain analysis and advice regarding pharmacological and non-pharmacological pain treatment, including advice on the appropriate duration of opioid treatment. Specific advice against continuing opioid treatment for chronic pain was also given. When risk factors associated with adverse outcome of long-lasting opioid therapy were identified, both the physician responsible for the patient on the ward and the patient were explicitly advised that opioids should not be continued beyond the advised time-frame for patients with similar injuries. When signs of inappropriate opioid use indicating the development of opioid dependence or problematic opioid use were identified, such as preference for intravenous administration, requests for opioids to reduce anxiety or rapid dose escalation without any new tissue injury, this was addressed by providing advice on how this should be managed.

2.3 Data collection

The study utilized individual data on the study sample from an existing national database on trauma care, SweTrau [22]. Additional data were collected from the electronic medical records (EMR) TakeCare^® on one occasion (July 2016) from the Stockholm Region as the basis for characterizing the participants. Data for the first year after the trauma occasion was retrieved retrospectively. Thus, although the patients were enrolled consecutively, data collection was retrospective.

3.1 Patient characteristics

The majority of the 29 patients were male (76%), with age ranging between 21 and 55 (median=34) years.

3.2 Trauma and injuries

Types of trauma and injuries are shown in Table 1. Regarding type of trauma, in 11 patients the cause of trauma was classified as ‘high energy fall”, and six of these were considered by the treating physician to be self-inflicted. Most patients (27) underwent surgery during their hospital stay. The median number of surgeries was two, ranging from 0 to 11. The most common type of surgery was orthopedic, which 21 patients underwent. Twenty patients were treated in the ICU, for a median of two days and ranging between 0 and 35 days. The median length of hospital stay was 23 days (range 3–113 days). One patient died by suicide during the follow-up period.

3.3 Opioid prescription patterns

In 48% (n=14) of the patients there were records of opioid prescription at least once during the six months preceding the trauma. In five patients (17%) the indication was something else than pain, and in nine (31%) the opioids were categorized as pain treatment. In four patients the prescription pattern clearly matches the indication of opiate substitution treatment following opioid abuse. This was in all cases secondary to the use of illegal opioids such as heroin, and not secondary to opioid use as part of pain treatment. In one patient the EMR stated ongoing heroin abuse, verified by the patient. For one patient a prescription of tramadol matching daily use was identified. Thus, in total six individuals were identified as having a continuous opioid exposure up to the time of the trauma.

In 25 patients (86%) there was an ongoing prescription of opioids corresponding to at least daily use three months after the trauma. At 6 months this was reduced to 21 patients (72%). One year after the trauma, a total of 18 patients (62%) still had an ongoing prescription for opioids in amounts corresponding to daily use or more, and two more patients were classified as intermittent users. Oxycodone was the most commonly prescribed long-term opioid. Individual data on the duration of continuous opioid prescription following trauma is shown in Fig. 1.

Fig. 1:

Opioid prescription prior to and following trauma presented as individual data (n=29). Dashed lines represent prescription corresponding to intermittent (less than daily) use during the 6 months before trauma. Continuous lines represent prescription corresponding to at least daily use. Two lines are blank because two participants had no opioid use before trauma and no continuous opioid use after trauma.

3.4 Psychiatric pharmacological treatment

Pharmacological treatment for psychiatric diagnoses were prescribed in 59% (17) of the patients during the 6 months period before trauma, most commonly Anti-Anxiety and Antidepressive agents as well as Hypnotics and Sedatives. During the follow-up period, medication was prescribed for psychiatric diagnoses in 93% (27) of patients, including benzodiazepines (seven patients before trauma, and another six at follow-up).

3.5 Psychiatric co-morbidity

Twenty patients (69%) in the present sample had a psychiatric diagnosis recorded with an ICD code in the EMR before the trauma. In the sample, 13 individuals had a diagnosed affective disorder (depression and/or anxiety), 13 had an addiction disorder, six had a diagnosed neuropsychiatric disorder (ADHD or autism), and one had a diagnosed psychosis. Several of the 29 patients had more than one diagnosis. Psychiatric diagnoses were reported in the EMR in five more patients but without an ICD code, typically because the medical records were outside the Stockholm Region.

4.1 Main finding

The participants displayed high injury severity, had extensive surgery as well as extended hospital stay. Notably, participants in the present study had higher injury severity (ISS 18, IQR 16-34) in comparison to the total trauma population at KUH (ISS 5, IQR 2-14) [13]. However, for most participants opioids were still prescribed 1 year after trauma. The majority were also prescribed psychiatric medication and had an identified psychiatric co-morbidity in their EMR before trauma. Thus, although tentative, this observational study indicate that long-term opioid prescription may be relatively common among patients with complex trauma and psychiatric co-morbidity.

4.2 Limitations of the study

This is a small observational study conducted to preliminarily investigate possible risk factors for continued problematic opioid use in a selected sample of patients referred to the pain consultant. However, this paper provides support for a larger longitudinal evaluation. This study utilized retrospective data, which implies a possible underestimation of the amount of prescription and psychiatric diagnoses. We only have ICD codes for the psychiatric diagnoses from EMR, in other words not from a standardized clinical interview. We have used the EMR from the Stockholm Region, which implies that we do not know if patients from outside the Stockholm Region received additional health care or diagnoses. The prescription data were retrieved from EMR, which includes all prescription data from 90% of primary care and five out of the six large hospitals in the Stockholm Region. It is not possible to know if all prescriptions were filled at a pharmacy or if these drugs were actually used by the participants. Also, the lack of adequate data on pain intensity and pain related disability prevents investigation of the impact of pain, prescription patterns, and the treatment effect.