Treatment of chronic noncancer pain in patients on opioid therapy in primary care: A retrospective cohort study

Abstract

Background:

This study sought to characterize the real-world treatment of chronic noncancer pain (CNCP) in patients on opioid therapy in primary care.

Methods:

A retrospective cohort study from 2014-18 was conducted at a multidisciplinary primary care clinic in Chilliwack, British Columbia. Included were adults on daily opioid therapy for CNCP. Patients receiving palliative care or ≤1 visit were excluded. Outcomes of interest included use of opioid/nonopioid pharmacotherapy, number/frequency of visits and proportion of patients able to reduce/discontinue opioid therapy.

Results:

Seventy patients (mean age 53 years, 53% male, 51% back pain) were included. Median follow-up was 6 visits over 12 months. Sixty-two patients (89%) reduced their opioid dose, 6 patients had no change and 2 patients required a dose increase. Mean opioid dose was reduced from 183 to 70 mg morphine equivalents daily. Twenty-four patients (34%) discontinued opioid therapy, 6 patients (9%) transitioned to opioid agonist therapy and 6 patients (9%) breached their opioid treatment agreement. Nonopioid pharmacotherapy included nonsteroidal anti-inflammatory drugs (64%), gabapentinoids (63%), tricyclic antidepressants (56%) and nabilone (51%).

Discussion:

Over half of patients were no longer on opioid therapy by the end of the study. Most patients had a disorder (e.g., back pain) for which opioids are generally not recommended. Overall mean opioid dose was reduced from baseline by approximately 60% over 1 year. Lack of access to specialized pain treatments may have accounted for high nonopioid pharmacotherapy usage.

Conclusions:

This study demonstrates that treatment of CNCP and opioid tapering can successfully be achieved in a primary care setting. Can Pharm J (Ott) 2020;153:xx-xx.

Knowledge Into Practice.

• There is currently a paucity of data regarding the practice-based treatment of patients with chronic noncancer pain on opioid therapy and success, or lack thereof, of tapering opioids in patients on high-dose opioid therapy.

• This retrospective cohort study described the treatment of 70 adult patients with chronic noncancer pain on daily opioid therapy in a multidisciplinary primary care clinic. Sixty-two patients were able to reduce their opioid dose, six patients had no change and two patients required a dose increase.

• Common nonopioid pharmacotherapy used as adjunctive/alternative therapy to opioids included nonsteroidal anti-inflammatory drugs, gabapentinoids, tricyclic antidepressants and nabilone.

• This study demonstrates that treatment of chronic noncancer pain and opioid tapering can be successfully achieved in a multidisciplinary primary care clinic.

Mise En Pratique Des Connaissances.

• À l’heure actuelle, il y a peu de données sur le traitement reposant sur la pratique visant les patients souffrant de douleur chronique non cancéreuse qui prennent des opioïdes ainsi que sur la réussite ou l’échec de la réduction de la dose des patients qui prennent des doses élevées d’opioïdes.

• La présente étude de cohorte rétrospective a décrit le traitement de 70 patients adultes souffrant de douleur chronique non cancéreuse qui prenaient quotidiennement des opioïdes, dans une clinique multidisciplinaire de soins primaires.

• La pharmacothérapie courante sans opioïdes utilisée comme thérapie d’appoint ou de rechange aux opioïdes comprenait des anti-inflammatoires non stéroïdiens, des gabapentines, des antidépresseurs imipraminiques et le nabilone.

• La présente étude démontre que le traitement de la douleur chronique non cancéreuse et la réduction des opioïdes peuvent être menés avec succès dans une clinique multidisciplinaire de soins primaires.

