System Dynamics Modeling of Medical Use, Nonmedical Use and Diversion of Prescription Opioid Analgesics Wayne Wakeland, Ph.D. Alexandra Nielsen, M.S. Teresa Schmidt, M.A. 30 th International System Dynamics Conference St. Gallen, Switzerland July, 2012
Overview Background Model Overview Model Testing Policy Analyses Limitations Future Research
Background Number of New Nonmedical Users of Opioid Analgesics (Thousands) Source: SAMHSA (2006). Overview of findings from the 2005 National Survey on Drug Use and Health. (Office of Applied Studies, NSDUH Series H-30, DHSS Publication No. SMA 06-4194). Rockville, MD.
Background Sources of Opioids for Nonmedical Users (% of Respondents) Plus, of those who got them free, 80% reported that their source got the drugs by prescription from a single doctor Internet = percent of people, not amount of drugs Other = multiple doctors, forged prescription, pharmacy theft. Source: Substance Abuse and Mental Health Services Administration. (2010). Results from the 2009 National Survey on Drug Use and Health: Volume I. Summary of National Findings (Office of Applied Studies, NSDUH Series H-38A, HHS Publication No. SMA 10-4586Findings). Rockville, MD.
Background Are there feasible policy changes that could help to address this major health health concern? Warner, M., Chen, L. H., & Makuc, D. M. (2009). Increase in fatal poisonings involving opioid analgesics in the United States, 1999 2006. NCHS Data Brief, 22.
Model Overview
Nonmedical Use Sector Rate of Initiation of Nonmedical Opioid Use US Population Aged Twelve Plus Rate of Initiation During Unlimited Accessibility Initiating Nonmedical Use Opioid Popularity Number of Individuals Using Drugs Nonmedically (Excluding Marijuana and Pharmaceutical Opioids) Low Frequency Nonmedical Opioid Users R Total Number of Individuals Using Opioids Nonmedically R Increasing Frequency High Frequency Nonmedical Opioid Users Link to Full SFD B Accessibility of Pharmaceutical Opioids B <US Population Aged Twelve Plus> Fraction of Demand Met from Chronic Pain Trafficking Supply of Opioids Diverted by Patients Distributing Total Demand for Opioids
Data Support Parameter Support NONMEDICAL USE SECTOR DIRECT INDIRECT PANEL 1 Base Level of Abuse Potential of Pharmaceutical Opioids 1.3 2 Fraction of Demand Met from Chronic Pain Trafficking.25 3 Fraction of Low Freq Users who switch to High Freq.06 4 High Frequency User All-Cause Mortality Rate.02 5 High Frequency User Cessation Rate.08 6 Low Frequency User All-Cause Mortality Rate.012 7 Low Frequency User Cessation Rate.15 8 Number of Days of Nonmedical Use Among High Freq Users 220 9 Number of Days of Nonmedical Use Among Low Freq Users 30 10 Number of Dosage Units Taken per Day 2 11 Overdose Mortality Rate for High Freq Nonmedical Users.002 12 Overdose Mortality Rate for Low Freq Nonmedical Users.0002 13 Rate of Initiation of Nonmedical Opioid Use.006 14 Table Function for the Impact of Limited Accessibility 15 Table Function for the Number of Individuals Using Illicit Drugs Excluding Marijuana and Pharmaceutical Opioids 16 US Population Ages 12 and Older
Number of Dosage Units Diverted Diverting Number of Prescriptions Diverted from Patients with Abuse or Addiction Dosage Units per Prescription Daily Excess Use by Patients with Abuse or Addiction Excess Prescriptions Used by Patients with Abuse or Addiction Excess Prescriptions used by a Patient with Abuse or Addiction B Diversion Sector Number of Excess Prescriptions Acquired through Forgery and Dr. Shopping Fraction of Prescriptions Acquired Through Forgery and Dr Shopping Multiplied by the Profit Motive Multiplier Number of Patients who Engage in Dr. Shopping or Forgery Number of Prescriptions Given to Patients with Abuse or Addiction Link to Full SFD Supply of Opioids Diverted by Patients Demand Among Nonmedical Users Months of Supply Available Profit Motive as a Function of Months of Supply Base Fraction of Excess Prescriptions Acquired Through Forgery or Dr. Shopping
Data Support Parameter Support DIVERSION SECTOR DIRECT INDIRECT PANEL 1 Average Number of Dosage Units Per Opioid Prescription 86 2 Average Number of Extra Dosage Units Taken/day Among 1.5 Patients with Abuse or Addiction 3 Fraction of those with Abuse/Addict who Engage in Dr..5 Shopping 4 Fraction of those with Abuse/Addict who Engage in Forgery.4 5 Number of Days of Extra Opioid Usage Among Patients 50 with Abuse/Addiction 6 Profit Multiplier 15 7 Table Function for Effect of Perceived Risk on Extra Rx Obtained
