OUP user menu

The Opioid Renewal Clinic: A Primary Care, Managed Approach to Opioid Therapy in Chronic Pain Patients at Risk for Substance Abuse

Nancy L. Wiedemer RN, MSN, CRNP, Paul S. Harden Pharm.D, Isabelle O. Arndt MD, PhD, Rollin M. Gallagher MD, MPH
DOI: http://dx.doi.org/10.1111/j.1526-4637.2006.00254.x 573-584 First published online: 1 October 2007


Objective. To measure the impact of a structured opioid renewal program for chronic pain run by a nurse practitioner (NP) and clinical pharmacist in a primary care setting.

Patients and Setting. Patients with chronic noncancer pain managed with opioid therapy in a primary care clinic staffed by 19 providers serving 50,000 patients at an urban academic Veterans hospital.

Design. Naturalistic prospective outcome study.

Intervention. Based on published opioid prescribing guidelines and focus groups with primary care providers (PCPs), a structured program, the Opioid Renewal Clinic (ORC), was designed to support PCPs managing patients with chronic noncancer pain requiring opioids. After training in the use of opioid treatment agreements (OTAs) and random urine drug testing (UDT), PCPs worked with a pharmacist-run prescription management clinic supported by an onsite pain NP who was backed by a multi-specialty Pain Team. After 2 years, the program was evaluated for its impact on PCP practice and satisfaction, patient adherence, and pharmacy cost.

Results. A total of 335 patients were referred to the ORC. Of the 171 (51%) with documented aberrant behaviors, 77 (45%) adhered to the OTA and resolved their aberrant behaviors, 65 (38%) self-discharged, 22 (13%) were referred for addiction treatment, and seven (4%) with consistently negative UDT were weaned from opioids. The 164 (49%) who were referred for complexity including history of substance abuse or need for opioid rotation or titration, with no documented aberrant drug-related behaviors, continued to adhere to the OTA. Use of UDT and OTAs by PCPs increased. Significant pharmacy cost savings were demonstrated.

Conclusion. An NP/clinical pharmacist-run clinic, supported by a multi-specialty team, can successfully support a primary care practice in managing opioids in complex chronic pain patients.

  • Opioids
  • Primary Care
  • Chronic Noncancer Pain
  • Substance Abuse
  • Pharmacy Costs
  • Pain Medicine


Chronic noncancer pain is a common problem facing primary care providers (PCPs) and the health care system [1]. Epidemiologic studies estimate that between one-third and one-half of adults live with some form of daily or recurrent pain [2]. Fifty percent of veterans seen in primary care report at least one type of chronic pain [3,4]. Timely, aggressive treatment of pain can prevent the adverse effects that often result from untreated or mismanaged pain, such as delays in healing, changes in the central nervous system (sensitization, neuronal plasticity, cortical reorganization, spontaneous pain), chronic stress and its medical consequences, family stress, depression, job loss, and suicide [5–10].

Comprehensive chronic pain management based on the biopsychosocial model of pain generation and perception improves outcomes [11–15]. Treatment methods include rational polypharmacy, physical therapy, psychotherapy, family therapy, interventional pain management, and complementary modalities. Opioids are usually considered after appropriate nonpharmacologic and pharmacologic modalities fail to relieve pain or when they are deemed safer than alternatives [5,16,17]. If monitored appropriately, opioids have safety advantages over non-steroidal anti-inflammatory medications, which are associated with known morbidity and mortality, for certain patients requiring daily medication [18–20]. However, despite a growing body of literature supporting the use of opioids in the treatment of chronic pain [21–25], and consensus statements available to guide practice [17,26–28], controversy over their use continues.

A major factor contributing to the controversy is the abuse potential of opioids. Clinicians are faced with a dilemma—how to safely incorporate opioids into treatment plans that maximize the possibility of successful pain control while minimizing the risk of misuse or abuse. This dilemma is heightened for PCPs. PCPs shoulder most of the burden of pain management [3,29,30] despite having received little specific training in pain medicine or addictionology [31,32] and being generally constrained to brief visits for evaluating and managing complex problems. The specter of state (medical practice boards) and federal (Drug Enforcement Agency) sanctions adds to their discomfort, even fear, of using opioids in clinical practice [33,34].

