DRUG SAFETY
Identifying Controlled Substance Patterns of Utilization Requiring Evaluation Using Administrative Claims Data Stephen T. Parente, PhD; Susan S. Kim, PharmD; Michael D. Finch, PhD; Lisa A. Schloff, MS; Thomas S. Rector, PharmD, PhD; Raafat Seifeldin, PharmD, PhD; and J. David Haddox, DDS, MD Objectives: To develop a systems approach to identify, for further evaluation, patients with potential controlled substance misuse or mismanagement using software queries applied to administrative health claims data. Study Design: Retrospective validation of the system using insurance claims. Patients and Methods: Data from administrative health claims databases representing nearly 7 million individuals younger than 65 years were used by multidisciplinary expert panels to develop and validate controlled substance patterns of utilization requiring evaluation (CS-PURE) criteria. Results: Thirty-four CS-PURE queries were developed in SAS and applied to administrative claims records to identify patients with potential controlled substance misuse or mismanagement. From these, we identified 10 CS-PURE with the highest expert agreement that intervention was warranted. Expert panel agreement that CS-PURE correctly identified cases ranged from 48% to 100%, with at least 50% agreement in 9 of 10 CS-PURE. The prevalence rates for CS-PURE ranged from 0.001% to 0.252%. This translates to identifying between 5 and 1116 patients for individual CS-PURE in a 500 000-member health plan. Conclusions: We developed and empirically validated a group of queries using CS-PURE to identify patients with potential controlled substance misuse or mismanagement that would warrant further evaluation by the treating physician, a quality assurance function, or the medical director. Claims-based CS-PURE identification is generalizable to most health insurers with access to medical and pharmaceutical claims records. Although CS-PURE are not direct measures of misuse, they can direct attention to potential problems to determine if intervention is needed. (Am J Manag Care. 2004;10:783-790)
he diversion, abuse, and inappropriate use of controlled substances are subjects of continuing concern among the medical community, insurers, and policy makers. However, a balance must be achieved between preventing diversion and abuse and encouraging the use of controlled substances for legitimate medical need, particularly for pain management.1-3 Several clinical practice guidelines, consensus statements from professional associations, and state
T
VOL. 10, NO. 11
laws and policies emphasize that it is essential for opioid analgesics to be available for the treatment of moderate-to-severe pain and that prescribing should be individualized to the patient.2,4-14 Although some progress has been made in treating pain, undertreatment of pain is still prevalent.15-17 Media coverage of diversion and abuse of controlled substances and uncertainty regarding potential disciplinary action may cause physicians to hesitate when considering treatment for a patient who may require long-term or high doses of opioids.1,18 This is exacerbated when physicians have difficulty discerning between a patient with a legitimate pain problem and one who is feigning pain to obtain drugs for abuse or diversion.19 Because pain is subjective and cannot be measured or ruled out by laboratory tests or physical examination, physicians rely largely on their interpretation of patient interviews and histories to determine a patient’s need for analgesics. However, they often find themselves in the predicament of wanting to treat seemingly legitimate patients without having information about their patients’ prescription drug and medical histories that would help them identify and address any problems. Indeed, a 1999 report from the Institute of Medicine stressed that most medical errors do not result from individual practitioners’ recklessness; rather, they are attributable to faulty processes and systems that lead people to make mistakes or fail to prevent them through lack of information and support in a complex working environment.20 Solving problems within healthcare requires the design of systems and processes to help
From the HSI Network LLC (STP, LAS) and Department of Healthcare Management, Carlson School of Management, University of Minnesota (STP), Minneapolis, and Center for Health Care Policy and Evaluation, United Health Group, Eden Prarie (MDF, TSR), Minn; and Purdue Pharma LP, Stamford, Conn (SSK, RS, JDH), and Department of Anesthesiology, School of Medicine, University of Connecticut, Storrs (JDH). This study was supported by Purdue Pharma LP. Address correspondence to: Stephen T. Parente, PhD, Department of Healthcare Management, Carlson School of Management, University of Minnesota, 321 19th Avenue South, Room 3-149, Minneapolis, MN 55455. E-mail:
[email protected].