Medical Use Sector Abuse, Addiction, and Overdose Deaths Among Medical Users Short Acting Patients with Abuse or Addiction Becoming Addicted on Short Acting Patients on Short Acting Opioids Treating New Patients with Short Acting New Chronic Pain Patients Long Acting Patients with Abuse or Addiction Patients on Long Acting or Short and Long Acting Becoming Addicted to Long Acting B Adding or Switching to Long Acting Bias Toward Prescribing Short Acting Treating New Patients with Long Acting B B Treatment Rate for Long Acting Treatment Rate for Short Acting Perceived Risk of Treating with Pharmaceutical Opioids Risk Adjusted Treatment Rate Link to full SFD
Data Support Parameter Support MEDICAL USE SECTOR DIRECT INDIRECT PANEL 1 All Cause Mortality Rate for Patients on Long-acting Opioids.012 2 All Cause Mortality Rate for Patients on Short-acting Opioids.01 3 All Cause Mortality Rate for Patients with Abuse/Addiction.015 4 Average Long-acting Treatment Duration (in years) 7 5 Average Short-acting Treatment Duration (in years) 5 6 Base Level of Abuse Potential for Pharmaceutical Opioids 1.3 7 Base Rate for Adding or Switching (to Long-acting).03 8 Base Rate of Opioid Treatment for Pain.05-.23 9 Base Risk Factor (degree Tx reduced in 95 due to risk) 1.3 10 Diagnosis Rate for Chronic Pain.05-.15 11 Overdose Mortality Rate for Patients Abusing Opioids.0015 12 Overdose Mortality Rate for Patients on Long-acting.0025 13 Overdose Mortality Rate for Patients on Short-acting.0005 14 Rate of Addiction for Patients on Long-acting.05 15 Rate of Addiction for Patients on Short-acting.02 16 Table Function for Short-acting Bias (function of perceived risk) 17 Tamper Resistance (baseline value) 1
Model Testing: Model vs. Initiates RBP KEY RBP: red Model: blue # of People
Model Testing: Model vs. Nonmedical Users RBP # of People KEY RBP: red Model: blue
Model Testing: Model vs. Opioid Deaths RBP Opioid Deaths KEY RBP: red Model: blue
Interventions Prescriber Education Simulated as halving the number of patients per year who become addicted to opioids And doubling prescribers perception of risk, which halved the fraction of pain patients prescribed opioids Popularity Suppression Simulated as reducing the rate of initiation by half
Results: Prescriber Education Opioid Deaths Key: Baseline Model Run: Plots 1, 3, 5 With Prescriber Education: Plots 2, 4, 6 Total Nonmedical Medical
Implications of Prescriber Education Intervention Decreased overdose deaths among medical users because wary prescribers offer opioid therapy to far fewer individuals But with possible denial of therapeutic treatment to some patients with legitimate chronic pain complaints Nonmedical overdose deaths also decrease Due to fewer individuals with abuse or addiction who could engage in trafficking And increased difficulty to obtain fraudulent prescriptions due to heightened prescriber risk perception
Results: Popularity Suppression Opioid Deaths Key: Baseline Model Run: Plots 1, 3, 5 With Popularity Suppression: Plots 2, 4, 6 Total Nonmedical Medical
Implications of Popularity Suppression Intervention Sharply reduced nonmedical initiation and overall population of nonmedical users Substantially reduced nonmedical and total overdose deaths Once the nonmedical user population declines, positive feedback leads to virtuous cycle of decreased use and decreased popularity, which further reduces use and associated deaths Medical usage-related deaths not impacted Pain treatment not inhibited
Limitations Spotty empirical support Focused on trafficking versus interpersonal sharing Excluded consideration of key issues: Effects of poly drug misuse Treatment alternatives Payor policies (formulary and co-pays)
Future Research Use of Monte Carlo analyses to better gauge impacts of parameter uncertainty Improve model regarding interpersonal sharing Create models at the individual behavior level
Acknowledgments This research was supported by the National Institute of Drug Abuse, grant # 1R21DA031361-01A1 Initial funding was provided by Purdue Pharma, LP Special thanks to J. David Haddox, John Fitzgerald, Jack Homer, Lewis Lee, Louis Macovsky, Dennis McCarty, Lynn Webster, and Aaron Gilson for their guidance, data support, and critical review
back
back
back