Another problem is the nomenclature used in clinical assessment and diagnosis. Confusion over terms used to define substance abuse disorders results in misconceptions about rates of addiction in the chronic pain population and in mislabeling, and stigmatizing patients. For example, patients taking chronic opioids for pain may be incorrectly diagnosed with addiction disorder when, in fact, they actually are just physically dependent on opioids. This confusion may bias clinicians against using or continuing to use opioids in patients with chronic pain [35,36]. The suggestion to a patient that their medical use of opioids for pain relief is drug addiction creates cognitive dissonance and distrust of the clinician, impairing the clinician–patient relationship. Of equal concern is the practice of providing opioids for chronic pain without following standard practice guidelines regarding assessment and monitoring [22]. This results in potentially mistreating chronic pain as well as missing the opportunity to identify and treat a comorbid substance abuse disorder. Either course, avoiding opioids when indicated for pain or providing opioids without following clearly documented clinical guidelines, risks a negative outcome.

Like for other common chronic diseases (e.g., hypertension, diabetes, depression, substance abuse), specialists in pain medicine cannot possibly assume care for all patients with chronic pain, numbering in the tens of millions. The majority of care must fall to primary care practitioners [11,12,37]. Although traditional specialties, such as orthopedics, rheumatology, and neurology, are often consulted in cases of chronic pain, they generally have neither the expertise nor interest in providing the longitudinal, biopsychosocial care, in a chronic disease management model, that is required for large numbers of patients with chronic pain. Thus, the common practice pattern is a succession of referrals to specialists and subspecialists with an interest in a narrow area of medical practice, the “sequential care model” of care [11] that does not address the totality of the biopsychosocial clinical challenge.

Since early, successful treatment impacts longitudinal outcomes [38], the challenge facing health care systems is how to incorporate evidence-based and consensus-based guidelines for treatment of pain disorders, including the use of opioids, into primary care practice where persistent pain usually first presents [39]. This challenge is formidable when considering the present nature of primary care practice in managed care, in which a typical visit lasts 10 minutes or less.

Thus, a new model of care—the Pain Medicine and Primary Care Community Rehabilitation Model—is proposed [11,12]. In this model, primary care practitioners use evidence-based algorithms supported by pain medicine specialty programs and community resources to care for chronic pain. Care is focused on the notion of secondary prevention, i.e., appropriate acute pain treatment and early recognition and effective early management of chronic pain disorders will reduce both the incidence and morbidity of chronic pain in the community. Recognizing enormous resource variability in communities and health care systems, the model emphasizes an approach based upon several principles: 1) primary care treatment supported by evidence-based algorithms; 2) timely access to pain medicine consultation and care, appropriate to priority of level of need, to prevent or reduce morbidity and improve functional outcomes; 3) goal-oriented, selectively multimodal, integrated outcomes-driven care; and 4) efficient use of available community resources to positively affect outcomes for the largest number in any health system or community population.

This article describes how a nurse practitioner (NP) and clinical pharmacist, supported by a multidisciplinary team of consultants, planned and implemented a structured approach to prescribing opioids in a large primary care practice at an urban, academic Veterans Affairs Medical Center, the Opioid Renewal Clinic (ORC) [40]. The article also describes the impact of that program on the use of opioids in the clinical practice of PCPs and their satisfaction as well as the impact of that intervention on patient outcomes and resource utilization in the medical center.

The development of the ORC coincided with a number of events nationally that brought pain management into the forefront of health care and in a negative turn, to the front pages of newspapers with the publicity about the misuse of Oxycontin® (Purdue Pharma LP, Stamford, CT) (referred to as oxycodone SA [sustained action]). The Veterans Health Administration (VHA) in 1998 and Joint Commission on Accreditation of Healthcare Organizations (JCAHO) in 2000 mandated through required standards that pain management is a right of all patients. The Philadelphia VA Medical Center (PVAMC), like all institutions across the country, adopted policies based on the current standards and guidelines [41,42]. The PVAMC's success in screening for pain in all veterans seeking care at our institution resulted in our identifying challenges in assessing and treating chronic pain in complex patients. The concurrent liberalization of the use of opioids for chronic noncancer pain and the trend toward using long-acting formulations also influenced our practice [16,17,21,23–25]. In 2000, the PVAMC was identified as one of the highest prescribers of oxycodone SA in the Veterans Administration (VA) system. The same year, the VHA removed oxycodone SA from the national formulary and mandated that all VHA medical centers to decrease oxycodone SA to 3% of all opioids prescribed. Oxycontin represented 22.5% in the first quarter in fiscal year 2001 at a cost of $129,793 at the PVAMC.

