Access to the PPM Journal and newsletters is FREE for clinicians.
9 Articles in Volume 8, Issue #1
Spine-related Pain in Sports Medicine
Outpatient Interventional Treatments for Migraines and Pain Flare-ups
Identifying Abusers Prior to Initiating Chronic Opioid Therapy
Urine Drug Tests in a Private Chronic Pain Practice
Platelet Rich Plasma (PRP) Matrix Grafts
Role of Sustained-release Opioids in Treating Chronic Pain
Adenoid Cystic Carcinoma of the Parotid Gland
Evaluation and Management Codes Drive Medical Necessity
Grappling with the Ethics of Practical Pain Management

Identifying Abusers Prior to Initiating Chronic Opioid Therapy

Strategies for prescribing physicians to identify chronic pain patients at high risk for inappropriate utilization of prescription opioid analgesics.

In the 1980’s until the mid-1990s, opioids were accepted as the mainstay of cancer pain treatment.1 Yet, in interdisciplinary programs for patients with chronic pain of nonmalignant origin, the use of opioid analgesics was considered problematic2 and detoxification from opioids was considered a primary goal of treatment.3,4 In 1996, Portenoy5 suggested that there was a subpopulation of patients with chronic pain of benign origin that could benefit from chronic opioid therapy—without intolerable side effects or the development of aberrant drug-related behavior. At that time, the pain management community witnessed a clear swing of the pendulum, resulting in the widespread use of progressively increasing dosages of strong opioid analgesics.

Today there still exists a degree of consensus that such treatment may be beneficial to certain patients suffering from chronic pain of nonmalignant origin. Yet, concerns regarding the potential abuse of prescription opioids are mounting. While initial estimates of potential prescription opioid abuse among chronic pain patients was thought to be quite low—with Portnoy and Foley6 estimating that only 5% of such patients were at risk—some recent estimations range significantly higher.7-14 In 2002, Joranson and colleagues15 determined that the rate of prescription opioid abuse in the United States increased 71% between 1997 and 2002. It has also been reported that while prescription opioids accounted for 5.75% of drug abuse in the United States in 1997, this figure had risen to 9.85% by 2002.16 Finally, Cicero and colleagues17 recently reported (in late 2007) that rates of prescription opioid abuse have increased every quarter over the past 3 ½ years. It is uncertain whether the increase in prescription drug abuse is by pain patients, per se, or diversion by pain patients, manufacturers, pharmacies, or wholesalers. Regardless, all prescribing physicians must attempt to balance abuse, diversion, and therapeutic risk.

While the physician is morally obligated to relieve pain and suffering, he/she must simultaneously attempt—within his or her practice mileu—to identify potential abusers and diverters. Unfortunately, the “science” of risk assessment is still in its infancy, and selection remains an overwhelming challenge. Ideally, physicians prescribing opioids to patients with chronic pain will consult with an appropriately trained and experienced pain psychologist in order to obtain an evaluation that will serve as a guide in the selection process. Tragically, however, access to qualified pain psychologists appears to be becoming progressively more limited, with this scarcity more pronounced in non-urban areas. Schatman18-20 has written extensively about the demise of the multidisciplinary chronic pain management clinic over the past decade. Many of these pain treatment facilities that have closed their doors due to financial factors were housed in teaching hospitals that offered fellowship training programs to psychologists that intended to specialize in pain management. Turk and Burwinkle21 have noted the paucity of training opportunities in interdisciplinary treatment available to psychologists. Within the Veterans Administration, the initial lack of availability of pain psychologists to treat injured veterans of Operations Enduring Freedom and Iraqi Freedom has been noted.22 Additionally, many fellowship-trained pain psychologists have chosen to pursue academic careers as opposed to serving as much-needed clinicians.23 Accordingly, the rest of this article will focus on strategies that physicians can implement in order to increase the likelihood that the patients to whom they are prescribing opioids for chronic pain are at the lowest possible risk for abuse of these potentially beneficial yet dangerous medications.