THE AMERICAN JOURNAL OF MANAGED CARE
783
DRUG SAFETY Table 1. Prevalence Results for Top 10 Controlled Substance Patterns of Utilization Requiring Evaluation (CSPURE) Most Positively Associated With the Probability of Identifying a Misuse Case Period Prevalence for 2000, %
CS-PURE
Pattern of Controlled Substance Misuse
Database 1 (n = 2 927 237)
Database 2 (n = 782 880)
Projected No. of Patients in a 500 000-Member Health Plan
01
Multiple prescribers (≥6 prescribers for same drug)
0.213
0.252
1116
02
Multiple pharmacies (≥4 different pharmacies for same drug)
0.132
0.135
664
03
Chronic use of carisoprodol (≥4 prescriptions in 6 mo)
0.125
0.105
597
04
Continuous overlap of ≥2 different benzodiazepines for ≥30 d, when 1 is for alprazolam
0.059
0.070
310
05
Estimated ≥4 g/d of acetaminophen
0.028
0.012
118
06
≥2 Prescriptions for meperidine hydrochloride with >2-day supply
0.016
0.023
88
07
Chronic use of butorphanol[AUTHOR: OR “BUTORPHANOL TARTRATE”?] (≥4 prescriptions in 6 mo)
0.015
0.019
81
08
Continuous overlap of ≥2 different benzodiazepines for ≥90 d, when 1 is for clonazepam
0.005
0.007
28
09
Continuous overlap of ≥2 different benzodiazepines for ≥90 d, when 1 is for diazepam
0.003
0.004
17
10
Overlap of ≥2 different sustained-release or long-acting opioids for ≥90 consecutive days
0.001
0.001
5
avoid errors, to minimize the damage caused by errors that occur, and to analyze the patterns of errors and discover ways to prevent them. Despite technological advances and the wealth of strategic knowledge within administrative health claims databases, only 17 states operate electronic prescription monitoring programs, which vary in their goals, structure, and oversight by the health profession.21-26 Presently, few health plans analyze the data to identify potential misuse of controlled substances. Access to this aggregate information on patients is not readily provided to physicians, restricting their ability to provide quality care. In response to this need, we developed a software program that identifies patients with potential prescription mismanagement or abuse and diversion issues. This article lays out a road map for a system to complement state programs, where they exist, and provide a stand-alone tool for physicians in other states. The system detects controlled substance patterns of utilization requiring evaluation (CS-PURE) that suggest need for further evaluation. The CS-PURE criteria, 10 of which are presented herein (Table 1), were developed and validated by experts primarily from the medical
784
profession but also from the pharmaceutical diversion investigation community, based on a combination of evidence and professional consensus.
METHODS Data The data for this study were drawn from health insurance claims databases collected during 2000. These databases were used to test the development and application of CS-PURE criteria. The database used to develop the original CS-PURE prototype consisted of a compilation of claims from different self-insured employers and third-party administrators from nearly all of the 50 states. It contains a mix of health plan types, including indemnity fee-for-service plans, preferred provider organizations, independent practice associations, and health maintenance organizations. Nearly 7 million covered lives are included in the single year of data, drawn from the workforce younger than 65 years and their covered dependents. Database 1 is from one of the nation’s largest managed care plans. Data were used from 5 distinct markets in which independent practice association and preferred provider organi-
THE AMERICAN JOURNAL OF MANAGED CARE
NOVEMBER 2004
Monitoring Controlled Substance Use zation products are offered, 3 in southern states and 2 in midwestern states. Together, these markets contained almost 3 million covered lives from the population younger than 65 not covered by Medicaid. Database 2, generated from a health plan that offered preferred provider organization and health maintenance organization products in the eastern United States, contained records for almost 800 000 covered lives younger than 65. This health plan is representative of most moderateto-large insurers in the United States in terms of demographics and structure. Development of Patterns of Controlled Substance Misuse An 11-member multidisciplinary expert panel, consisting of 2 addictionists, 3 pain physicians, 2 psychologists, 1 psychiatrist, and 3 pain management nurses, convened in December 2001. Their goal was to define CS-PURE criteria that could be applied to claims databases to identify possible abuse or diversion of controlled substances by patients or mismanagement by prescribers. The CS-PURE are not conclusive of inappropriate use; rather, they aim at improving patient safety and outcomes by alerting prescribers and insurers of potential problems so that further evaluation can be conducted by them. The expert panel reached consensus on 38 prototypical CS-PURE for evaluation and did not exclude particular patient groups in general, with 2 exceptions (≥3 prescriptions for injectable opioid for