This mandate created a problem for the Primary Care Service. PCPs were screening for pain as mandated by JCAHO, but most PCPs were inexperienced in prescribing opioids for chronic pain. Following trends in the literature [16,21] and advice from experts who were invited to update the PCPs, prescribing long-acting opioids made sense. As many of our patients were already prescribed daily short-acting oxycodone/acetaminophen, the switch to oxycodone SA seemed easy. However, managing patients on chronic opioids already presented problems for PCPs and now they were mandated to convert their patients to another opioid.

Opioid-prescribing guidelines promoted by professional organizations [26,27] and federation of state medical boards [28] recommend consideration of the use of opioid treatment agreements/contracts which include the use of urine drug testing (UDT). Although efficacy of these tools remains unproven, their inclusion as policy in clinical settings is widespread. When used routinely for all patients, these tools can minimize conflicts associated with treatment with chronic opioids [43–46]. However, Fishman et al. [47] emphasize the limitations of relying solely on treatment agreements and UDTs to monitor adherence to chronic opioids.

Using precise terminology for substance misuse and addiction disorder reduces confusion and mislabeling of patients [48]. A range of aberrant drug-taking behaviors in patients treated with opioids for chronic pain has been described. Differential diagnosis includes under-treated pain (pseudoaddiction) [49], recreational drug use, undiagnosed psychiatric disorder (i.e., anxiety disorders, unipolar and bipolar depression, schizophrenia, personality disorder), encephalopathy, dementia, addiction, and diversion [50,51]. In the context of opioid treatment for pain, the disease of addiction is characterized by a persistent pattern of aberrant opioid use over time, including one or more of the following: loss of control over use of opioids, continued use despite harm, compulsive use, and craving [27]. There is no validated tool to predict the risk of substance abuse in patients with chronic pain being considered for opioid therapy [47]. Patients' self-report of drug use is frequently unreliable in this setting [44]. Observation and documentation of aberrant drug-taking behaviors while managing pain with a range of strategies that acknowledge the potential for abuse, misuse, and addiction allows for effective pain management and the identification of patients at risk for substance misuse and addiction [35].



The PVAMC is an urban university affiliated tertiary care center which provides health care for approximately 50,000 veterans (FY 2003). The PVAMC Primary Care Center has an enrollment of 17,000. Primary care is staffed by 19 physicians, 15 NPs, and one physician assistant. Guidelines and policies for management of chronic pain with opioids were made available to these PCPs in conjunction with education in the form of grand rounds and in-service seminars in 2000. An audit after 18 months revealed that clinical pra-ctice guidelines were rarely followed, indicating the need for a new plan for improving pain management.


The mandate to decrease oxycodone SA to 3% of all prescribed opioids presented an onerous task to the Primary Care Service. To assist, the pharmacy donated a full-time clinical pharmacist, which presented an opportunity to improve practice. This led to the formation of the ORC, which developed in three phases. In Phase I, one of the authors (NW) reviewed the literature, interviewed leaders in the field locally and nationally. She also held focus groups with the PCPs to obtain their input in designing a program. Several themes emerged that influenced the design of the ORC. PCPs believed that available pain management guidelines were impractical in primary care for several reasons: the complexity of veterans with pain, who tend to have multiple medical and psychiatric comorbidities; the brevity of outpatient encounters; inexperience in using opioids for chronic pain; and, the added time burden of utilizing opioid treatment agreements (OTAs), which included monitoring of chronic opioids with random UDT. They reported that conflicts with patients about opioid use were common, often resulting in angry interactions and interfering with the PCPs ability to focus on patients' medical comorbidities. PCPs asked specifically for help with what they viewed as “the problem patients.” The literature and the opinions of leaders in pain medicine reinforced the importance of strategies such as OTAs [47], UDT [44,45], a formal clinic policy [43], and the use of precise terminology for addiction disorder, physical dependence and tolerance to minimize confusion and labeling [31,35,48].