Behaviors Indicative of Prescription Opioid Abuse

Confusion and disagreement regarding what constitutes prescription opioid abuse has been apparent. Butler et al.24 offered a reasonably comprehensive list of behaviors that are indicative of abuse:

  • Selling prescription drugs
  • Forging prescriptions
  • Stealing drugs
  • Injecting oral formulations
  • Obtaining prescription drugs from non-medical sources
  • Concurrently abusing ETOH or illicit drugs
  • Escalating doses on multiple occasions or otherwise failing to comply with the prescribed regimen despite warnings
  • “Losing” prescribed medications on multiple occasions
  • Repeatedly seeking prescriptions from other physicians or ER’s without informing the original physician
  • Deterioration of function relating to drug use.

Risk Factors for Prescription Opioid Abuse

While the body of published research addressing risk factors for opioid abuse in general is substantial, investigations identifying chronic pain patients at risk for prescription opioid abuse are relatively scarce. It is important to recognize that neither the risk factors identified in the general population nor the assessment tools utilized to identify can necessarily be generalized to the chronic pain population, as the circumstances of their distress and predisposing psychological factors of the groups are not identical. Accordingly, the studies cited in this analysis shall be limited to those specific to patients suffering from chronic pain. For this population, predictors of prescription opioid abuse that have been investigated are summarized in Table 1.

Table 1. Risk Factors to Know Before Prescribing Opioids
  • History of substance abuse
  • Cigarette smoking
  • Under age 25
  • Male
  • History of mental treatment

Lifetime History of Substance Abuse. One of the strongest predictors of prescription opioid abuse among chronic pain patients is thought to be a lifetime history of substance abuse.6,8,11,12,25-27 Interestingly, while current substance abuse has been found to be predictive of prescription opioid abuse, this risk factor has been examined in only several investigations.8,13,28 Not surprisingly, histories of polysubtance abuse have been found to be more predictive of prescription opioid abuse than have histories of single substance abuse.12,13,29-31 Several studies have identified the risk factor of arrests for DUI’s or drug-related offenses.7,13,32 While a Medline search yields thousands of studies indicating that family histories of substance abuse put relatives at heightened risk of abuse in the general population, only investigations by Michna et al.33 and Akbik et al.34 identify family history as a risk factor for prescription opioid abuse.

Cigarette Smoking. Cigarette smoking has also been determined to be a risk factor for prescription opioid abuse among chronic pain patients.13,29,33,35,36 This finding is not surprising, given the addictive nature of nicotine and what is widely considered in the medical world bad judgment given the abundant evidence of the serious health risks associated with smoking.

Patient’s Age. In terms of demographic predictors, only patient age has been consistently implicated, with younger chronic pain patients found to be at higher risk than older patients.8,11,26,34 This finding is consistent with the literature on drug abuse in general indicating a negative correlation between age and risk for substance abuse,37-41 and may relate to differences in values between cohorts.42

Patient’s Gender. While a number of studies have identified gender differences in risk of prescription opioid abuse among chronic pain patients, a careful review of the literature indicates that the findings are mixed, and the evidence is accordingly inconclusive.

Emotional Status. Emotional status has been strongly implicated as a risk factor for inappropriate utilization of prescription opioids among chronic pain patients. Numerous studies10,12,13,28,43-45 have indicated that depressed chronic pain patients are at greater risk for prescription opioid abuse than are patients who are not depressed. While anxiety has been identified as a risk factor for prescription opioid abuse among people suffering from chronic pain, the number of studies that have investigated this variable is limited.28,31,44 Although there exists a relative paucity of empirical literature supporting anxiety as a risk factor, results of the three studies that have been published are consistent with animal models, as stress has been found to be one of the main stimuli provoking opioid consumption in rats.46

Assessing Risk of Abuse

Clinical Interviews and Psychometric Tools. As physicians often sense that in-depth patient interviews and psychometric testing are too time-consuming to be considered a routine aspect of their pain practices, they will often rely upon their “gut feelings” in assessing risk of medication abuse in their practices. Unfortunately, doing so may not be prudent. For example, investigators have found that physicians judged only 13.9% of chronic pain patients who were prescribed opioids as manifesting aberrant drug-related behaviors, when approximately 50% produced positive urine toxicology screens for illicit drugs.45 Patient self-report has been found to be notoriously unreliable in regard to opioid use in chronic pain.45,47-49 Additionally, Akbik and colleagues34 found that patients who believe that their responses may determine whether they are considered for opioid therapy may underreport their behavior. Accordingly, more formal assessment tools should be used to predict potential aberrant drug-related behavior.

