Prescription Behavior Surveillance System Len Paulozzi, MD, MPH Centers for Disease Control and Prevention Centers for Disease Control and Prevention National Center for Injury Prevention and Control Harold Rogers PDMP National Meeting Washington, DC, September 26, 2013 TM
Prescription Behavior Surveillance System Why? Google s Chief Economist, Hal Varian: The sexy job in the next 10 years will be statistician. Because now, we really do have essentially free and ubiquitous data. So the complementary factor is the ability to understand that data and extract value from it. 2
Itinerary of the Talk Prescription drug monitoring programs (PDMPs) as surveillance systems Background on PBSS Descriptive measures: patients Descriptive measures: providers 3
PDMP Attributes As a Surveillance System: Strengths Simplicity: single data source, few data elements Representativeness: population-based Timeliness: excellent Data quality: insurance and system error checks Acceptability: mandatory Cost: little additional cost to analyze data See: Lee et al, eds., Principles and Practice of Public Health Surveillance, 3 rd edition, 2010. 4
PDMP Attributes As a Surveillance System: Weaknesses Predictive value positive: unclear, metrics untested Flexibility: fields collected not easily changed Stability: funding unstable or inadequate in places See: Lee et al, eds., Principles and Practice of Public Health Surveillance, 3 rd edition, 2010. 5
Role of PDMP Data in CS Prescribing Process Provider PDMP Prescription Feedback Policy Maker Patient Policy 6
Role of PDMP Data in CS Policy Development Provider PDMP Prescription Feedback Policy Maker Patient/ Population Policy 7
Prescription Behavior Surveillance System (PBSS) Support from CDC s Injury Center and FDA Bureau of Justice Assistance funded the PDMP Center of Excellence (COE) at Brandeis University COE tasked to establish Prescription Behavior Surveillance System Purposes: To create a public health surveillance and evaluation tool based on de-identified, longitudinal data from state PDMPs. To inventory and evaluate prescriber educational initiatives that aim to enable safer prescribing of controlled substances, using the PBSS database when feasible to assess the effectiveness of selected prescriber initiatives. 8
Time Course for Surveillance Component of PBSS May 2012: COE sent first invites to PDMPs Dec 2012: PBSS indicators finalized August 2013: 7 states completed Data Use Agreements (DUAs) for the PBSS project, and of these: 5 states provided de-identified PDMP data to COE (CA, DE, FL, ID, ME) 2 states are preparing de-identified data to send to COE (IN, KY) Together represent 23% of the US population Preliminary results in 40+ report tables from FL Draft surveillance report 9
PDMP Information Repurposed in PBSS Patient characteristics Sex and age group Residence zip code Prescriber/pharmacy characteristics Practice zip code Prescription characteristics Date dispensed Drug class, schedule, subtypes (derived from NDC codes) Opioid dosage (derived from other variables) Source of payment (where available) 10
Descriptive Measures/Indicators in PBSS: Patients and Prescriptions Population-based prescription rates Mean daily opioid dosage Percentage of prescribed days with overlap Multiple provider episodes (MPE) per 100,000 residents (using BJA definition) Percentage of rx involved in MPE Specific drug combinations Payer types, e.g., cash (where available) 11
Notes for PBSS Results: Florida PDMP operable in October, 2011 PDMP CS II-IV Dispensers not required to report controlled substances for patients under 16 years of age Results restricted to state residents unless indicated otherwise All results unpublished and preliminary
Rate per 1,000 residents Opioid and benzodiazepine prescription rates, Florida and Maine, by quarter, 2012 (PBSS) 250 200 150 100 50 Opioid FL Opioid ME Benzo FL Benzo ME 0 1 2 3 4 Quarter of 2012 Limited to state residents. 13
Rate per 1,000 residents Opioid prescription rates by age group, Florida and Maine, 2012 1,400 1,200 1,000 800 600 400 Florida Maine 200 0 <18 18-24 25-34 35-44 45-54 55-64 65+ Age Group Limited to state residents. 14