During Phase II, we designed and developed the ORC based upon the information provided in Phase I and the needs of our particular case mix of patients, which included a high percentage of patients with pain and psychiatric and addiction comorbidity. The ORC aimed to: 1) provide appropriate treatment for each patient—opioid therapy when indicated and addiction therapy when indicated; 2) improve PCP confidence in prescribing opioids; 3) improve monitoring and documentation; and 4) reduce overall costs of care both by decreasing misuse or overuse of resources (i.e., reducing emergency room [ER] and walk-in visits and complaints to the patient advocate requiring administrator and clinician time) and by meeting pharmacy budget goals for decreasing oxycodone SA use.

The program was managed by an NP and clinical pharmacist. PCPs referred patients by ordering a consult on the VA's Computerized Patient Record System. Consultation required a signed OTA with their PCP and UDT performed prior to enrollment. For convenience and speed, the following tools were made readily available in the electronic record for ease of use by PCPs: 1) an assessment template to document the key domains of pain treatment outcomes, called the “4As”—analgesia, activities of daily living, adverse events, aberrant drug-taking behaviors (Appendix I) [52]; 2) an electronic note for the OTA; and 3) a UDT order set placed in the primary care order screen.

A multidisciplinary pain management team (addiction psychiatrist, rheumatologist, orthopedist, neurologist, and physiatrist) met biweekly to support the NP and Pharm.D by reviewing cases and advising on treatment plans based on multimodal management including, besides opioids, NSAIDs and acetaminophen for osteoarthritis, transcutaneous electrical stimulation (TENS) units, antidepressants and anticonvulsants for neuropathic pain, and reconditioning exercises. They were also available for phone consultation. Early on, many cases were reviewed by the team, but as the PCPs became more comfortable with the program and the NP and Pharm.D acquired experience, most cases were managed by the Pharm.D and PCPs. The NP and Pharm.D were located in the primary care clinic (PCC) so that PCPs could drop in as needed to discuss cases face to face or by telephone as well as consulting the electronic medical record.

After referral, all patients were given the same instructions regarding the ORC structured program: they were expected to follow their individualized multimodality pain treatment plan (e.g., physical therapy, chronic pain school, tests and visits requested by PCP or Pain Team), and agree to be monitored with frequent (UDTs). Patients with aberrant drug-taking behaviors required more structured prescribing and monitoring, including frequent visits, prescribing small quantities of opioids, more frequent UDT, pill counts, and education and counseling.

Phase III consisted of maintenance and evaluation. The program was monitored by regular team meetings and review of patients and protocols. Evaluation of the program included patient adherence to OTA, PCP satisfaction and rates of use of OTAs and UDTs by PCPs.

Outcome Measures

The outcomes of the program on providers, pharmacy budget, and patients were evaluated through December of 2003, 22 months after the start of Phase I.

Providers: The effect of the program on PCPs was measured by behavioral changes and by satisfaction. PCP behavior change was assessed by counting the absolute number of UDTs and OTAs ordered by PCPs on the electronic medical record. PCP satisfaction with the program was assessed by a questionnaire (Appendix II). Pharmacy: The effect on the pharmacy budget goals for 2001–2002 was measured by extracting cost data on patients. Patients: The impact on patient care was assessed by the percentage who were adherent to the OTAs and by results of UDTs.


Provider Behavior

The number of OTAs more than doubled from their baseline in 2001 (63) to 2002 (144) and more than tripled by 2003 (214). The increase in UDT testing is illustrated in Figure 1, which shows a slight increase from the first month, February 2002 (74), to the second month, March 2002 (84), a relatively stable rate over the next 6 months (range 84–148), and then a steady increase to an average of 200 per month over the last 6 months of data collection (July–December 2003). Following the initiation of the program in February 2002, the largest percentage increase in use occurred between May 2002 and October 2002. Pharmacy budget goals for reducing costs of oxycodone SA were met (see Table 1), with a reduction from $129,793 (Q1 FY 01) to $5,236 (Q1 FY 03). The number of prescriptions for opioids per month remained constant, demonstrating a shift from oxycodone SA to other less costly long-acting opioids.

Figure 1

Rate of increase in use of UDT by PCPs. UDT = urine drug testing; PCP = primary care provider.