As mentioned earlier, one of the most unfortunate mistakes made in efforts to predict opioid abuse among chronic pain patients is the practice of relying on interviews and psychometric measures that were developed for the general population or drug abusers without chronic pain. In doing so, clinicians fail to recognize the uniqueness of chronic pain sufferers, as their qualitative experience of pain and its sequelae need to be considered from a more phenomenological perspective than is typically the case.50 It is essential that physicians recognize that the motivation underlying opioid abuse is not necessarily identical to that of “purely recreational” drug abusers.

Research on prediction of opioid abuse among chronic pain patients often relies on structured and semi-structured clinical interviews for data collection. One commonly utilized assessment is the Structured Interview for DSM-IV Substance Abuse Module.51 This assessment is commonly used by chemical dependence counselors and researchers. Savage52 has noted that the DSM-IV criteria emphasize tolerance and physical dependence, which are normal and expected consequences of long-term opioid treatment, and do not specifically indicate misuse. Additionally, this DSM-IV Structured Interview has never been validated with a chronic pain population. Given these factors in conjunction with its length of administration (30-60 minutes), the Substance Abuse Module is an inappropriate tool for risk assessment among chronic pain patients in a clinical setting.

Similarly, the Addiction Severity Index (ASI)53 is a semi-structured clinical interview that provides multidimensional measurement of seven functional domains. Unfortunately, the current version (ASI-6) requires 45-75 minutes to complete, and also lacks validity data on prediction of prescription opioid abuse. Another semi-structured interview, the Prescription Drug Use Questionnaire (PDUQ)54 is psychometrically appropriate for prescription opioid abuse risk assessment. However, the PDUQ is also a relatively lengthy interview to conduct, as well as to score, and requires a trained clinician for its administration, thereby limiting its utility in the fast-paced chronic pain clinical setting.

Appropriate Opioid Risk Screening Tools

The assessment tools designed specifically for the prediction of prescription opioid abuse in chronic pain patients that appear to be most reliable and valid are the Screener and Opioid Assessment for Patients with Pain (SOAPP)55 and the Pain Medication Questionnaire (PMQ).56 While extremely valuable in assessing risk, it is important to remember that psychometric measures are not litmus tests; rather, they should be used within the context of a process of chart reviews, thorough histories, and other supportive materials in order to understand the gestalt of a patient’s risk profile.

The SOAPP is comprised of 24 Likert-scale items addressing issues including substance abuse history, family history of substance abuse problems, problems with medication use, psychosocial problems, tobacco use, and legal problems. Of the 24 items, the 14 that were found to be most predictive of aberrant behavior are included in the scoring, and accordingly, either the 24- or 14-item version can be administered. The sensitivity and selectivity of the SOAPP were found to be adequate by its authors. A longitudinal multicenter study to develop and cross-validate a revised version of the SOAPP that includes more subtle items has recently been completed.57 Perhaps the greatest weakness of the original version of this widely used tool is that it provided a single cut-off score, thereby allowing for only a dichotomous identification of patients as being either at high or low risk of prescription opioid abuse. The SOAPP-R, however, allows the clinician to categorize the patient as being at low, medium, or high risk.