patients without a cancer diagnosis during a year, and any patient with a prescription for benzodiazepine or opioid with a prior substance abuse diagnosis). Some CS-PURE criteria were based on similar patterns, but reflected variations in specific medications used and changes in the duration of consecutive or overlapping days of medication use, for example, continuous overlap of 2 or more benzodiazepines for at least 30, 60, or 90 days. Operationalization of CS-PURE Computer programs based on the expert panel’s original 38 CS-PURE were developed using SAS, version 8.2, to operationalize and apply CS-PURE to the initially developed database. Detailed use profiles were produced for the patients identified by each of the prototypical CS-PURE. These profiles were reviewed and assessed for the accuracy of the computer coding by a project team comprising pharmacists, computer programmers, and health services researchers. At the conclusion of this process, the original 38 CS-PURE were reduced to 34 CS-PURE (detailed in a list available from the author). This change reflected the deletion of 4 of the original CS-PURE criteria because they identified a low number of patients.
VOL. 10, NO. 11
Assessing the Generalizability and Validity of CS-PURE The CS-PURE specification was assessed in terms of generalizability and validity. The CS-PURE were considered to be generalizable if they could be implemented using data from different insurers and managed care organizations and if they identified approximately the same percentage of patients in each plan, indicating the presence of general or similar trends. Alternatively, there could be a plausible explanation for dramatic variation in prevalence rates across plans, such as differences in state laws regarding prescribing practices or the presence of prescription monitoring programs. The CS-PURE criteria were considered to have a high degree of validity if they identified the target population with a low rate of false-positive results. The methods for assessing these dimensions are described herein. Generalizability Assessment. Estimation of whether the data requirements of CS-PURE were too stringent to be widely applicable across plans occurred naturally during the refinement of the profiles, as already described. Once CS-PURE were developed in such a way that the required data elements would be readily available from most insurers, they were applied to data from the 5 database 1 markets and database 2 to compute period prevalence rates. Several multivariate tests of generalizability across the health plans were also completed. Logistic and linear probability regression analyses were used to determine whether the health plans were systematically different in their probability to predict patients’ CSPURE after controlling for general health status using the ambulatory diagnostic group system developed by Johns Hopkins University.27 Validation Assessment. To validate the specification of CS-PURE, a second expert panel meeting was convened in December 2002. The 10-member panel, 7 of whom were members of the original panel in December 2001, consisted of 6 physicians specializing in pain medicine or addiction, a pain management nurse, and 1 current and 2 former diversion investigators. In preparation for this meeting, 180 patient profiles of medical and pharmacy claims history were generated by applying the initial CS-PURE to database 2. This database was chosen because it provided multiregion variation and it was likely to be most similar to an average managed care plan database. A given patient could be flagged by 1 or multiple CS-PURE. Using these results, a random selection of proportionately sampled patients was generated based on the period prevalence of the 34 CSPURE. Each profile represented the medical and pharmacy claims history for a given insured person, displayed in chronological order of service date or pre-
THE AMERICAN JOURNAL OF MANAGED CARE
785
DRUG SAFETY scription fill date. In addition, minimal information about the patient (eg, age and sex), as well as each patient’s total medical care expenditure, was presented. The profiles flagged all CS-PURE found for a given patient for 2000. Panelists were divided into 3 teams of 3 or 4 members each, with each team containing at least 1 pain physician and a diversion investigator. As each profile was reviewed, panelists discussed the available information. Each expert was then asked to individually score each profile along 3 dimensions: (1) whether the case of prescription medication use appeared to be outside the scope of accepted practice, (2) whether further evaluation by the health plan would be recommended, and (3) when an intervention was judged to be needed, whether the intervention should be with the patient, the prescriber, or both. Following the meeting, scores of individual experts were evaluated, and agreement scores were computed for the profiles within CS-PURE. The scores were tabulated to construct the share of agreement in the responses of the 3 questions and provide an assessment of validity among the expert panel.