View this table:
Table 1

Pharmacy data

Total No. Opioid RxTotal Cost for All Opioids% Oxycontin®/All OpioidsOxycontin® Costs
Q1 FY 015,202$190,68122.5$129,793
Q2 FY 015,273$206,24222.3$127,528
Q3 FY 015,662$174,44716.0$100,067
Q4 FY015,265$153,28812.3 $67,495
Q1 FY024,821$130,133 9.6 $47,595
Q2 FY024,801$121,059 6.2 $31,136
Q3 FY025,034$111,898 0.8  $6,161
Q4 FY024,909$101,579 0.5  $4,721
Q1 FY034,959$109,868 0.4  $5,236

Reduction in Utilization of Health Care Services

We conducted a retrospective analysis of the first 108 patients enrolled in the ORC. We compared resource utilization, calculated as the average number of visits monthly during the 12 months before enrolling in the ORC compared with the average number of visits per month after enrollment. Preliminary analyses showed that ORC enrollees demonstrated an average decline, per patient, in ER visits of 72.7% and unscheduled PCP visits of 59.6%.

Provider Satisfaction

Table 2 lists how the providers rated the impact of the program on their practice. A total of 35 PCPs received questionnaires; 19 (54%) were returned. In total, 84% of the those who responded referred to the service. The majority of the PCPs who completed the survey found the program helpful in their practice, both in changing abnormal illness behavior (walk-ins, medication complaints) and in freeing up more time to deal with important medical problems. Table 3 lists the comments written about the program at the end of the questionnaire. These comments were uniformly positive.

View this table:
Table 2

Results of provider satisfaction survey

Change in PCP practice
I use the opioid agreement more often.
  37%Strongly agree
I routinely order urine drug screens.
  63.2%Strongly agree
The consistent approach to chronic opioid therapy promoted by the Pain Team and put into practice by the Opioid Renewal Clinic has helped me feel comfortable in managing chronic pain.
  61%Strongly agree
I receive fewer complaints regarding pain medications.
  65%Strongly agree
There are fewer walk-ins for pain management issues.
  76%Strongly agree
I can spend more time with the patients' other medical problems when they are being medical problems when they are being followed by this service.
  76.4%Strongly agree
  • PCP = primary care provider.

View this table:
Table 3

Comments from PCPs

Comments from PCP Satisfaction Questionnaire
• This service is excellent and a big help with our most difficult patients.
• The service has made a positive impact on our patients.
• I appreciate the service availability by phone for questions while I am seeing patients.
• This program has made my life and my patients life easier. It gives me time to address all the other important health care issues during the visit.
• It has helped me to improve my relationship with my chronic pain patients.
• I strongly am in favor of this program. Where I came from, primary care was buried in opioid renewals and behavioral issues.
• This service has been a godsend for primary care providers. It has enabled us to effectively treat chronic pain.
It allows us the time to care for other medical problems. It also gives us an effective mechanism for dealing with very prevalent substance abuse issues.
• Before this service, patients seeking opioids would disrupt the delivery of care. Shouting in the waiting area was common.
Security frequently had to be called in. Providers felt threatened. Prescriptions for opioids were not managed systematically.
Providers who could not handle these patients had nowhere to turn. The establishment of this service has ended the disorganized way we managed pain and introduced a safer, more effective approach. It's one of the best innovations I've seen in primary care.
  • PCP = primary care provider.

Patient Behavior

A total of 335 patients were referred to ORC over a 22-month period for the structured treatment strategy. In total, 171 (51%) were referred because of documented aberrant drug-taking behaviors. Of this latter group, 166 patients had urine toxicology testing positive for illegal drugs, unprescribed medications or were consistently negative for prescribed medication (see Table 4). Five patients' aberrant behavior consisted of one or more of the following behaviors: frequent early renewal requests, not following plan for renewal, getting opioids from multiple providers.

View this table:
Table 4

Outcomes of referred patients (N = 335)

OutcomesNumber (%)
171 (51%) documented aberrant drug-taking behaviors
  Resolution of aberrant behaviors 77 (45)
  Self-discharged from ORC 65 (38)
  Referred for addiction treatment 22 (13)
  Consistently negative UDT (weaned from opioids)  7 (4)
164 (49%) no documented aberrant drug-related behaviors at referral
  Adherence to OTA164 (100)
  • ORC = Opioid Renewal Clinic; UDS = urine drug testing; OTA = opioid treatment agreement.