The PMQ56 is comprised of 26 items scored on a 5-point Likert scale on which patients indicate their degree of agreement or behavioral conformity. In addition to providing short-term prediction of aberrant opioid-related behavior, longer-term predictive validity of the PMQ for abuse-related behavior six months after evaluation and treatment has been established.57 Like the SOAPP-R, the PMQ categorizes patients into three risk level groups. One of the benefits of the PMQ over the SOAPP is that its selectivity and sensitivity are considerably higher, with both figures approaching 0.90.58 The SOAPP-R and the PMQ hold great promise, and their authors are in the process of conducting additional research in order to further establish their validities and reliabilities. To this point, no one has established that any one instrument or interview format is superior to any other in predicting opioid misuse among chronic pain patients. Accordingly, there is not yet a “gold standard.”

A somewhat less valid measure for the assessment of addiction risk than those mentioned above is the Opioid Risk Tool (ORT).43 The ORT has actually demonstrated high sensitivity and selectivity in predicting future aberrant opioid-related behavior among chronic pain patients. Easy to administer and score, the ORT consists of 10 items that are endorsed by the patient, with the items weighted and summed in order to categorize patients as being at low, moderate, or high risk. Unfortunately, the authors failed to test the validity of the weights assigned to the individual risk factors, thereby limiting the overall validity of the measure.

In addition to utilizing a measure designed specifically to predict prescription opioid abuse in chronic pain patients, the abundance of studies indicating that depression and anxiety are potential predictors of abuse suggests that physicians should also assess these variables prior to initiating chronic opioid therapy. Given physicians’ difficulties in identifying depression and anxiety during routine appointments,59-65 easily-used psychometric measures of depression and anxiety should assist the prescriber in his/her efforts to determine which chronic pain patients are at risk for using opioids so as to inappropriately medicate their emotional distress.

A number of well-established, highly reliable and valid measures of depression that can be quickly administered, scored, and interpreted are available to physicians. Most commonly used are the Beck Depression Inventory-2 (BDI-2),66 the Center for Epidemiological Studies Depression Scale (CES-D),67 and the Zung Self-Rating Depression Scale.68 All of these measures are brief and easily scored and interpreted, requiring 5-10 minutes. Additionally, all three of these inventories have been validated among both chronic pain and chemical dependence populations in myriad investigations. The CES-D, however, has the advantage of actually having been determined to predict prescription opioid abuse in chronic pain patients.60,61 While these inventories are thought to measure slightly different aspects of depression, the correlations between the three tools have been found to be quite high.69 Accordingly, each is appropriate for screening patients with chronic pain prior to initiating opioid therapy, and the decision regarding which measure to use can be left to clinician preference.

While numerous well-validated tools are available for the measurement of depression, there exist fewer highly-regarded instruments for the measurement of anxiety. Most commonly used is the Beck Anxiety Inventory (BAI).70 Like the BDI-2, the BAI consists of 21 symptoms rated on a 4-point scale of progressive severity, and requires only 5-10 minutes to complete and score. The BAI has been psychometrically-supported as a measure of state anxiety in chronic pain patients.71 Another commonly used measure of anxiety is the State-Trait Anxiety Inventory (STAI).72 This tool is composed of two 20-item scales, offering the advantage of distinguishing between temporary (“state”) and characterological (“trait”) anxiety. The STAI’s ability to make this distinction may be important, as trait anxiety has been determined to predict substance abuse in the general population.73-76 Additionally, trait anxiety, but not state anxiety, has been found to relate to substance abuse relapse.77 On the other hand, although widely used in medical settings, the STAI has been determined not to clearly differentiate anxiety disorders from depressive disorders.78,79 The studies that have identified anxiety as a risk factor for prescription opioid abuse specifically in chronic pain populations have relied upon structured interviews rather than the STAI or the BAI to measure anxiety. However, as lengthy structured interviews are impractical in the clinical setting, either of these brief measures is appropriate as a source of valuable information to the prescribing physician.

