Prescription Drug Monitoring Programs Montana State Fund 13 th Annual Medical Conference November 1, 2013 Peter Kreiner, Ph.D. PDMP Center of Excellence at Brandeis University
Primary non-heroin opiates/synthetics admission rates, by State (per 100,000 population aged 12 and over) 2
Primary non-heroin opiates/synthetics admission rates, by State (per 100,000 population aged 12 and over) 3
Primary non-heroin opiates/synthetics admission rates, by State (per 100,000 population aged 12 and over) 4
Primary non-heroin opiates/synthetics admission rates, by State (per 100,000 population aged 12 and over) 5
Primary non-heroin opiates/synthetics admission rates, by State (per 100,000 population aged 12 and over) 6
Primary non-heroin opiates/synthetics admission rates, by State (per 100,000 population aged 12 and over) 7
Overview What are prescription drug monitoring programs (PDMPs) and how can they help? A quick aside: Why is the evidence base for PDMPs so limited? Selected best and promising practices of PDMPs Using identified PDMP data Using de-identified PDMP data
What Data Fields Does PDMP Data Contain? Patient first and last name, street address, town/city, Zip Code, birth date, gender Prescriber and pharmacy DEA license #, street address, town/city, Zip Code Prescription information: date prescribed, date filled, drug name, drug NDC code, dosage, days supply, source of payment
System Overview Pharmacists Reports Sent Reports Sent Dispensers State PDMP Data Submitted Reports Sent Prescribers *Other groups may also receive reports other than those listed Law Enforcement & Professional Licensing Agencies
60 Number of Prescription Drug Monitoring Programs (PDMPs): Authorizing Legislation Passed Between 1939 and 2012 50 50 PDMPs Number of PDMPs 40 30 20 10 0 Years
Why Is the Evidence Base So Limited? Nearly half of currently operational state PDMPs become so since 2008 (21 out of 46 = 46%) 13 PDMPs (28%) became operational since 2011 Limited years for studying effects Great variation in characteristics across PDMPs Difficulty in separating out PDMP effects
PDMP Characteristics Location of PDMP in state government Health dept., board of pharmacy, single state authority Law enforcement agency Professional licensing board No. of states 38 6 2 Drugs that can be monitored No. of states Only Schedule II drugs 1 Only Schedule II and III drugs Schedule II, III, and IV drugs 2 45 Schedule II V drugs 29 Source: National Alliance for Model State Drug Laws
PDMP Characteristics II Access to law enforcement No. of states Access other than to law enforcement No. of states For probable cause, search warrant, subpoena, other judicial process Pursuant to active investigation On request from law enforcement 17 29 1 To prescribers and dispensers To patient, parent, or guardian To licensing or regulatory boards To Medicare, Medicaid, or state insurance programs 45 35 44 29
PDMP Characteristics III Frequency of pharmacy submission of data to PDMPs No. of states Real-time data submission 1 Weekly data submission 22 Monthly data submission 6 No provision for electronic submission 4
PDMP Characteristics IV Interstate data sharing Share data with other PDMPs Share with users in other states Share with both other PDMPs and authorized users No. of states 19 8 15 Provide unsolicited reports No. of states No reports 7 To prescribers only 2 To law enforcement only 2 To prescribers and pharmacists only To prescribers, pharmacists, law enforcement, and licensing entities 5 20
Nevertheless, Some Suggestive Within-State Evidence
Multiple provider episode rates* for CS II drugs, Quarter 4 of 2011 vs. Quarter 4 of 2012, Florida Rate per 100,000 residents 8 7 6 5 4 3 2 1 0 6.9 4.7 2.6 1.8 1.9 0.8 0.0 0.0 <18 18-34 35-54 55+ Age Group Q4 2011 Q4 2012 *Having CSII rx from 5+ prescribers dispensed at 5+ pharmacies during one quarter. 18
Florida s PDMP Continued The PDMP became operational in September, 2011 The law authorizing the PDMP was accompanied by several other provisions: A requirement for pain clinics to register with the state Increased penalties for operating a pill mill Drug overdose deaths associated with Schedule II opioids declined in 2012 from 2011 Oxycodone: 33%; Oxymorphone: 35%
Selected PDMP Best Practices Unsolicited reporting to providers Interstate data sharing with other PDMPs Weekly or more frequent data submission Use of PDMP data for surveillance and to support prevention (de-identified data)