Table 4 presents the outcomes of these referred patients. Of the patients who were referred for aberrant drug-taking behaviors (N = 171), 45% adhered to the OTA, but 38% self-discharged from the practice when the structured program was offered. Only 22 (13%) required referral for addiction treatment. Seven (4.09%) were weaned from opioids due to consistently negative urine drug tests. In total, 164 of referred patients (48.9%) had no documented aberrant behaviors. They were referred due to complexity including h/o substance abuse, conflicts with PCP regarding opioids, opioid rotation or titration. This group continued to adhere to the OTA without any aberrant behaviors.

Table 5 presents the number and percentage of each drug that was found collectively in the UDTs, with cocaine and THC being by far the most commonly abused.

View this table:
Table 5

Urine drug testing (UDT) results of the N = 166 referred with abnormal UDT

Abnormal UDT DrugN (%)
Cocaine61 (37.75)
THC60 (33.14)
Morphine24 (14.46)
Benzodiazepines22 (13.25)
Oxycodone13 (7.83)
Propoxyphene13 (7.83)
Hydrocodone11 (6.63)
Codeine 8 (4.82)
Methadone 4 (2.41)
6-acetyl morphine (heroin) 3 (1.81)
Butalbital 1 (0.60)
Meperidine 1 (0.60)
PCP 1 (0.60)
  • UDT = urine drug testing.


During the last decade, the use of opioids for the treatment of noncancer pain, once almost solely in the realm of pain specialists, has increased in the primary care setting [3,4,29]. PCPs are now expected to manage this controversial and technically difficult therapy, often without guidance from pain management specialists. Although guidelines exist [17,26–28] and hospital policies support opioid therapy, PCPs often feel ill-prepared and overwhelmed by its demands. Thus, opioids are usually either avoided or, if prescribed, guidelines are followed variably for various reasons, particularly time constraints, but also the personal opinions of the provider rather than evidence. The implementation of the ORC demonstrates that a primary care-based pain service that supports PCPs in managing opioids can overcome barriers to opioid prescribing. To our knowledge, there is no other literature documenting such results in a pain management program in primary care.

After 22 months, the ORC demonstrated improvement in all outcomes: 1) change in provider practice measured by increased use of urine drug toxicology testing, increased documentation of the OTA, and decreased prescribing of oxycodone SA; 2) improved patient adherence to the OTA; and 3) the differentiation of addiction and other aberrant behaviors as well as documentation, by negative UDT, of possible diversion.

Change in PCP Practice

The use of OTAs and monitoring with UDTs are widely accepted tools intended to improve patient adherence with opioid therapy [47]. Although the Federation of State Medical Boards of the United States guidelines [28] for opioid prescribing recommend these tools, consistent with our experience, they are not routinely used [29,44]. Prior to the implementation of the ORC, an average of 20 UDTs were ordered per month by Primary Care. OTAs were documented on average two to four a month in the 6 months prior to starting the program. PCPs reported that they only required OTAs and UDTs for patients they deemed as potentially problematic patients.

After implementation of the program, the added support provided by immediate access to the clinical pharmacist and NP came with a prerequisite: an OTA and routine UDTs on all patients treated with opioids for chronic noncancer pain. When the PCPs first started performing UDTs on patients they were already following, in some cases for years, the number of positives (positive for illicit or unprescribed medications or negative for prescribed medications) was surprising (see Table 3). This is similar to the findings of Katz and Fanciullo [44] who found positive UDTs (either positive for illicit drug, unprescribed controlled substance or negative for prescribed medication) in 27% of patients with no previous behavioral issues and another series [53] which found that 21% had concealed substance misuse from their providers when it was discovered by UDT. Positive urines illustrated for PCPs the importance of UDTs for all patients on chronic opioid therapy, rather than their prior practice of selecting patients based on opinion. Our findings support Gourlay et al.'s proposal for “universal precautions” for all patients being considered for opioid therapy [46]. This universal application of the chronic opioid policy (including OTA and UDT) in a respectful and matter-of-fact manner was surprisingly well received by our patients.