Inappropriate Questionnaires and Tools

Questionnaires are typically easier to administer than are structured and semi-structured interviews, and are less time-consuming. One of the briefest and most commonly used questionnaires adapted for the prediction of drug abuse is the CAGE-AID.80-82 Adapted to include drugs from the original CAGE screen for alcoholism, the CAGE-AID consists of four simple questions regarding a perceived need to cut down on substance use, whether others are annoyed by one’s substance use, guilt relating to substance use, and the perceived need to use substances as “eye-openers” in order to steady one’s nerves. Despite its popularity, the CAGE-AID has never been validated as a predictor of prescription opioid abuse, and its use for this purpose accordingly ought to be avoided. Other commonly used tools that should not be used due to their lack of established predictive validity for prescription opioid abuse among chronic pain patients include the Michigan Alcohol Screening Test (MAST),83 the Drug Abuse Screening Test (DAST),84 the Screening Instrument for Substance Abuse Potential (SISAP),85 the Substance Abuse Subtle Screening Inventory (SASSI-3),86 and the Screening Tool for Addiction Risk (STAR).87 Also inappropriate for use for prediction of prescription opioid abuse among chronic pain patients are three scales from the Minnesota Multiphasic Personality Inventory-2 (MMPI-2): The MacAndrew Alcoholism Scale-Revised (MAC-R), the Addiction Potential Scale (APS), and the Addiction Acknowledgement Scale (AAS).88 While substance abusers have produced higher mean scores than non-abusers on all three of these scales, they were developed for assessing potential alcohol abuse. And, as is true of so many instruments that are inappropriately used, their predictive validities for prescription opioid abuse in chronic pain patients has never been established.

Urine Toxicology Screens

While it is beyond the scope of this article to discuss the benefits of urine toxicology screens (UTSs) in detail, one notable investigation should be mentioned. Katz and colleagues26 determined that use of UTS in conjunction with monitoring patient behavior was found to identify 17% more opioid abuse among chronic pain patients than behavior monitoring alone. Typically, physicians who require UTSs from their patients order them during the course of chronic opioid therapy. In a 2001 study,89 investigators found that only 8% of primary care physicians who prescribed opioids utilized UTSs. Results of a 2006 study by Bhamb and colleagues90 suggest that UTSs are becoming more widely used, with 15% of primary care physicians utilizing them. However, only 6.9% reported obtaining this test prior to prescribing opioids. In addition to making urine screens a standard component of opioid prescription practice, more physicians should consider requesting UTS before initiating chronic opioid therapy as well. Results consistent with a patient’s use of medications and/or illicit substances can help establish his/her veracity, thereby providing additional risk data to the prescribing physician.

Recommendations and Conclusions

In summary, prior to initiating chronic opioid therapy, physicians should consider the following steps:

  • Taking a detailed history in order to assess substance abuse problems, including questioning the patient regarding DUI’s and drug-related arrests
  • Considering the patient’s use of tobacco
  • Screening for depression (using the BDI-2, CES-D, or Zung Self-Rating Depression Scale) and anxiety (using either the BAI or STAI)—as depression and anxiety are both identified risk factors for abuse, considering initiation of psychopharmacotherapy and/or referral for counseling prior to initiating opioid treatment is probably prudent
  • Using either the SOAPP or the PMQ as standard risk assessment tools
  • Performing UTS at time of initial consideration, as well as during the course of chronic opioid therapy to assess trustworthiness
  • Not putting the patient or physician at risk by going with one’s “gut feeling”
  • Documenting efforts to assess risk is critical—psychological screening prior to initiation of chronic opioid therapy will never be 100% accurate; however, proof of effort goes a long way in terms of reducing physician liability.

At present, there exists no irrefutable “formula” to identify chronic pain patients at high risk for inappropriate utilization of prescription opioid analgesics. What represents a “yellow flag” to one physician may represent a “red flag” to another. The science/art of risk prediction is still young and accordingly uncertain. However, the risk factors and measures that have been reviewed provide the prescribing physician with a group of tools that can be easily used to make chronic opioid therapy safer for him/herself as well as for the patient. “Risk” is not synonymous with a substance abuse disorder or addiction. However, identification of a patient at-risk may warrant additional attention, more in-depth evaluation, discussion with the patient, and more aggressive monitoring. All of these physician behaviors will enhance the overall quality of pain management services provided while minimizing risk for the patient and the medical practitioner.

Last updated on: February 21, 2011
close X