PrescriptionMonitoring ProgramCenter for Excellence, Brandeis University April 10-12, 2012 Walt Disney World Swan Resort
Accepted Learning Objectives:1. Explain the current capabilities and contributions of PMPs.2. Describe ways PMP data can be used to predict patterns of opioid overdoses and how this can be used to target prevention efforts.3. Evaluate the benefits of unsolicited reporting as a means to reduce drug abuse.
Disclosure Statement• All presenters for this session, Thomas Clark, David R. Hopkins and Leonard Young, have disclosed no relevant, real or apparent personal or professional financial relationships.
Massachusetts Department of Public Health Drug Control Program Prescription Monitoring Program An Evaluation of the Impact of Providing Unsolicited “Questionable Activity” Reports to Prescribers Len Young
• Purpose of the MA PMP: To help promote safe prescribing and dispensing To help prevent drug diversion and abuse.• MA PMP collects data on Schedule II-V (e.g., narcotic, stimulant, sedative) prescriptions dispensed by community, clinic and outpatient pharmacies as well as out-of-state mail order pharmacies that deliver to MA.• Over 11 million Schedule II-V prescription records (8.7 million new prescriptions) were reported to MA PMP in CY 2011.
Estimated Number of Individuals per 100,0001 Showing Questionable Activity2 by Fiscal Year in MA 7,411 (0.85%) Individuals 121,238 (5.8%) Prescriptions1 Populationincludes all individuals (identified by customer ID) who received at least one Schedule II opioid prescription in a fiscal year.2 Questionable activity is defined as having received Schedule II opioid prescriptions from a minimum of 4 providers and 4 pharmacies during the reported fiscal year.
WHY SEND UNSOLICITED REPORTS?• Evidence from PMPs (e.g., Nevada1 and Wyoming2 ) suggests that sending unsolicited reports reduces questionable activity (e.g., number of prescribers and pharmacies visited).• The total economic cost of substance abuse has reached $245 billion. Includes treatment and prevention costs, healthcare, losses on job productivity, crime and social welfare.31 PMP Center of Excellence Notes from the Field. Nevada’s Proactive PMP: The Impact of Unsolicited Reports. October 20112 PMP Center of Excellence Notes from the Field. Trends in Wyoming PMP Prescription History Reporting: Evidence for a Decrease in Doctor Shopping? September 2010.3 National Institute on Drug Abuse and the National Institute on Alcohol Abuse and Alcoholism in the US. 8
Unsolicited reports consist of: Overview of initiative Patient prescription history Provider contact information Guidance on use of information Two-part optional prescriber survey Part 1: Complete upon receipt of report Part 2: Three month follow-up
• 336 baseline surveys received as of Jan, 2012• 72% of respondents said Unsolicited Reports (both electronic and hard copy) “very” or “somewhat helpful”• 8% said they were “aware of all or most of other prescribers” in report• 9% said “based on current knowledge, including report, patient appears to have legitimate medical reason for prescriptions from multiple prescribers”Source: MDPH and P. Kreiner et al., Brandeis University
PURPOSE OF ANALYSIS• To date, previous analyses (noted in previous slide) of the impact of unsolicited reports to prescribers have not included a comparison group.• MA PMP evaluated the impact of unsolicited reports on the controlled prescription drug use of individuals who met specified thresholds of questionable activity for whom such reports were sent.• A non-intervention comparison group was included to provide more accurate measures of the impact of unsolicited reports.
METHODOLOGYIdentifying Individuals for Unsolicited ReportsAll individuals considered for the “Unsolicited Report”group met a pre-specified minimum questionableactivity threshold: ─ Having a Schedule II prescription prescribed by 4 or more different prescribers AND ─ Having a Schedule II prescription dispensed at 4 or more different pharmacies ─ During a 6 month reporting period
METHODOLOGY (Continued)Identifying Individuals for Unsolicited Reports • Individuals who meet this threshold are considered to be the at risk population for questionable activity and comprised the pool of eligible people selected for an unsolicited report. • Prescription records of individuals selected for unsolicited reports were carefully reviewed by experienced PMP staff to verify that the individual was exhibiting questionable activity. • Only a small percentage (< 5 percent) of all the eligible at risk individuals in MA were selected for an unsolicited report.
METHODOLOGY (Continued)Identifying Individuals for Unsolicited Reports• Prescribers received the selected individual’s Schedule II prescription history for the most current 1-year time period.• The individuals (i.e. case population) included in the analysis were identified between January and August, 2010.
METHODOLOGY (Continued) Selection of Comparison Group• A comparison group was selected from the same pool of eligible individuals who were selected (i.e., cases) for unsolicited reports.• Any individual who was selected for an unsolicited report was ineligible to be in the comparison group.• A methodology referred to as “propensity scoring” was used to identify suitable matches so that key measures and other characteristics were similar to the distribution of the case group.
*Pre-Intervention data includes all Schedule II prescription records reported to MA PMP for a specified 12 month time period prior to sending a case report for a specified 12 month time period prior to when the comparison individual is selected.
* Excludes individuals where no controlled prescription records were found in the post-intervention period; cases: n = 62, controls: n = 61† Statistically significant at p < 0.05
CONCLUSIONS• Findings suggest that providing unsolicited reports reduces key proxy measures of questionable activity.• All key measures declined more in the case group compared with the comparison group, although the differences were, in most instances, not statistically significant.• There was a statistically significant difference, between the case and control group, in the decline in the number of pharmacies visited from the pre- and post-intervention period.• The results validate other studies (without a comparison group) that showed marked declines in measures of questionable activity for individuals for whom unsolicited reports were sent.
STUDY LIMITATIONS• Difficulty in identifying suitable comparison population.• MA PMP only collected Schedule II Rx during the pre- intervention time frame; therefore pre- and post- intervention comparisons for Schedule III-V are not possible.• Limited patient health information - no way to confirm whether all individuals in the study were engaging in inappropriate use of controlled drugs; also no way to follow up on individuals who did not have any controlled prescription activity during the post-intervention period.• Incomplete prescription records - some controlled drug prescription records may not be included in the analysis; (e.g., individual may intentionally conceal or misrepresent identifying information.
POSSIBLE NEXT STEPS• Expand study – include larger sample size to improve power to detect differences.• Calculate morphine milligram equivalents for pre- and post-intervention cases and controls.• Sub group analyses (e.g., individuals who combine different controlled drug products versus those who primarily use opioid products; individuals who obtain small versus large dosage quantities).• Look at longer time trends - assess whether individuals, whose use of controlled prescription drugs declined markedly in the post-intervention time period, maintained that decline over time.• Why did the comparison group show such large decreases in prescriptions, prescribers, pharmacies, etc?
Acknowledgement• Portions of this project were supported by grants awarded by the U.S. Bureau of Justice Assistance. Points of view or opinions in this presentation are those of the author and do not represent the official position or policies of the United States Department of Justice.