Numbers of histories 70 60 50 40 30 20 10 66 39 Notes from the Field : Wyoming PDMP Unsolicited Prescription Histories per Month, 10/2008 9/2009 27 32 28 31 40 33 26 18 26 15 Numbers of histories 1000 900 800 700 600 500 400 Solicited Prescription Histories per Month, 10/2008 9/2009 524 459 541 681 682 750 726 685 730 651 773 949 0 300 Source: Wyoming PDMP Source: Wyoming PDMP
Why Unsolicited Reports Are Important MA PDMP survey physicians receiving unsolicited reports: Only 8% of respondents were aware of all or most of other prescribers Only 9% said based on current knowledge, including PDMP report, patient appears to have legitimate medical reason for prescriptions from multiple prescribers Alert prescribers of persons receiving more than 100 mg morphine equivalents of opioids per day 8.9 times higher risk of death than low dose
Massachusetts: Evaluation of Unsolicited Reporting MA initiated unsolicited reporting in 2010 Schedule II only Unsolicited reports sent on a small fraction of patients who met questionable activity threshold We constructed profiles of patients on whom reports were sent and developed a comparison group (on whom reports not sent) based on propensity score matching Intervention and comparison groups matched on age, gender, and # prescriptions, # prescribers, # pharmacies in the 12 months prior to sending of reports Both groups tracked for subsequent 12 months in PDMP data
Massachusetts: Evaluation of unsolicited reporting Preliminary results Case Group (N = 84) Comparison Group (N = 84) Pre Post % Change Pre Post % Change Probability Total # of Schedule II Rx 48.3 24.0 50.3 49.1 30.0 38.9.08 Average # of Prescribers 18.5 8.2 55.7 18.0 9.7 46.1.19 Average # of Pharmacies Average Dosage Units 11.0 5.3 51.8 11.7 7.0 40.2.02 2,309 1,404 39.2 2,428 1,700 30.0.32 Average days Supply 473 272 42.6 475 359 24.4.02
Interstate Data Sharing All PDMPs receive data on prescriptions written in every other state and filled in their state 3 hubs currently enable provider access to patient prescription history data from multiple state PDMPs Interstate sharing varies: Provider requests for data from other states mostly focus on neighboring states More comprehensive data should lead to better clinical decision-making, but no evidence as yet
Weekly or More Frequent Data Submission: Physician Use of PDMP Data OH study of Emergency Department 41% of prescribers who received PMP report altered prescribing for patients receiving multiple simultaneous narcotics prescriptions Of these providers, 63% prescribed no narcotics or fewer 39% prescribed more For non-ed physicians, need for data frequency not as clear
Use of PDMP Data for Surveillance and to Support Prevention: Massachusetts
2005 Opioid-related Health Overdose Problems Rate per 100,000 by Town Rate per 100,000 Quintiles 0 0.01-19.82 19.82-37.5 37.5-56.92 56.92-225.51
2005 Prescriptions Associated with Questionable Activity (Rates per 100,000 Prescriptions) by Pharmacy Town Questionable activity rates 0 1-1095 1096-1897 1898-2882 2883-14184
Massachusetts Geospatial Analysis Do rates of questionable activity predict subsequent changes in rates of opioid overdoses at the community level? Controlling for community socio-demographic variables and for spatial association
Spatial regression: questionable activity rate as a predictor of subsequent change in opioid overdose rate (Data from Massachusetts PDMP in partnership with Brandeis University) Variable Coefficient Probability Constant -.543.185 Opioid OD rate 2001-03 average.519 <.001 Population density 2000.143 <.001 Poverty rate 2000.096.025 Ethnic heterogeneity 2000 -.066.145 Population mobility 2000.034.334 Percent > 65.078.019 Questionable activity rate 2001-03 average.226 <.001 Nonprofit intensity -.030 <.001 Spatial lag (Opioid OD rate 2004-06 average).191.025 Lambda (spatial association error term).034.833 Dependent variable: Opioid OD rate 2004-06 average Pseudo R-squared:.699
Implications Highlights importance of PDMP-based measures for surveillance Questionable activity measure predicts subsequent increases in rates of overdoses Timeliness of PDMP data compared to health outcome data Importance of PDMP-based measures for prevention Identify areas at high risk for increase in opioid overdoses Identify clusters of communities at high risk: targeting cluster for intervention may be more effective Identify low-risk islands amidst high-risk communities: what can be learned from them?
