Impact of a Managed Controlled-Opioid Prescription Monitoring Program on Care Coordination

,
The American Journal of Managed Care, September 2012, Volume 18, Issue 9

Managed care organizations have the opportunity to identify potential opioid misuse and implement care coordination interventions, which can enhance safety and streamline patient pain management.

Introduction:

Inappropriate opioid medication utilization has grown tremendously in recent years. Managed care organizations have the opportunity to identify potential opioid misuse and implement care coordination interventions.

Methods:

This randomized controlled study evaluated the impact of providing actionable information to prescribers of members who received opioid prescriptions from 3 or more prescribers at 3 or more pharmacies in a 3-month identification period. Impact was assessed through change in number of prescribers, dispensing pharmacies,

and filled opioid prescriptions over a 1-year period following identification.

Results:

Members randomly assigned to the intervention group demonstrated greater reductions in the number of prescribers (23.98%), dispensing pharmacies (16.28%), and filled opioid prescriptions (15.25%) over the 1-year period. Regression analyses identified group assignment and the number of opioid prescribers in the 3-month identification period as statistically significant predictors of reductions in the number of prescribers, pharmacies, and filled prescriptions.

Conclusions:

This intervention provided action-able information to prescribers regarding member opioid utilization, in addition to available managed care resources. It resulted in significantly greater reductions in the number of prescribers, pharmacies, and prescriptions compared with a general information letter notifying prescribers of available managed care resources. Implementation of this intervention has the potential to enhance coordination of care for members potentially at risk for poor health outcomes.

(Am J Manag Care. 2012;18(9):516-524)Managed care organizations have the opportunity to identify potential opioid misuse and implement care coordination interventions.

  • The impact of the implemented program was assessed through analysis of change in the number of prescribing physicians, dispensing pharmacies, and opioid prescriptions filled over a 1-year period following identification.

  • This study utilized a shorter identification period than previous programs implemented by other organizations.

  • The shorter time frame allows the health plan to provide prescribers with more timely information, potentially allowing them to intervene sooner to enhance safety and coordination of care.

The overwhelming majority of global opioid prescription medications are consumed by residents of the United States.1 A direct correlation exists between the increase in prescription opioid utilization over the last decade and the growing prevalence of opioid misuse.2 Coordinating efforts to improve management of pain, the most common chief complaint encountered by physicians, has been acknowledged as a national priority due to the considerable increase in prescription opioid misuse.3,4

Uncoordinated pain management leads to poor health outcomes for patients5 and increases societal cost burden.6 Opioid-related admissions to substance abuse treatment centers increased by 400% between 1996 and 2006.7,8 Emergency department visits associated with opioid use increased by over 100% between 2004 and 2008.9 Recent measures taken by the legal and medical communities are a clear response to these dramatic upward trends.6,10,11

Approximately 35 US states, in addition to several local health organizations, have implemented Prescription Monitoring Programs (PMPs).12 Many state PMPs collect controlled substance prescription claims data from dispensing pharmacies, usually on a monthly basis, and make that information available to prescribers to facilitate informed treatment decisions.13-17 These prescribers are alerted when prescription claims for a given member meet predetermined inclusion criteria, usually based on prescription claim patterns associated with doctor and/ or pharmacy shopping. These patterns can be indicators of potential misuse of prescription medication.

A standardized methodology for identifying opioid misuse based on prescription claims information has yet to be established.18 The Massachusetts PMP defines questionable controlled substance dispensing as 5 or more prescribers and pharmacies during a 12-month time frame.19 The Upper Peninsula Health Plan (UPHP), a Michigan Medicaid health maintenance organization (HMO), intervenes when members have received 3 or more prescriptions from 2 or more prescribers over a 6-month period.20 PMPs typically disseminate prescription claims information to controlled substance prescribers, as part of the intervention, via mail or a secure website. Pharmacy claims data provided to prescribers may include information related to the dispensed controlled substances, the prescribers, the member, and the dispensing pharmacies. Challenges exist with each type of notification. When information is mailed, it is difficult to determine whether the prescriber received the communication and whether the information was integrated into clinical practice. When information is available via a secure website, prescribers must access the site, an added step which often does not occur. A New York—based managed care organization (NYMCO) developed a Managed Controlled-Opioid Prescription Monitoring Program to identify and address uncoordinated treatment of pain with opioid analgesic medications. The clinical pharmacy department used its internal database to analyze adjudicated schedule II-IV opioid prescription claims quarterly between the 2008 fourth quarter and the 2010 first quarter. Approximately 980,000 members were covered under the included NYMCO line of business during this time frame and a total of 79,754 members obtained at least 1 opioid prescription using the pharmacy benefit. Members who obtained controlled opioid medications from 3 or more prescribers and 3 or more pharmacies within a 3-month time period were identified through this database. Prescription claims data was the driving force behind this initiative which integrated the skills of clinical pharmacy, behavioral health, and medical management specialists at the NYMCO to deliver comprehensive information/service resources to facilitate care coordination when a member received opioid prescriptions from multiple prescribers. The NYMCO hypothesized that use of these resources would streamline pain management and enhance care coordination and medication safety for identified members by reducing the number of prescriptions filled, the number of opioid-dispensing pharmacies, and the number of opioid prescribers utilized by the member.