positively associated with the probability of identifying a misuse case. The prevalence results are displayed in Table 1. Five of the 6 health plans’ prevalence data were combined to represent database 1 plans. The other plan was database 2. Prevalence estimates between the 2 insurers were similar. Notable exceptions were CSPURE06, in which meperidine hydrochloride prescribing was nearly 2 times greater in database 2, and CS-PURE05, in which the number of prescriptions for 4 g/d or more of acetaminophen was roughly 2 times greater in database 1. To appreciate the size of the populations identified as an estimate of potential workload for the prescribers and administrators of the plans, the number of patients who would be identified in a 500 000-member health plan was calculated. In almost all cases within the top 10 population, the number of patients identified by each of the CS-PURE criteria was fewer than 1000. In some cases, the estimated target population was fewer than 20 patients per 500 000 covered lives. Although these estimates are small, they are consistent across the different health plans and are found in more than 1 type of health insurer or region of the country.
CS-PURE Refinement Using the expert panel results, a logistic regression analysis was undertaken for 2 reasons: (1) to examine the validity of the ratings for each of the CS-PURE criteria and (2) to order CS-PURE by their ability to identify cases of interest. The dependent variable in the regression was the rating assigned by each expert reviewer for each profile that he or she reviewed. These were regressed on 34 dichotomous variables, 1 for each of the CS-PURE criteria. No omitted variable was necessary because profiles could, and often did, trigger multiple CS-PURE. This generated odds ratios and tests of statistical significance for each of the 34 CS-PURE. We then culled the list of 34 CS-PURE, concentrating on only the top 10 CS-PURE. The choice of 10 as the number of CS-PURE to examine was somewhat arbitrary, relying on natural breaks in the analysis and on our estimate of the resources that would be necessary to implement them. The selection of the 10 CS-PURE was based on the union of statistically significant positive odds ratios from the logistic regression analysis results and the highest expert agreement scores that intervention was warranted.
CS-PURE Validity Table 2 presents the results of the expert panel’s assessment of the validity of CS-PURE. The overall panel agreement that a case of potential misuse was correctly identified is summarized, as well as the clinical and law enforcement representatives’ assessments of specificity. The overall agreement of correct identification of a case was at least 50% for 9 of the top 10 CSPURE. In CS-PURE01, in which patients had 6 or more different prescribers for the same controlled substance, the overall percentage of patients the experts agreed on as being valid identifications was 55%. In almost all cases, law enforcement and medical professionals were in agreement on the share of patients to be identified. The expert panel also largely agreed that a patient correctly identified as a potential misuse candidate should receive further evaluation and that some form of intervention should be directed at the prescribing physician or the patient (in some cases, both). For CSPURE05 and CS-PURE06, the expert panel had 100% agreement that all of the patients sampled were correctly specified and should be evaluated, with further action possibly warranted. Both of these CS-PURE are clear indicators of inappropriate prescribing, based on current guidelines. The maximum safe daily dose of acetaminophen has long been established at 4 g/d, and guidelines from professional associations warn that meperidine should not be used for more than 48 hours for acute pain and should not be prescribed for persis-
RESULTS CS-PURE Prevalence Six health plans were used to generate the period prevalence estimates of the 10 CS-PURE criteria most
786
THE AMERICAN JOURNAL OF MANAGED CARE
NOVEMBER 2004
Monitoring Controlled Substance Use
Table 2. Validation of Top 10 Controlled Substance Patterns of Utilization Requiring Evaluation (CS-PURE) Most Positively Associated With the Probability of Identifying a Misuse Case Experts Agree on Misuse, %
Experts Agree on Evaluation, %
Experts Agree on Intervention, %
Pattern of Controlled Substance Misuse
No. of Cases Reviewed
Overall
Clinical
Legal
01
Multiple prescribers (≥6 prescribers for same drug)
86
55
59
48
60
59
02
Multiple pharmacies (≥4 different pharmacies for same drug)
58
59
64
51
64
64
03
Chronic use of carisoprodol (≥4 prescriptions in 6 mo)
19
64
68
58
68
71
04
Continuous overlap of ≥2 different benzodiazepines for ≥30 d, when 1 is for alprazolam
30
56
58
50
56
55
05
Estimated ≥4 g/d of acetaminophen
3
100
100
100
100
100
06
≥2 Prescriptions for meperidine hydrochloride with >2-day supply
2
100