Other Surveillance Applications Examine prescribing rates in different states/regions and over time Broken out by age groups and gender Examine (trends in) measures of risky patient and provider behavior Examine geographic variation and factors associated with this variation
Rate per 1,000 residents Opioid prescription rates by age group, Florida and Maine, 2012 1,400 1,200 1,000 800 600 400 200 Florida Maine 0 <18 18-24 25-34 35-44 45-54 55-64 65+ Age Group 34
Daily opioid dosage in MME and high dosage by quarter, Florida, 2011-2012 120 Percent and MME/day 100 80 60 40 20 MME/day % > 100 MME/day 0 Q1 11 Q2 11 Q3 11 Q4 11 Q1 12 Q2 12 Quarter/Year Q3 12 Q4 12 Note: First 3 quarters of 2011 data is incomplete and should be interpreted with caution. 35
Percent of prescriptions accounted for by prescriber decile by CS type, Florida, 2012 Percent 80 70 60 50 40 30 20 10 0 1-4 5 6 7 8 9 10 Prescriber Deciles Opioid Benzodiazepine Stimulant 36
Mean daily opioid dosage by prescriber decile by quarter, Florida, Q4 2011 to Q4 2012 120 Mean daily dosage (MME) 100 80 60 40 20-13.4% Top Fifth Tenth TOTAL 0 Q4 2011 Q1 2012 Q2 2012 Q3 2012 Q4 2012 Calendar Quarter Prescriber deciles are based on number of opioid prescriptions. 37
Percent of a prescriber s patients seeing multiple providers by distance deciles, Florida, 2012 Percent 1.0 0.8 0.6 0.4 0.2 0.0 1 2 3 4 5 6 7 8 9 10 Prescriber Distance Deciles 180 160 140 120 100 80 60 40 20 0 Mean miles to prescriber. Prescribers are divided into deciles according to the mean distance between them and their patients for all CS prescriptions. Multiple providers means 5+ prescribers and 5+ pharmacies in 3 months. Includes out of state residents. 38
New and Promising Practices Collect ID of person picking up prescription Mandatory provider registration with and use of PDMP Batch data sharing with 3 rd party payers (Medicare, Medicaid, public Workers Comp) Interoperability of PDMP data with health information exchanges, electronic health record systems, pharmacy dispensing software Evaluate prescriber education initiatives
Collect ID of Person Picking Up Prescription Require pharmacies: To do photo ID check before dispensing a controlled substance Rx to verify who has the drug To submit ID information on who picks-up each prescription -- so PDMP knows who actually has the drug MA PDMP has mandated such reporting and positive ID for Schedule II prescriptions since 1/2/2009 MA found 38% of the persons who dropped off or picked up the prescriptions are not the patient As of 1/1/2011, MA requires reporting and positive ID for all Schedule II to V prescriptions
Mandatory Provider Registration and Use Recently begun in Kentucky (July, 2012) Provider must check PDMP at: First C-II or C-III hydrocodone prescription or change in drug Continued prescribing of these Rx at three months Average weekday requests to KASPER: Before mandate -- 2,900 After mandate -- 19,000 Legislation passed in three other states (MA, NY, TN), in process of being implemented No evidence as yet of effects on patient care and outcomes
Batch Data Sharing with 3 rd Party Payers Most PDMPs allow Medicare, Medicaid, and/or state insurance programs to access PDMP data on individual patients WA PDMP first example of batch data sharing, with Medicaid agency and state Workers Comp agency > 2,000 Medicaid patients were found to have obtained prescriptions using both Medicaid and cash on the same day in 2012
Interoperability of PDMPs with HIEs, EHRs, Pharmacy Dispensing Software The Substance Abuse and Mental Health Services Administration (SAMHSA) has recently awarded grants to 9 state PDMPs to improve interoperability: Integrate PDMP data into EHRs (e.g., for hospital ED) and in pharmacy dispensing software Many of these projects build on earlier pilot studies conducted by MITRE Cross-site evaluation by the Centers for Disease Control and Prevention
Evaluate Prescriber Education Initiatives Many efforts underway nationwide to influence prescriber behavior The FDA has engaged PDMP Center of Excellence to inventory these efforts and summarize the evidence base PDMP data can be used to evaluate whether desired changes in prescriber behavior have occurred Population-based evaluation only is possible, since PDMP data contain no medical information
PDMP Center of Excellence White Paper on PDMP Best and Promising Practices Available at: http://www.pewhealth.org/uploadedfiles/phg/content_level_pages/r eports/pdmp_full%20and%20final.pdf
Contact Information Peter Kreiner, Ph.D. Principal Investigator PDMP Center of Excellence Brandeis University 781-736-3945 pkreiner@brandeis.edu www.pmpexcellence.org