METHODSEligibility for Randomization

Eligible NYMCO members were identified using adjudicated pharmacy claims data. Members had 1 of the following types of insurance coverage: commercial (HM), Medicare Advantage Part D Plan (MR), or Medicare Part D Prescription Drug Plan (MP). HM and MR policies included both medical and pharmacy coverage with the health plan. MP policies included only pharmacy benefits. Members with MP policies typically had medical benefits, but these benefits were managed by another health plan.

On a quarterly basis between first quarter 2009 and first quarter 2010, members filling controlled opioid medications prescribed by 3 or more prescribers and filled at 3 or more pharmacies in a 3-month period were identified. A quarterly average of 195 members met the aforementioned criteria. First Databank therapeutic class codes (GC3) were used internally to identify prescription claims for opioid pain medicine (H3A and H3U). Reports were executed 1 week after the end of each quarter.

Upon identification, members were randomly assigned to either the intervention group (IG) or the comparison group (CG). Members identified in multiple quarterly reports remained in the group they were initially assigned to. All members included in the 12-month outcomes analysis maintained pharmacy insurance coverage throughout the study period. Prescribers who wrote opioid prescriptions for these identified members included healthcare practitioners, contracted with the NYMCO, given authorization by the Drug Enforcement Agency to prescribe controlled substances (ie, physicians, physician assistants, nurse practitioners, dentists, etc).

Members in both the IG and CG had access to the NYMCO’s pharmacy services, including member services, the ability to fill prescriptions at a vast array of network pharmacies, access to all medications covered under the member’s pharmacy benefits, drug utilization evaluation, and safety alerts for a given prescription relayed as a result of processed claims submitted to the NYMCO by the dispensing pharmacy. Prescriptions paid for in full by the patient were not transmitted to the NYMCO for reimbursement processing and could not be accounted for because these claims were not available within the database used for this analysis.

Intervention Group

For each member in the IG, all prescribers who wrote prescriptions for controlled opioid medications during the 3-month identification period were isolated. Each prescriber was mailed a letter along with a clinical medication report (CMR). The letter informed the prescriber that the NYMCO member under his or her care had received controlled opioid medications from multiple prescribers and filled the prescriptions at multiple pharmacies. The letter discussed the importance of collaboration between prescribers of controlled opioid medications and the added safety benefits of limiting the number of pharmacies dispensing a member’s controlled opioid medications. The letter also informed the prescriber of various NYMCO resources, including contact information for the following: the clinical pharmacy department, which provided an opportunity to collaborate with a NYMCO clinical pharmacist; the behavioral health department, to arrange for mental health or substance abuse services; and the name and phone number of a certified alcohol and substance abuse counseling (CASAC) specialist, in case prescription medication abuse or dependence was suspected.

The CMR, included with the IG letter, provided each prescriber with NYMCO data for each controlled opioid prescription filled during the 3-month identification period. Data included medication name, dose, days supply, date filled, prescriber name, and phone number. The prescriber phone number was included to promote ease of collaboration between prescribers. It typically took at least 1 month from the end of the quarter for a prescriber to receive the letter and CMR. When a member was identified in multiple quarterly reports, a new letter and CMR with the corresponding opioid prescription utilization data was generated and mailed to each prescriber.

The CASAC specialist proactively reached out to each prescriber who prescribed controlled opioid medications to members identified in multiple quarters to ensure he or she received and reviewed the CMR and to discuss the case and offer any appropriate resources or assistance to the prescriber. The specialist was also available to members to discuss treatment options and provide brief screening interventions over the telephone. If a member did not have medical benefits with the NYMCO, the specialist was able to provide other resources available through the member’s prescription benefit with the NYMCO. A total of 330 phone calls were made by the CASAC specialist to these identified prescribers during the study period. Care coordination and resource dissemination was the focal point of these conversations. An average of 5 hours, out of the 35-hour work week, was spent on weekly prescriber phone calls during the months after identification. When combined with member identification and mail preparation, all activities associated with execution of the program required less than 0.5 full time equivalents (FTEs).