100
100
100
100
07
Chronic use of butorphanol tartrate (≥4 prescriptions in 6 mo)
3
56
50
67
100
100
08
Continuous overlap of ≥2 different benzodiazepines for ≥90 d, when 1 is for clonazepam
12
63
67
54
65
63
09
Continuous overlap of ≥2 different benzodiazepines for ≥90 d, when 1 is for diazepam
10
63
65
60
60
60
10
Overlap of ≥2 different sustained-release or long-acting opioids for ≥90 consecutive days
4
63
67
50
69
69
CS-PURE
tent pain. This is because of the formation of an active metabolite called normeperidine, which is a central nervous system excitotoxin that produces anxiety, tremors, myoclonus, and generalized seizures when it accumulates with repetitive dosing.4 CS-PURE Generalizability To assess generalizability, we first computed period prevalence for each of the CS-PURE criteria. Next, using a logistic regression model, we regressed a binary response variable, which corresponded to whether patients did or did not have CS-PURE, on dummy variables for each plan from database 1, with the single plan from database 2 serving as the reference category and an ambulatory care group score to control for case mix.27 The overall sample size was 221 836 patients having at least 1 of the CS-PURE for services received in 2000. Pseudo r2 estimates ranged from 0.01 to 0.45. Statistically significant health plan–specific effects on prevalence were found. All of the models had at least 1 health plan dummy variable that was significant, sug-
VOL. 10, NO. 11
gesting geographic variation in the distribution of CSPURE results. No single health plan consistently had the largest positive or negative effect on the probability of CS-PURE prevalence, indicating that, while the health plan populations were different, they were not systematically biased in a particular direction or degree in the magnitude of their effect.
DISCUSSION We developed a claims-based, computerized method to identify (1) patients who may be misusing controlled substances and (2) prescribers who may be providing pharmacologic management that warrants evaluation. Using claims data to improve quality of care is a specialized art requiring collaboration between clinical experts, programmers, systems analysts, and health services researchers to correctly interpret and identify CS-PURE from items found in a typical database. The major components of claims data are medical and pharmacy
THE AMERICAN JOURNAL OF MANAGED CARE
787
DRUG SAFETY claims, member eligibility, and provider data.28,29 This usually includes the patient’s diagnoses, tests and procedures performed, prescriptions filled (including the specific drug, quantity, and number of days supplied), duration and level of hospitalization with dates and costs, and a unique identifier of the provider who rendered the service. The CS-PURE can be applied using these data elements, which are commonly available to most health plans, and are written in SAS programming language, which is also common to most plans (SAS programming is available on CD-ROM from the author). Practicing and specialty-specific physicians from different parts of the country, together with pharmaceutical diversion investigators, were involved in all aspects of the development of our claims-based CS-PURE. This type of tool can be implemented with little additional expenditure by insurers and is unobtrusive, with low administrative burden to physicians because the data have already been collected. The top 10 CS-PURE criteria identified patients with a high probability of controlled substance misuse or mismanagement so that resources can be targeted and practitioners are not overwhelmed with data, as the projected number of patients identified in a 500 000-member health plan ranged from 5 to 1116. These numbers are manageable for possible intervention, but also show that the prevalence of potential problems with controlled substances cannot be ignored. The lowest prevalences were associated with CS-PURE09 and 10, which were most restrictive in terms of overlap for 90 or more consecutive days. None of the 10 CS-PURE included a threshold for the number of pills or dosage of a controlled substance, because the clinical experts were emphatic that there is no such thing as underprescribing or overprescribing; rather, prescribing should be categorized as appropriate or inappropriate based on an analysis of individual patient characteristics. The only dosage threshold occurred in CS-PURE05 for 4 g/d or more of acetaminophen, which has long been established as the maximum safe dosage of this analgesic.4 Acetaminophen is included in CS-PURE because it is commercially available in fixed combination with opioid analgesics, which should, but often in practice does not, limit the maximum dosage of these combination products. Carisoprodol was included in CS-PURE because it is frequently abused, although it is not a controlled substance by federal regulations; an active metabolite of carisoprodol is meprobamate, a schedule IV anxiolytic.30 Application of Results The CS-PURE are not direct measures of quality, but are surrogates to direct physicians’ attention to poten-