Based upon the authors' formal and informal daily contact while working in close proximity to the PCP practice, the process of building confidence in PCPs' use of opioids appears to be due to at least four factors. First, daily access to the support of the NP, Pharm.D, and the multidisciplinary Pain Team meant that PCPs never felt isolated and on their own in making decisions. Second, the Pharm.D's documentation provided them with laboratory data and clinical outcomes data to support their clinical decisions and to increase their confidence in discussing these decisions with their patients—for example, enabling them to say “it wasn't just my decision.” Third, when good outcomes were documented, this increased their confidence that they could prescribe opioids with success even in the context of apparent aberrant behaviors. Fourth, by outlining clear responsibilities in the treatment agreement, PCPs gave the patient an opportunity to participate in the clinical process of demonstrating that opioids were safe, manageable, and effective rather than burdening the PCPs with that decision based upon bias and unreliable data.

Patient Outcomes

All of the 335 patients referred to the ORC over a 22-month period were classified as having a medically stable noncancer pain condition. A total of 164 referrals were considered complex and “at risk”—these will be discussed in more detail below. In total, 171 referrals were patients with objective aberrant behaviors, and of these 77 (45%) were able to resolve their aberrant drug-taking behavior and continue opioid medication; 65 (38%) chose not to follow the structured program and left the ORC and 22 (13%) remained in the program but were unable to adhere to the OTA and were referred for addiction treatment. Based on the entire sample (N = 335), the rate of manifest addiction was 6.5%. However, if one assumes that those who dropped out have addiction or substance abuse, the rate in the entire sample (N = 335) may be as high as 26% (22 referred for addiction treatment and 65 self-discharged).

Importantly, our structured program reduced the apparent number of patients with addiction (assuming that aberrant drug-taking behaviors equals addiction, as many clinicians erroneously do), from N = 171 to the actual N = 22 with confirmed addiction who were referred for treatment. These results suggest that a health care system can anticipate considerable savings in professional hours by avoiding unnecessary referrals for addictionology consultations for the large group of patients, in this case 77 patients whose aberrant behavior resolved in the ORC.

Of the 171 patients referred for aberrant behavior, we hypothesize that the 65 (38%) who initially refuse participation may have done so for obvious reasons such as ongoing addiction, diverting medications for profit or to obtain illegal drugs or, simply not wanting to be bothered with the structured program. However, this study cannot test this. We do know from our longitudinal clinical work in primary care that a few of the patients who initially refused monitored opioid therapy in ORC eventually returned and agreed to participate—these “returnees” are not reflected in our data as it was outside the time frame of our project. Further studies will be needed to elucidate causes of initial refusal. Consideration of the problem of diversion is particularly important. The recent national mandates to identify and treat pain have occurred with a concomitant increase in illicit use of prescription opioids [54]. To reduce diversion and to preserve the legitimate treatment of chronic pain with opioids, published clinical practice guidelines for prescribing opioids for noncancer pain include monitoring both for effectiveness and for aberrant drug-taking behaviors [55,56].

Available data suggest addiction rates of 6–16.7% in the general US population [57], 20–26% in hospitalized patients, and 40–65% in trauma patients [58]. Our addiction rates fall within these ranges although meaningful comparisons are difficult due to methodological differences such as in criteria for selection of study samples, in the definitions of abuse and addiction, and in methods of assessment. This problem manifests as well in estimates of prevalence rates in primary care populations. Reid et al. [3], in a retrospective review of diagnoses recorded in the medical record of patients being treated with opioids for chronic noncancer pain, reported an 18% lifetime prevalence of narcotic abuse/dependence in a VA PCC (VA) compared with a 38% prevalence in a university-based PCC. Documentation of “prescription opioid abusive behaviors” was found in 24% (VA) and 31% (PCC). Other retrospective chart reviews documented rates of drug abuse rates of 6% in a large university-based primary care practice [29] and 21% in a VA sample [4]. In our review of 208 patients referred for a chronic noncancer pain consultation at the Philadelphia VA from 2000 to 2002, 22% had documented addiction disorder.

In contrast to these cross-sectional chart reviews, confidence in our categorization lies in the strength of the longitudinal, prospective assessment of actual substance abuse behaviors within a structured program. The diagnosis of addiction was made based upon observed behaviors over time while pain was managed aggressively, therefore eliminating the possibility that behaviors were due to pseudoaddiction.