Comparison Group

The prescribers providing controlled opioid medications to members in the CG received a letter, but did not receive a CMR. This general information (GI) letter was not member specific; it did not indicate which members under the prescribers’ care received controlled opioid prescriptions from 3 or more prescribers and filled the prescriptions at 3 or more pharmacies in the 3-month identification period. The letter simply described rising national trends in prescription medication abuse and informed the prescriber of the same NYMCO resources provided in the letter to prescribers of members in the CG.

An average of 540 letters were mailed out quarterly to the prescribers who wrote prescriptions for all identified members (IG and CG combined) during the previous 3-month period. The pharmacist performing the pharmacy claims data analysis also prepared the letters for mailing. An average of 5 hours out of the 35-hour work week, each quarter, was spent on mail preparation. This depended on availability of staff to assist.

Measures

The outcome measures for this study assessed the following changes in controlled opioid medication utilization patterns over a 12-month period:

Change in Number of Prescribers. The number of controlled opioid medication prescribers in the 12th month following identification minus the number of prescribers providing prescriptions for controlled opioid medication in the first month following identification.

Change in Number of Dispensing Pharmacies. The number of pharmacies filling prescriptions for controlled opioid medication in the 12th month following identification minus the number of pharmacies filling prescriptions for controlled opioid medication in the first month following identification.

Change in Number of Controlled Opioid Prescriptions. The total number of prescriptions filled for controlled opioid medications in the 12th month following identification minus the total number of prescriptions filled for controlled opioid medications in the first month following identification.

Statistical Analysis

Differences between member groups in terms of age, percentage of male members, type of insurance coverage (HM, MR, or MP), type of policy holder (primary, secondary, or dependent), number of prescribers during the 3-month identification period, and number of pharmacies in the 3-month identification period were examined using t tests for continuous variables and χ2 analyses for categorical variables.

Three forward stepwise linear regression models were created using the following dependent variables: change in number of prescribers, change in number of pharmacies, and change in number of prescriptions. Change in prescribers, pharmacies, and number of prescriptions was defined as the number in the 12th month following member identification minus the number in the 1st month following identification. Dependent variables were normally distributed. Variance inflation factors and tolerance were assessed. No issues of multicollinearity were present. For all models, the independent variables consisted of group assignment (IG or CG), centered age,

gender, line of business (HM, MR, or MP), type of policy holder (primary, secondary, or dependent), number of prescribers in the 3-month identification period, and number of pharmacies in the 3-month identification period. The age variable was centered by subtracting the mean age of 54 from each member’s age, and dummy variables were created to identify line of business and type of policy holder. HM was used as the reference category to compare against MR and MP. Primary policy holder was used as the reference category to compare against secondary and dependent. Statistical significance for all tests was established at α <.05.

RESULTS

A total of 754 NYMCO members met the study eligibility criteria for initial inclusion, accounting for 0.95% (9.5 members per 1000) of the opioid-utilizing population. Of these, 382 were assigned to the IG and 372 were assigned to the CG. A total of 292 (38.72%) members did not maintain pharmacy coverage for the 1-year outcome period and were removed from all outcome analyses. Of these, 140 (36.65%) were IG members and 152 (40.86%) were CG members, resulting in 242 members in the IG and 220 in the CG for the outcome analysis.

Excluded Versus Included Members

Table 1

There were no statistically significant differences between the excluded members who did not maintain coverage and the members included in the outcome analyses on any of the following variables: percentage assigned to the IG, gender, number of prescribing doctors in the 3-month identification period, number of pharmacies in the 3-month identification period, or type of policy holder (primary, secondary, or dependent). However, significant differences in age and insurance type were found and are displayed in .

Intervention Group Versus Comparison Group Members

Table 2

Table 3

eAppendices A

B

There were no statistically significant differences between the 242 IG members and the 220 CG members on any of the following preintervention variables: noncentered age, percentage of male members, type of insurance coverage, and type of policy holder (). There were also no statistically significant differences between the 2 groups during the 3-month identification period when comparing the average number of prescribers and average number of pharmacies dispensing controlled opioid prescriptions (). Prescriber and pharmacy frequencies for each group are provided in and , available at www.ajmc.com.