788
tial problems requiring evaluation. Therefore, analysis of claims data to identify CS-PURE should be followed by a process of communication with physicians to determine if intervention is needed. For example, the medical director of a health plan could generate a timely letter or an e-mail alert to the primary or all prescribers of a patient flagged by CS-PURE so that any problem could be promptly addressed. The cases would then be stratified; some complex patients may require a case manager, while other individuals may be triaged into appropriate treatment, for example, an interdisciplinary pain clinic or substance abuse treatment center. It is important that physicians are supported with mechanisms, such as clinical pathways tied to differential diagnoses, to assist in responding to the data sent to them, for example, to differentiate between people with inadequately treated pain and those seeking drugs for abuse or diversion. Current information on proper pain management should be available to physicians, with targeted education for physicians who have been “scammed” or who may be outdated in their prescribing practices to improve their management of patients. Although physicians have been known to resist notices about how to prescribe or treat based on claims data,31 in this case, physicians are provided with valuable information, some of which they would otherwise have no way of knowing, that can help them not only provide better care but also protect their prescriptive authority. Ongoing Refinement As the prototypical CS-PURE criteria are applied by health plans, analysis of the results with subsequent refinement will increase the usefulness of CS-PURE logic to physicians and medical directors. The medical profession should be involved in evaluating the quality and accuracy of data sources, methods, and results. This also requires the development of reliable outcome indicators to assess the effectiveness of applying CSPURE to claims data in reducing prescription drug abuse and in improving quality of care. Limitations Our study has several limitations. Claims data were designed to support a financial transaction rather than to convey clinical information. Pharmacy claims data represent only filled prescriptions and do not generally include information on prescriptions paid for in cash. In addition, the billing diagnosis code includes many terms that may be associated with pain. Commonly used coding and billing logic allow physicians to arbitrarily cite only 1 of multiple diagnoses.32 Other limitations of claims databases include the omission of some nonre-
THE AMERICAN JOURNAL OF MANAGED CARE
NOVEMBER 2004
Monitoring Controlled Substance Use imbursable services, inconsistent or inaccurate coding, the possibility of corrupted data, and the absence of codes that capture disease or symptom severity.28 Underreporting due to unmet deductibles can also lead to incomplete data.29 The data requirements to apply CS-PURE are stringent. Clear identification of unique prescribing physician and pharmacy identifiers is required. Many times, a group practice may be represented by a single prescribing physician identifier, and this may damage the credibility of CS-PURE. New standards on health insurance data, imposed by the Health Insurance Portability and Accountability Act, will guarantee the future specification of a unique prescribing identifier. An additional limitation is that the presence of a data field does not guarantee its quality. Many CSPURE require specific information on dosage from pharmacy claims, but health insurers often do not collect this information. Without a dosage listed on the claim record, CS-PURE must rely on the National Drug Code dosage differences. To explore new CS-PURE specification, accurate dosage information will be essential. Despite these limitations, claims data are readily accessible to plans and constitute the most detailed and uniform body of information available.
stances. They should therefore desire that insurers provide them with useful analyses of available information that will assist them in their practice. However, technology can ignite fears about accuracy, liability, privacy, and security. Physicians have reason to be cautious, which is why physician leadership is critical in developing and implementing such systems and ensuring that safeguards are in place. Acknowledgments We gratefully acknowledge the members of the 2 expert panels: Susan Derby, NP; Daniel Doleys, PhD; Douglas Gourlay, MD; Steven Hahn, MD; Howard Heit, MD; Madelyn Kitaj, MD; Eliot Krames, MD; Bridget McDonough, RN; Fred Sheftel, MD; Richard Stieg, MD; Julie Waight, FNP; J. David Haddox, DDS, MD; Dale Ferranto, MS; Landon S. Gibbs, BA; and Scott Gunn.