Introduction

North America is currently in the midst of an opioid crisis. The prevalence of opioid use disorder and rate of opioid-related overdose deaths have risen dramatically over the past 2 decades.¹ In the United States between 2000 and 2014, the age-adjusted rate of drug overdose deaths involving opioids more than doubled.² Furthermore, long-acting opioid therapy for the treatment of chronic noncancer pain, as compared to nonopioid pharmacotherapy, has been associated with an increase in death, including death from causes other than overdose.³ Long-term opioid therapy has also been associated with hyperalgesia, fractures, myocardial infarction and sexual dysfunction.^4,5 The opioid crisis has given rise to a variety of guidelines, standards and position statements encouraging safe prescribing of opioids.^6-11 The authors of a Canadian opioid therapy guideline, published in the Canadian Medical Association Journal in 2017, recommend that opioid therapy for chronic noncancer pain should only be considered after optimization of nonopioid pharmacotherapy and nonpharmacologic therapy.⁶ Additionally, they advocate that the opioid dose, if trialed, should be limited to less than 50 to 90 mg of morphine equivalents daily (MED) and that patients currently taking high-dose opioid therapy (defined as ≥90 mg MED) should undertake a taper to achieve the lowest effective dose. However, other than recommending a multidisciplinary approach, the authors do not provide guidance on how best to taper opioids.

There is currently a paucity of data regarding the practice-based treatment of patients on preexisting opioid therapy and success, or lack thereof, of tapering opioids in patients on high-dose therapy. The objective of this study was to characterize the real-world treatment of chronic noncancer pain in patients on preexisting opioid therapy in a primary care setting and evaluate how treatment of these patients compares to the recently published guideline.⁶

Methods

Study design

This was a retrospective cohort study at a primary care clinic in Chilliwack, British Columbia. The clinic provides longitudinal care for complex or unattached patients from the Chilliwack area. During the study period, 1 primary care physician with an interest in chronic pain treatment (although no formal training) accepted consultative referrals for patients with chronic pain. The goal of this service was to manage chronic pain rather than specifically to taper opioid therapy, although most referred patients had uncontrolled pain while on high-dose opioid therapy and thus had an indication for opioid tapering. Many patients were also referred to the clinical pharmacist at the clinic to provide consultative services. The pharmacist, who often followed up with patients between clinic visits with the physician, primarily focused on assessing the safety and effectiveness of nonopioid pharmacotherapy for pain and recommending treatment for opioid withdrawal symptoms, while the physician primarily focused on assessing the opioid therapy. Patients who received longitudinal care at the clinic were also under the care of a nurse practitioner or physician for non-pain-related medical conditions. All patients received education and patient resources on nonpharmacologic therapy for pain management (e.g., physiotherapy, cognitive behavioural therapy) and were referred for specialized pain treatments as required. All patients undergoing an opioid therapy taper received motivational interviewing. All patients were required to sign an opioid treatment agreement—patients who broke this agreement (e.g., for being prescribed opioids from another health care provider or using illicit or street-purchased opioids) were tapered off of their opioid therapy typically over a 2-week period and offered treatment for opioid use disorder. Patients were still provided care through the clinic, including treatment of their chronic pain, but were no longer prescribed opioid therapy. The Fraser Health Research Ethics Board approved this study.

Study population

All adult patients (≥19 years of age) with chronic noncancer pain who received care from the primary care pain physician were reviewed. Chronic noncancer pain was defined as a painful condition that persists for at least 3 months and is not associated with malignant disease.⁶ Included were patients on chronic daily opioid therapy—patients exclusively on intermittent or nonprescription opioid therapy (e.g., low-dose codeine [8 mg or less per dosage unit] products) were excluded. Opioid therapy was defined as codeine, fentanyl, hydromorphone, oxycodone, methadone (for pain) or morphine prescribed by a registered health care provider. Patients taking tramadol, buprenorphine, illicit opioids (e.g., diacetylmorphine) or opioid agonist therapy (OAT) for opioid use disorder were excluded. Patients with cancer, enrolled in a palliative care program or with ≤1 clinic visit (e.g., as requested by the referring health care provider or the patient did not attend the follow-up appointment) were also excluded. Patients were typically followed in this study until they were able to achieve adequate pain control and/or successfully tapered off of opioids.