Of the 335 referrals, 164 (49%) were complex “problem patients” (e.g., past history of addiction, conflicts with PCP over opioids, pharmacological complexity) who the PCPs felt unprepared to manage on their own and were referred for the “potential” of developing substance misuse. It is important to note that all of these patients successfully adhered to the OTA using the minimum amount of structure (e.g., brief monthly visits to the pharmacist) in the ORC protocol. The risks of exposing individuals with past or present addiction disorder to opioid analgesics are not well studied. Many experienced pain clinicians believe that the majority of individuals with chronic pain can be managed safely with opioids if monitored appropriately [10,16,23,26,50], whereas some practitioners believe that the risks outweigh the benefits [59]. A more refined, nuanced assessment of risk, based upon evidence not opinion, is not possible given our present knowledge. Thus, we as a field are challenged to develop methods of distinguishing patients with chronic pain who have, or are at risk for, a substance abuse disorder.

Our study's several limitations preclude generalizations. We cannot be sure how often and to what degree patients in our sample received other health care outside the VA system, as many patients have additional insurance in their family. We did not measure important pain treatment outcomes, such as physical and psychosocial functioning. The sample size is relatively small and derived from a Veterans Medical Center in an urban setting. Other health care systems, without an integrated electronic medical record and the availability of easy team consultation, may be challenged to implement this system. Thus, testing this intervention in larger, more demographically diverse samples drawn from several different settings would provide results that could be more easily generalized to the larger health care system. For example, would this program be cost-effective or even feasible within a suburban context with less poverty and addiction comorbidity? Finding demographically similar groups matched for disease and illness severity would enable us to analyze differences between this approach and treatment as usual in different settings. Each clinical settings will differ in the experience and training of available practitioners that can be convened to form a multidisciplinary team to support an ORC.

Future studies of the Pain Medicine and Primary Care Community Rehabilitation Model, instructing PCPs to use evidence-based medication management algorithms for different types of pain (e.g., neuropathic, nociceptive, myofascial) and including a broader range of treatments, including acupuncture, behavioral therapies, and interventional pain medicine, should further improve performance as measured by patient outcomes and reduced system costs. The relative success of this intervention indicates that further multicenter studies of the ORC approach are warranted, using additional measures of outcome such as pain levels and psychosocial, physical, and occupational functioning. It is hoped that a gold standard of outcomes, measurable behavior, such as is available in an integrated and self-contained health system such as VA, will be utilized in such studies.


We implemented a structured program for opioid management for chronic noncancer pain in a primary care setting, based on the current standards and guidelines for opioid prescribing. Strategies utilized in the NP/clinical pharmacist-run clinic included: standardized documentation, OTAs, UDT, frequent visits, and patient education. We demonstrated that PCPs increased their use of OTAs and UDT and prescribed opioids with more confidence. We suggest that an ORC-like model, tailored to the needs of each particular setting, can help make this important treatment modality available to more patients at lower risk.

Appendix I


Appendix II

Opioid Renewal Clinic

Provider Satisfaction and Program Evaluation

1. The goals of the program and services offered are clear to me.54321
2. The strategies utilized by this service to manage patients on chronic opioid therapy have influenced my practice.54321
  2a. I use the opioid agreement more often.54321
  2b. I routinely order urine drug screens when indicated.54321
  2c. Once I make the decision to start chronic opioids, I titrate to effectiveness.54321
  2d. I automatically initiate constipation prophylaxis when I start opioids.54321
3. Have you referred patients for this service?YesNo
4. I have never referred but I use the service as a resource for assistance with management.frequentlyoccasionallynever
5. I have referred to the service and I also utilize the service as a resource for assistance with other cases.frequentlyoccasionallynever
If you have referred please complete the following:
6a. The methods of communication between the Opioid Renewal Clinic Clinicians and me are helpful.54321
6b. The Pharm.D's progress note provides pertinent information.54321
6c. I prefer to be notified by View Alert (note requiring cosignature) and email.YesNo
6d. I prefer to be notified by email and would prefer not to have to cosign the note.YesNo
Complete whether you have referred or not:
7a. The service has a positive impact on the Primary Care practice.54321
7b. I receive fewer complaints regarding pain medications.54321
7c. There are fewer walk-ins for pain management issues.54321
7d. I can spend more time with the patient's other medical problems when they are being followed by this service.54321
8. The consistent approach to chronic opioid therapy promoted by the Pain Team and put into practice by the Opioid Renewal Clinic has helped me feel comfortable in managing chronic pain.54321
9. Your patients have reported their level of satisfaction with the Opioid Renewal Clinic asVery SatisfiedSatisfiedDissatisfiedNo Comment


View Abstract