Effects of the Intervention on Change in Number of Prescribers

Figure 1

Table 4

Compared with CG members, IG members demonstrated a greater reduction in the number of controlled opioid medication prescribers from month 1 to month 12 during the outcome period. On average, the reduction in number of prescribers was 0.24 (23.98%) greater among IG members (Table 3). depicts the average number of prescribers of controlled opioid medication for both groups at 1, 6, 9, and 12 months following identification. The forward stepwise regression model for change in number of prescribers of controlled opioids from the 1st month to the 12th month following identification yielded 2 significant variables: group assignment and number of prescribers in the 3-month identification period. The results for this model are presented in .

Effects of the Intervention on Change in Number of Pharmacies

Figure 2

IG members demonstrated a greater reduction than CG members in the number of pharmacies filling controlled opioid medications over the 1-year outcome period. On average, the reduction in the number of pharmacies was 0.19 (16.28%) greater among IG members (Table 3). depicts the average number of pharmacies filling controlled opioid medications for both groups at 1, 6, 9, and 12 months following identification. The forward stepwise regression model yielded 2 significant variables: group assignment and centered age (Table 4).

Intervention on Change in Number of Controlled Opioid Prescriptions

Figure 3

IG members demonstrated a greater reduction than CG members in the number of controlled opioid prescriptions over the 1-year outcome period. On average, the reduction in opioid prescriptions was 0.26 (15.25%) prescriptions greater among IG members (Table 3). depicts the average number of controlled opioid prescriptions filled for both groups at 1, 6, 9, and 12 months following identification. The forward stepwise regression model yielded 3 significant variables: group assignment, number of opioid prescribers in the 3-month identification period, and centered age (Table 4).

CONCLUSIONS

The intervention implemented in this study was designed to improve the coordination of care for individuals given opioid medications by 3 or more prescribers and who filled the prescriptions at 3 or more pharmacies in a 3-month identification period.

Three regression models assessed the association between the intervention and several demographic and utilization measures on the change in the number of prescribers, pharmacies, and prescriptions as indicated by comparing the change in each measure during the 12th month postidentification with the change in each measure during the 1st month postidentification. It took more than 1 month from the time a member was identified for the letters/reports to be created/mailed by the NYMCO pharmacy staff and received by the prescriber. It was anticipated that it would take up to 12 months before the intervention would yield the desired utilization changes. This time frame was anticipated because of the time involved in generating and disseminating reports to prescribers, who in many cases would wait until the next scheduled visit to address the clinical care coordination issues with the member. For members rotating between a large number of prescribers, the next visit could take several months. It is also possible that once a member discovered that a prescriber was now aware of the other provider prescribing opioids, the member might discontinue seeing that provider and begin seeing a new prescriber. This might increase the number of prescribers the member visits in a short time period. In such a case, it was anticipated that over time the member would discover that all prescribers were aware of the uncoordinated opioid utilization and that they could collaborate to ensure the member received the most clinically appropriate care possible. It was anticipated that in the 12th month, members of prescribers receiving the intervention would demonstrate a significantly greater decrease in the number of prescribers, pharmacies, and opioid prescriptions compared with the first month than did the members of prescribers who did not receive the full intervention.

It is important to note that the beneficial results of the intervention did not occur immediately. Graphs depicting changes in the number of prescribers, pharmacies, and prescriptions at the 1st, 6th, 9th, and 12th month demonstrate that the differences between the IG and CG did not emerge until approximately 12 months after the initial intervention. This suggests that NYMCO commitment over time is necessary to achieve results.

Results of the regression analyses demonstrated that the intervention was associated with a significantly greater reduction, over a 12-month period, in the number of opioid prescribers, pharmacies, and opioid prescriptions compared with members whose prescribers did not receive the intervention. In addition, members with a higher number of preintervention prescribers demonstrated a greater reduction in the number of opioid prescribers and prescriptions. This may be due to the fact that members starting out with the highest number of prescribers and prescriptions have the greatest opportunity for reductions. Finally, age was significantly associated with reductions in opioid prescriptions and pharmacies, with younger members demonstrating greater reductions (Table 4). The regression results support the benefits of providing prescribers with a report of each member’s opioid medication utilization, contact information for all prescribers, and information about accessing available NYMCO resources. This finding has important implications for all managed care organizations.

This study utilized a shorter identification period than previous programs implemented in Massachusetts and Michigan.19,20 The shorter time frame allows the health plan to provide prescribers with more timely information, potentially allowing them to intervene sooner to enhance safety and coordination of care.