Data sources and collection

Data were collected from the clinic electronic medical record by 1 investigator. Data were available from February 2014 to April 2018, and all data were collected in March and April 2018. Only clinic visits to the primary care pain physician were reviewed. The following data were collected in a deidentified manner: patient age and sex, etiology of chronic noncancer pain, presence/absence of a mental health disorder (defined as addiction, anxiety, bipolar disorder, depression, personality disorder, posttraumatic stress disorder, schizoaffective disorder or schizophrenia), opioid and nonopioid pharmacotherapy for pain at the initial clinic visit, number of clinic visits, overall duration of follow-up and opioid and nonopioid pharmacotherapy for pain at the final or most recent clinic visit. Nonopioid pain medications of interest included gabapentinoids, mirtazapine, nabilone, nonsteroidal anti-inflammatory drugs (NSAIDs), serotonin-norepinephrine reuptake inhibitors, selective serotonin reuptake inhibitors, tricyclic antidepressants and topiramate. Data were not collected on use of muscle relaxants (e.g., baclofen, cyclobenzaprine) or acetaminophen. Data were also collected on use of clonidine (for opioid withdrawal), illicit or medical cannabis (for pain) and specialized pain treatments (e.g., trigger point injections, epidural steroid injections). Of note, data were collected prior to the legalization of cannabis for nonmedical purposes in Canada.

Data analysis

Outcomes of interest included mean opioid dose at initial and final or most recent clinic visits, number and frequency of clinic visits, proportion of patients able to reduce or discontinue opioid therapy (including those on ≥90 mg MED) and use of nonopioid pharmacotherapy for pain. All opioid doses were converted to milligrams MED by multiplying by the following conversion factors: codeine by 0.15, hydromorphone by 5 and oxycodone by 1.5.⁶ Fentanyl was converted to milligrams MED by multiplying the dose in micrograms/hour by 2.4 and methadone (for pain) by 4 (for 1-20 mg/day), 8 (for 21-40 mg/day), 10 (41-60 mg/day) or 12 (61-80 mg/day).¹²

Statistical analysis

Results were reported using descriptive statistics. Continuous variables were reported as means with standard deviations for normally distributed data and medians with interquartile ranges for nonnormally distributed data. Categorical variables were reported as frequencies with percentages. All statistical analyses were performed with IBM SPSS Statistics (version 21; SPSS, Inc., an IBM Company, Chicago, IL).

Results

A flow diagram of patients is included as Figure 1. Ninety-six of 166 patients were excluded, the reasons for which are included in Table 1. Baseline characteristics of included patients are described in Table 2. Twenty-four patients (34%) reported their pain disorder was the result of a motor vehicle collision. Forty-eight of the 49 patients (98%) with a mental health disorder had depression. Data regarding follow-up and mean opioid dose at baseline and final or most recent clinic visit are included in Table 1. Six patients (9%) broke their opioid treatment agreement. Of the 26 patients who continued opioid therapy at a lower dose, 17 (65%) were taking ≤90 mg MED.

Figure 1.

Flow diagram of reviewed patients

Table 1.

Reasons for excluded patients (n = 96)

Reason	n	%
One clinic visit	46	47.9
Not taking opioids at baseline	18	18.8
Intermittent (i.e., as needed) opioid use	10	10.4
Illicit or street-purchased opioid use	8	8.3
Did not attend clinic visit or cancelled without rescheduling	6	6.3
Opioid agonist therapy for opioid use disorder	4	4.2
Tramadol use at baseline	2	2.1
Cancer	1	1.0
Unable to determine baseline opioid dose	1	1.0

Table 2.

Patient characteristics (n = 70)

Characteristic	Value
Age (years), mean ± SD	53.4 ± 12.0
Male sex, n (%)	37 (52.9)
Concurrent mental health diagnosis, n (%)	49 (70.0)
Etiology,* n (%)
Chronic back pain	37 (52.9)
Osteoarthritis	12 (17.1)
Headache disorder	12 (17.1)
Fibromyalgia	11 (15.7)
Rheumatic autoimmune disorder	4 (5.7)
Peripheral neuropathy	3 (4.3)
Chronic pancreatitis	2 (2.9)
Complex regional pain syndrome	2 (2.9)
Costochondritis	2 (2.9)
Multiple sclerosis	2 (2.9)
Other	11 (15.7)
Baseline opioid, n (%)
Morphine	18 (25.7)
Oxycodone ± acetaminophen	17 (24.3)
Codeine/acetaminophen	12 (17.1)
Hydromorphone	9 (12.9)
Fentanyl	7 (10.0)
Methadone (for pain)	2 (2.9)
Fentanyl and oxycodone	2 (2.9)
Hydromorphone and codeine/  acetaminophen	1 (1.4)
Morphine and codeine/acetaminophen	1 (1.4)
Oxycodone/acetaminophen and codeine/  acetaminophen	1 (1.4)
Follow-up
Duration of follow-up (months),   median ± IQR	12 ± 16
Number of follow-up clinic visits,   median ± IQR	6 ± 7
Opioid dose (milligrams MED), mean ± SD
Baseline opioid dose	183.4 ± 212.6
Opioid dose at final or most recent clinic visit	69.5 ± 141.0

IQR, interquartile range; MED, morphine equivalents daily; SD, standard deviation.