REFERENCES 1. Hoffmann DE, Tarzian AJ. Achieving the right balance in oversight of physician opioid prescribing for pain: the role of state medical boards. J Law Med Ethics. 2003;31:21-40. 2. American Pain Society Web site. Promoting pain relief and preventing abuse of pain medications: a critical balancing act: a joint statement from 21 health organizations and the Drug Enforcement Administration. 2001. Available at: http://www.ampainsoc.org/advocacy/promoting.htm. Accessed Febuary 20, 2002. 3. Zacny J, Bigelow G, Compton P, Foley K, Iguchi M, Sannerud C. College on Problems of Drug Dependence taskforce on prescription opioid non-medical use and abuse: position statement. Drug Alcohol Depend. 2003;69:215-232. 4. American Pain Society. Principles of Analgesic Use in the Treatment of Acute Pain and Cancer Pain. 4th ed. Glenview, Ill: American Pain Society; 1999.
CONCLUSION We have demonstrated the development of an empirically validated group of queries that draws on pattern recognition within existing claims databases from different health plans to detect prescribing and medical utilization patterns of controlled substances that may represent improper use. This has potential use as a valuable tool to assist physicians in managing their patients who require treatment with controlled substances. The strategies outlined in this report cannot be implemented by an individual physician or the medical profession as a whole, but are an initiative that requires resources that are only available to insurers, that is, claims data. The development of such a systems approach is an opportunity for insurers to assist physicians to address the possibility of inappropriate use of controlled substances. A partnership between plans and prescribers is needed to systematically provide physicians with timely access to information that is vital to achieve their ultimate goal of improving clinical care and outcomes. The medical profession has a strong incentive to take a leadership role in developing such programs. Physicians are, first and foremost, responsible for the quality of care provided to their patients. In addition, by federal and state law, they are responsible for their prescribing of controlled sub-
VOL. 10, NO. 11
5. Jacox A, Carr DN, Payne R, et al. Management of Cancer Pain: Clinical Practice Guideline No. 9. Rockville, Md: Agency for Health Care Policy and Research, US Dept of Health and Human Services, Public Health Service; March 1994. AHCPR publication 94-0592. 6. American Pain Society Web site. The use of opioids for the treatment of chronic pain: a consensus statement from American Academy of Pain Medicine and American Pain Society. 1997. Available at: http://www.ampainsoc.org/advocacy/opioids.htm. Accessed February 20, 2002. 7. American Pain Society. Guideline for the Management of Pain in Osteoarthritis, Rheumatoid Arthritis and Juvenile Chronic Arthritis: Clinical Practice Guideline No. 2. Glenview, Ill: American Pain Society; 2002. 8. American Geriatrics Society Panel on Persistent Pain in Older Persons. The management of persistent pain in older persons. J Am Geriatr Soc. 2002;50(suppl 6):S205-S224. 9. American Medical Association Web site. Report 11 of the Council on Scientific Affairs (A-99): use of opioids in chronic noncancer pain. 1999. Available at: http://www.ama-assn.org/ama/pub/article/2036-2341.html. Accessed March 1, 2002. 10. Dickinson BD, Altman RD, Nielsen NH, Williams MA. Use of opioids to treat chronic, noncancer pain. West J Med. 2000;172:107-115. 11. American Pain Society. Guideline for the Management of Acute and Chronic Pain in Sickle-Cell Disease: Clinical Practice Guideline No. 1. Glenview, Ill: American Pain Society; 1999. 12. Gilson AM, Joranson DE, Maurer MA. Improving state medical board policies: influence of a model. J Law Med Ethics. 2003;31:119-129. 13. Federation of State Medical Boards Web of the United States, Inc. Model Policy for the Use of Controlled Substances for the Treatment of Pain. May 2004. Available at: http://www.fsmb.org. Accessed October 1, 2004. 14. Dahl JL, Bennett ME, Bromley MD, Joranson DE. Success of the state pain initiatives: moving pain management forward. Cancer Pract. 2002;10(suppl 1):S9-S13. 15. Jacob E, American Pain Society. Pain management in sickle cell disease. Pain Manag Nurs. 2001;2:121-131. 16. Katz WA. Musculoskeletal pain and its socioeconomic implications. Clin Rheumatol. 2002;21(suppl 1):S2-S4. 17. Anderson KO, Richman SP, Hurley J, et al. Cancer pain management among underserved minority outpatients: perceived needs and barriers to optimal control. Cancer. 2002;94:2295-2304. 18. Robeznieks A. Pain specialists fear chilling effect after arrest. October 27, 2003. Available at: http://www.ama-assn.org/amednews/2003/ 10/27/prsc1027.htm. Accessed February 20, 2004. 19. Savage SR. Assessment for addiction in pain-treatment settings. Clin J Pain. 2002;18(suppl 4):S28-S38.