While other studies have discussed the use of both mailings and web portals to provide prescribers of opioids with detailed information,12,14-16 the intervention used in this study incorporated proactive outreach to prescribers of members who repeatedly filled opioid prescriptions from 3 or more prescribers at 3 or more pharmacies in the 3-month period of time. This enhancement helped ensure that prescribers of members in the greatest need of care coordination were aware of each member’s pharmacy utilization and the resources available to help both the member and provider.

Although no standardized identification criteria have been established,18 the easily replicable criteria used in this study yielded a reasonably sized cohort for whom actionable steps could be taken by prescribers in a timely manner. In addition, this intervention was well tolerated by both prescribers and members. No member or prescriber complaints were received during the study period.

The 12-month time frame needed to achieve results reflected both the time involved to deliver the intervention and the intricacy of the cases. Identification of members by analysis of prescription claims, in addition to the printing, collating, and mailing of the letters with the CMR to the prescribers, typically took at least 1 month. Time was then needed for the prescriber to receive the information, collaborate with other healthcare professionals, complete a follow-up appointment with the member and, if necessary, create an enhanced treatment plan for a member with possible complex issues and conditions.

Although this study utilized a randomized controlled design, several limitations existed. Identification criteria required a member to receive opioid medications from 3 or more prescribers and filled at 3 or more pharmacies within a relatively short 3-month time frame. These stringent criteria were used to ensure a high likelihood of identifying members with inappropriate opioid medication utilization. However, these criteria likely missed members who may have filled prescriptions at 2 pharmacies but received prescriptions from many prescribers, and vice versa. The effectiveness of the intervention for members with less than 3 prescribers or less than 3 pharmacies is unknown. The study could not account for prescriptions that were not obtained through the NYMCO prescription benefit because data associated with prescriptions paid for in full by the member are not shared with the NYMCO by the dispensing pharmacy. Prescription claims from the identification period could not be used to determine if the member was being treated for an acute or chronic pain condition, thus impact of diagnosis on the results of this study could not be accounted for. A prescriber’s ability to make an informed clinical decision to appropriately treat a clinical condition is enhanced by awareness of the prescription history of a given patient. If continued prescribing is not appropriate and a prescriber elects to cease prescribing medications to manage pain, this study cannot account for out-of-pocket expenses paid for prescriptions. In addition, members’ reaction to being confronted by prescribers may have been to seek opioid medication or treatment from sources outside of the NYMCO healthcare system, which the NYMCO could not account for.

This study also was unable to include Medicaid members, and thus the impact of the intervention on a Medicaid population is unknown. In addition, the study lacked the power to properly assess the benefit of the intervention if the criteria were made more stringent (eg, 4 prescribers and 4 pharmacies) or the outcomes analysis extended to more than 12 months.

A healthcare practitioner’s ability to design the best pain management plan for a given patient is severely hindered by the absence of up-to-date prescription utilization information. Receiving opioid prescriptions from multiple prescribers and pharmacies during short time intervals is evidence of uncoordinated pain management and suggests misuse of opioid medications. Although dependence or addiction to opioid medications cannot be diagnosed by analysis of prescription claims alone, prescription claims may be used as a tool to guide future prescribing and to pinpoint factors associated with the uncoordinated care of a member.

Enhancing prescriber access to member-specific controlled opioid prescription claims information and health plan opioid abuse resources is a relatively low-cost intervention. Implementing this type of intervention has the potential to facilitate informed treatment decisions and improve member safety. As inappropriate use of opioid medications continues to grow nationally, leading to higher costs and increased safety risks for members receiving opioid medication from multiple prescribers and pharmacies, managed care organizations have the opportunity to implement programs to enhance coordination of care and safety for members at highest risk for opioid medication misuse.

Acknowledgments

The authors would like to thank Arthur Naliboff, MS, RPh, Dennis Liotta, MD, MBA, Lenord Reich, PhD, and Neil Meyerkopf, MBA, MHS, for assisting with the administrative implementation and development of this pilot program. The authors would like to thank Juan Arias, CASAC, for assisting with prescriber and patient outreach. The authors would like to thank Chirag Patel and Stephanie Gaston, for assistance with membership data analysis.

Author Affiliations: From Clinical Pharmacy Department (AMG), Provider Group Clinical Management (AK), Emblem Health, New York, NY.

Funding Source: None.