^*Does not add to 100% as patients may have had more than 1 pain disorder.

Thirty-eight of 70 patients (54%) were on ≥90 mg MED at baseline. Of those, 33 patients (87%) had an opioid dose reduction, 4 patients (11%) had no change in opioid dose and 1 patient (3%) had an increase in opioid dose. Seven of 38 patients (18%) were able to discontinue opioid therapy, 3 patients (8%) broke their opioid treatment agreement and 2 patients (5%) were converted to OAT.

Nonopioid pharmacotherapy was recommended in 69 of 70 patients (99%) (Table 3). Twenty-three patients (33%) were recommended clonidine for opioid withdrawal symptoms, of whom 18 (78%) perceived a benefit. Twenty-five patients (36%) and 12 patients (17%) were referred for trigger point injections and epidural steroid injections, respectively. Twenty-one patients (30%) were referred to a tertiary pain clinic. Eighteen patients (26%) reported using cannabis (primarily nonmedical) for pain control.

Table 3.

Use of nonopioid pharmacotherapy

	Trialed, N (%)	Provided benefit, n (%)	Did not provide benefit or patient was intolerant, n (%)
NSAIDs	45 (64.3)	25 (55.6)	20 (44.4)
Gabapentinoids	44 (62.9)	23 (52.3)	21 (47.7)
TCAs	39 (55.7)	24 (61.5)	15 (38.5)
Nabilone	36 (51.4)	28 (77.8)	8 (22.2)
SNRIs	32 (45.7)	18 (56.3)	14 (43.8)
SSRIs	22 (31.4)	15 (68.2)	7 (31.8)
Mirtazapine	18 (25.7)	13 (72.2)	5 (27.8)
Topiramate	7 (10.0)	4 (57.1)	3 (42.9)

NSAIDs, nonsteroidal anti-inflammatory drugs; SNRIs, serotonin-norepinephrine reuptake inhibitors; SSRIs, selective serotonin reuptake inhibitors; TCAs, tricyclic antidepressants.

Discussion

This study sought to characterize the real-world treatment of chronic noncancer pain in patients on opioid therapy in a multidisciplinary primary care setting. The majority of patients had a pain disorder for which opioids are generally not recommended, such as back pain, headache disorder and/or fibromyalgia.⁸ Furthermore, 70% of patients had a concurrent mental health disorder, most commonly depression, although it is not known whether the disorder was preexisting before opioid therapy was initiated and the relative control was not assessed. Eighty-nine percent of patients (62/70) were able to taper their opioid therapy and 34% (24/70) were able to discontinue therapy. Seventeen percent of patients (12/70) transitioned to OAT for opioid use disorder or broke their opioid treatment agreement—when accounting for these patients, the proportion of patients no longer on opioid therapy at the end of the study period was 51% (36/70). The mean opioid dose was reduced from baseline by approximately 60% over 12 months (i.e., 10% every 2 months). For patients on high-dose opioid therapy (≥90 mg MED) at baseline, a similar number of patients (33/38, 87%) were able to taper their opioid therapy, but fewer (7/38, 18%) discontinued therapy. All but 1 patient was recommended some type of nonopioid pharmacotherapy, most commonly NSAIDs, gabapentinoids and tricyclic antidepressants. The nonopioid pharmacotherapy with the highest proportion of patients perceiving a benefit was nabilone at 78%, while the lowest proportion was gabapentinoids at 52%. A relative lack of access to local specialized pain treatments (e.g., trigger point injections, epidural steroid injections) may have negatively affected referral to these services and accounted for the high use of nonopioid pharmacotherapy. Furthermore, the nearest tertiary pain clinic is approximately 80 km away, which may explain why only 30% of patients were referred.