Percent and MME/day Daily opioid dosage in MME and high dosage by quarter, Florida, 2011-2012 120 100 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. 15
Rate per 100,000 residents Multiple-provider episode rates* for CS II drugs, Quarter 4 of 2011 vs. Quarter 4 of 2012, Florida 8 7 6.9 6 5 4.7 4 3 2 1.8 2.6 1.9 Q4 2011 Q4 2012 1 0 0.8 0.0 0.0 <18 18-34 35-54 55+ Age Group *Having CSII rx from 5+ prescribers dispensed at 5+ pharmacies during one quarter. Limited to state residents. 16
Descriptive Measures/Indicators in PBSS: Prescribers Prescription volume by prescriber decile Deciles defined by how many prescriptions per time period Prescription volume by pharmacy decile Deciles defined by how many prescriptions per time period Mean daily opioid dosage prescribed Percent cash payment (where available) Distance travelled 17
Percent of prescriptions 100 90 80 70 60 50 40 30 20 10 0 Two Years Ago at Harold Rogers: Percent of CS II-V prescriptions by prescriber decile by year, KY, 2009 64.3 (top decile) 17.9 8.4 2009 Top 20% of prescribers accounted for 82.2% of all prescriptions. Blumenschein, K, et al. Independent Evaluation of the Impact and Effectiveness of the Kentucky All Schedule Prescription Electronic Reporting Program (KASPER) Institute for Pharmaceutical Outcomes and Policy, Univ of Kentucky, 2010 18
Percent of prescriptions Percent of CS prescriptions by prescriber decile by year, KY, 2009 (CSII-V) and FL, 2012 (CSII-IV) 100 90 80 70 60 50 40 30 20 10 0 64.3 63.2 17.9 17.4 8.4 8.7 KY 2009 FL 2012 Sources: Kentucky: Blumenschein, K, et al. Independent Evaluation of the Impact and Effectiveness of the Kentucky All Schedule Prescription Electronic Reporting Program (KASPER) Institute for Pharmaceutical Outcomes and Policy, Univ of Kentucky, 2010. Florida: PBSS 19
Distribution of CS II-IV prescriptions by prescriber percentiles, Oregon, Jan-Sept, 2012 % of Prescribers % of CS Prescriptions 4 4 21 19 60 92 Oregon Health Authority. Prescription Drug Dispensing in Oregon, October 1, 2011 March 31, 2012 20
Percent Percent of prescriptions accounted for by prescriber decile by CS type, Florida, 2012 80 70 60 50 40 30 20 10 0 1-4 5 6 7 8 9 10 Prescriber Deciles Opioid Benzodiazepine Stimulant 21
Percent Percent of prescriptions accounted for by pharmacy decile by CS type, Florida, 2012 80 70 60 50 40 30 20 10 0 1-4 5 6 7 8 9 10 Pharmacy Deciles Opioid Benzodiazepine Stimulant 22
Mean daily dosage (MME) Mean daily opioid dosage by prescriber decile by quarter, Florida, Q4 2011 to Q4 2012 120 100 80 60 40 20 0 Q4 2011 Q1 2012 Q2 2012 Q3 2012 Q4 2012 Calendar Quarter -13.4% Top Fifth Tenth TOTAL Prescriber deciles are based on number of opioid prescriptions. 23
Percent Mean miles to prescriber Percent of a prescriber s patients seeing multiple providers by distance deciles, Florida, 2012 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. 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. 24
Conclusions about PBSS Compilation and analysis PDMP data from multiple states with less than 6 months lag Millions of records transformed into population-based, actionable information about both patients and providers Information relevant to developing and evaluating state policy initiatives With results from larger numbers of states, relevance to the national situation. 25
Acknowledgements Gail Strickler, Lee Panas, Peter Kreiner, Pat Knue, John Eadie and other members of the Center for Excellence in PDMPs at Brandeis University for producing the information used in this presentation. PDMPs of Florida and Maine for providing initial PBSS data Colleagues at BJA and FDA for working together to provide technical support and funding for this effort. 26
Thank You Len Paulozzi, MD, MPH lpaulozzi@cdc.gov The findings and conclusions in this report are those of the author and do not necessarily represent the official position of the Centers for Disease Control and Prevention/the Agency for Toxic Substances and Disease Registry.