THE AMERICAN JOURNAL OF MANAGED CARE
789
DRUG SAFETY 20. Kohn LT, Corrigan JM, Donaldson MS, Committee on Quality of Health Care in America, Institute of Medicine, eds. To Err Is Human: Building a Safer Health System. Washington, DC: National Academy Press; 1999. 21. Ziegler SJ, Lovrich NP Jr. Pain relief, prescription drugs, and prosecution: a four-state survey of chief prosecutors. J Law Med Ethics. 2003;31:75-100. 22. Joranson DE, Carrow GM, Ryan KM, et al. Pain management and prescription monitoring. J Pain Symptom Manage. 2002;23:231-238. 23. Forgione DA, Neuenschwander P, Vermeer TE. Diversion of prescription drugs to the black market: what the states are doing to curb the tide. J Health Care Finance. 2001;27:65-78. 24. US General Accounting Office Report to the Subcommittee on Oversight and Investigations, Committee on Energy and Commerce, House of Representatives. Prescription Drugs: State Monitoring Programs Provide Useful Tool to Reduce Diversion. Washington, DC: US General Accounting Office; May 2002. Publication GAO-02-634. 25. US Department of Justice, Drug Enforcement Administration, Diversion Control Program. Diversion and Abuse of Prescription Drugs: A Closer Look at State Prescription Monitoring Programs. Alexandria, Va: Drug Enforcement Administration; April 2000. 26. Brushwood DB. Maximizing the value of electronic prescription monitoring programs. J Law Med Ethics. 2003;31:41-54.
790
27. Weiner JP, Starfield BH, Steinwachs DM, Mumford LM. Development and application of a population-oriented measure of ambulatory care case-mix. Med Care. 1991;29:452-472. 28. Armstrong EP, Manuchehri F. Ambulatory care databases for managed care organizations. Am J Health Syst Pharm. 1997;54:1973-1983. 29. Weiner JP, Powe NR, Steinwachs DM, Dent G. Applying insurance claims data to assess quality of care: a comparison of potential indicators. QRB Qual Rev Bull. 1990;16:424-438. 30. Drug Abuse Warning Network Web site. Emergency department trends from DAWN: final estimates 1994 to 2001. August 2002. DAWN Series, No. D-21. DHHS publication SMA-02-3635. Available at: http://dawninfo.samhsa.gov/pubs_94_02/edpubs/2001final/default.asp. Accessed October 10, 2003. 31. Hofer TP, Hayward RA, Greenfield S, Wagner EH, Kaplan SH, Manning WG. The unreliability of individual physician “report cards” for assessing the costs and quality of care of a chronic disease. JAMA. 1999;281:2098-2105. 32. Deyo RA, Taylor VM, Diehr P, et al. Analysis of automated administrative and survey databases to study patterns and outcomes of care. Spine. 1994;19(suppl):2083S-2091S.
THE AMERICAN JOURNAL OF MANAGED CARE
NOVEMBER 2004