Author Disclosures: The authors (AMG, AK) report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (AMG); acquisition of data (AMG); analysis and interpretation of data (AMG, AK); drafting of the manuscript (AMG, AK); critical revision of the manuscript for important intellectual content (AMG, AK); statistical analysis (AMG, AK); administrative, technical, or logistic support (AMG, AK); and supervision (AMG).

Address correspondence to: Arsenio M. Gonzalez III, PharmD, RPh, 55 Water St, New York, NY 10041. E-mail: Agonzalez@emblemhealth.com.1. Solanki DR, Koyyalagunta D, Shah RV, Silverman SM, Manchikanti L. Monitoring opioid adherence in chronic pain patients: assessment of risk of substance misuse. Pain Physician. 2011;14(2):E119-E131.

2. Boscarino JA, Rukstalis M, Hoffman SN, et al. Risk factors for drug dependence among out-patients on opioid therapy in a large US health-care system. Addiction. 2010;105(10):1776-1782.

3. Brauser D. REMS requires manufacturers to initiate and pay for physician training programs. Medscape News. http://www.medscape.com. Published April 19, 2011. Accessed July 1, 2011.

4. Brauser D. Nation’s prescription drug abuse problem continues to grow. Medscape News. http://www.medscape.com. Published June 30, 2011. Accessed July 1, 2011.

5. Hallinan R, Osborn M, Cohen M, Dobbin M, Wodak A. Increasing the benefits and reducing the harms of prescription opioid analgesics. Drug Alcohol Rev. 2011;30(3):315-323.

6. Birnbaum HG, White AG, Schiller M, Waldman T, Cleveland JM, Roland CL. Societal costs of prescription opioid abuse, dependence, and misuse in the United States. Pain Med. 2011;12(4):657—667.

7. Substance Abuse and Mental Health Services Administration, Office of Applied Studies. The TEDS Report: Substance Abuse Treatment Admissions Involving Abuse of Pain Relievers: 1998 and 2008. Rockville, MD. July 15, 2010.

8. Substance Abuse and Mental Health Services Administration. Results from the 2008 National Survey on Drug Use and Health: National Findings (Office of Applied Studies, NSDUH Series H-36, HHS Publication No. SMA 09-4434). 2009. Rockville, MD.

9. Webster LR, Fine PG. Approaches to improve pain relief while minimizing opioid abuse liability. J Pain. 2010;11(7):602-611.

10. Wang J., Christo PJ. The influence of prescription monitoring programs on chronic pain management. Pain Physician. 2009;12(3):507-515.

11. Spiller H, Lorenz DJ, Bailey EJ, Dart RC. Epidemiological trends in abuse and misuse of prescription opioids. J Addict Dis. 2009;28(2):130-136.

12. US Dept of Justice. Drug Enforcement Administration website. Office of Diversion Control. Questions and Answers: State Prescription Monitoring Programs. http://www.deadiversion.usdoj.gov/faq/rx_monitor.htm. Updated May 2011. Accessed July 1, 2011.

13. Fishman SM, Papazian JS, Gonzalez S, Riches PS, Gilson A. Regulating opioid prescribing through prescription monitoring programs: balancing drug diversion and treatment of pain. Pain Med. 2004;5(3):309-324.

14. Joranson DE, Carrow GM, Ryan KM, et al. Pain management and prescription monitoring. J Pain Prescrip Manag. 2002;23(3):231-238.

15. Katz N, Houle B, Fernandez KC, et al. Update on prescription monitoring in clinical practice: a survey study of prescription monitoring program administrators. Pain Med. 2008;9(5):587-594.

16. Katz NP, Adams EH, Benneyan JC, et al. Foundations of opioid risk management. Clin J Pain. 2007;23(2):103-118.

17. Wilsey BL, Fishman SM, Gilson AM, et al. An analysis of the number of multiple prescribers for opioids utilizing data from the California Prescription Monitoring Program. Pharmacoepidemiol Drug Saf. 2011;20(12):1262-1268.

18. Chou C, Fanciullo G, Fine P, et al. Clinical guidelines for the use of chronic opioid therapy in chronic noncancer pain. J Pain. 2009;10(2):113-130.

19. Katz N, Panas L, Kim M, et al. Usefulness of prescription monitoring programs for surveillance—analysis of Schedule II opioid prescription data in Massachusetts, 1996-2006. Pharmacoepidemiol Drug Saf. 2010;19(2):115-123.

20. Braker LS, Reese AE, Card RO, Van Howe RS. Screening for potential prescription opioid misuse in a michigan medicaid population. Fam Med. 2009;41(10):729-734.