There are limited data regarding the practice-based management of patients on high-dose opioids and success or lack thereof of tapering opioid therapy. Harden and colleagues¹³ performed a retrospective review of 50 patients actively being tapered off of chronic opioid therapy. This review included a very similar patient population to the present study—mean age was 54 years, most patients were male and back pain was the most common diagnosis. In the Harden et al.¹³ review, 94% of patients tapered their opioid with a mean dose reduction of 46% over 12 months and 13% of patients discontinued therapy. In the present study, 89% of patients tapered their opioid with a mean dose reduction of 62% over 12 months and 34% of patients discontinued therapy. However, the present study had broader inclusion, as it accounted for patients with aberrant behaviour and patients did not have to agree to an opioid taper. A retrospective study by Baron and McDonald¹⁴ identified 23 patients undergoing “detoxification” from high-dose opioid therapy. All patients were switched from their opioid to ibuprofen and a subgroup was treated with buprenorphine. Twenty-one of these patients (91%) described a “significant” decrease in self-reported pain after detoxification.

The success with opioid tapering observed in this study was likely positively influenced by the multidisciplinary approach. The role of the pharmacist in the management of these patients focused on assessing nonopioid pharmacotherapy for pain (e.g., suggesting dose titration, monitoring adverse effects) and opioid withdrawal management (e.g., clonidine, adjunctive therapy for gastrointestinal symptoms). As well, the pharmacist provided education to patients regarding the long-term adverse effects of opioid therapy and inappropriateness of opioid therapy for certain conditions (e.g., back pain). Patients were able to consult with the pharmacist between clinic visits with the physician, thereby increasing their access to a health care professional. Pharmacists can play a key role in the management of patients with chronic noncancer pain through additional interventions, such as performing a thorough medication history that identifies previously trialed therapies for pain and possible interactive medications, drafting an opioid tapering schedule (if applicable) and providing patient education. Pharmacists are often more accessible than primary care physicians, thereby allowing for more opportunities for patients to engage with a member of the health care team.

This study highlights factors that may contribute to the success of managing chronic noncancer pain in real-world practice. It is imperative to develop a therapeutic plan with patients that includes clear expectations of both parties. Clinicians must be able to build trust and rapport with these patients, as some may not be willing to engage in the process, particularly if it involves opioid tapering. Clinicians also need to be patient but assertive and have flexibility when negotiating changes to the therapeutic plan as necessary. It is crucial to have a reimbursement model that allows for additional time to spend with patients and frequent appointments. All members of the multidisciplinary team should document progress with the therapeutic plan to ensure consistent messaging—this is likely better achieved in a practice setting where the team is co-located in the same clinic. Last, all members of the multidisciplinary team must be aware of the criteria for opioid use disorder to identify possible signs.

This study captured contemporary data on treatment of chronic noncancer pain with opioids over a 4-year period. However, the results must be interpreted within the context of the study limitations. As this was a retrospective review, it relied on the accuracy and completeness of the electronic health record. Only patients at a single primary care clinic were included; however, many patients were referred by other health care providers in the broader geographical area and thus represent a more diverse and challenging population with respect to pain management. Tramadol was excluded, as at the time of the study, it was not scheduled under the Canadian Controlled Drugs and Substances Act and Narcotic Control Regulations. Furthermore, it was challenging to account for self-prescribing of low-dose codeine products (other than patient self-report), as these sales are not tracked in British Columbia. Illicit drug use may have been underrepresented, as routine urine drug testing was instituted at the clinic in March 2017. After the final clinic visit, it is not known whether patients restarted or escalated their opioid dose from another health care provider and/or started using illicit or street-purchased opioids. No data about the objective assessment of pain (e.g., brief pain inventory) were collected. However, as stated, the goal of the service was to manage patients’ chronic pain to ensure adequate analgesic therapy and not to taper opioids despite exacerbating pain. Finally, a multidisciplinary team approach that includes a primary care physician with an interest in chronic pain treatment may not be feasible in all practice settings.

Conclusion

This observational cohort study of the primary care–based treatment of patients with chronic noncancer pain on opioid therapy demonstrated that almost 90% were able to successfully taper their opioid with an approximate 60% mean dose reduction over 12 months. Over one-third of patients were able to discontinue opioid therapy, while few patients transitioned to OAT or breached their opioid treatment agreement. This study demonstrates that treatment of chronic noncancer pain and opioid tapering can successfully be achieved in a multidisciplinary primary care clinic. ■