Connecting state and local government leaders
Analytics could help us stem the opioid crisis. We just need to solve a few red tape and data problems.
Earlier this year I attended the National Prescription Drug Abuse and Heroin Summit in Atlanta, where Prescription Drug Monitoring Programs (PDMPs) were a frequent topic of panels and presentations. The basic purpose of PDMPs is to collect, house and standardize data from pharmacies and dispensing providers in regard to controlled substances. PDMPs provide that information back to providers and pharmacists so they can see a patient’s prescription history and, ideally, help them identify high-risk patients who may need early intervention.
The opioid epidemic has put PDMPs in the spotlight, but they’ve actually existed for quite some time.
Believe it or not, California set up the first PDMP in 1939 to monitor narcotic prescriptions and tracked them by numbered forms. The early 1990s saw the first PDMPs to collect information in electronic format and the first state—Nevada—to provide data directly to pharmacists and prescribers. To date, 49 states, the District of Columbia and the Territory of Guam have some form of PDMP. Recently Missouri stated that it is working on a public-private partnership to develop a version of a prescription monitoring program.
I give this history lesson to make it clear that PDMPs are not some new concept, suffering growing pains and struggling towards maturity.
At the summit, many states touted their PDMP for the improvements in their user interfaces, the reporting capability and the new movement toward data integration. These are all important aspects to a successful “monitoring” program, and I would never downplay a PDMPs role in pulling a person from the depths of addiction. What no state could claim, however, is that their PDMP led to a decrease in the number of overall opioid-related deaths.
This should come as no surprise. Opioid epidemic numbers are appalling. According to the American Society of Addiction Medicine, drug overdose is the leading cause of accidental death in the U.S., with 60 percent of overdose deaths attributable to opioids in 2015. More than 20,000 were due to prescription opioids and nearly 13,000 resulted from heroin. As heroin is cheaper and easier to score than prescription drugs, it is no wonder that four out of five new heroin users start out misusing prescription painkillers.
My great fear is that, at the 2018 National Prescription Drug Summit, we will still be talking about surveilling a problem that will take thousands of lives between now and next year’s event. How can we move from monitoring to more widespread intervention, and saving more lives?
Combining new data sources with PDMP data, and applying analytics, can give us new insights to aid people in the throes of addiction.
This is not easily done, and requires slogging through entrenched obstacles and webs of red tape. But it is worth it, and more than ever before, IT departments, politicians, providers and pharmacists are clamoring for new ways to combat the opioid scourge.
It starts with good policy that ensures cross-agency cooperation and facilitates data use agreements between agencies so that new, and revelatory, data sources can be combined with PDMP and other data. For instance, substance abuse treatment data is critical to establishing recidivism rates. How frequently are individuals with a substance use disorder who went through rehab having relapses? That’s an important indicator of program efficacy, and yet I’ve heard many times how that data cannot be combined with PDMP data.
While there are many rules and regulations around this information there are ways to keep this data in practitioner hands so that they can use it within the confines of the patient-provider relationship. Similar to the way a doctor uses an electronic medical record now, where a system may alert them when they miss a diagnostic test, or when there could be an adverse drug interaction, etc., the opioid approach is no different. The system would notify the doctor when there is a hospitalization, ER visit, death or overdose, and provide risk scores and assessments, much like their current systems. Aggregate data is kept at a high level, and interventions remain between provider and patient.
Massachusetts accomplished it, and it generated new insights into strategies for confronting the commonwealth’s opioid problem. A prime example of good policy, Chapter 55 of the Acts of 2015 empowered the Massachusetts Department of Public Health to analyze 10 datasets from five government agencies, and produce a comprehensive report on opioid abuse trends and actionable steps to combat the problem.
According to the report, “In total, 29 groups from government, higher education, and the private sector provided information and expertise. This level of partnership is what makes the Chapter 55 report a milestone achievement in Massachusetts. Before this legislation was passed, such a comprehensive look at the opioid epidemic in the Commonwealth would not have been possible.”
Some of the report’s findings include:
- More than two-thirds of people who died from an opioid-related overdose had a legal opioid prescription at some point from 2011-2014. However, only about 1 in 12 of those who died had an opioid prescription in the month before their death.
- The fatal opioid-related overdose rate for individuals with three or more opioid prescribers is seven times higher than the rate for other people.
- Illegal use of Fentanyl is a serious problem. Data from 2013-2014 shows Heroin, Fentanyl, or both substances were present in an overdose victim’s system in 85 percent of cases. However, only about 3 percent of people who died from an opioid-related overdose had a prescription for Fentanyl at the time of death.
- Men are more likely than women to have Heroin in their systems, while prescription opioids are more likely to be found in women at time of death.
- For individuals age 25–34, opioids were responsible for more than a third of all deaths, rising to more than 40 percent for men in this group.
- Individuals recently released from prison are 56 times as likely to die from an opioid-related overdose, so treatment and overdose prevention services should be expanded in correctional facilities and be standardized, evidence-based, and monitored.
Massachusetts was able to answer more questions by integrating additional data sources with the data in the Commonwealth’s PDMP, and they’ve put that new knowledge to work. Learn more about the steps they’ve taken, and future plans, here.
The U.S. government funds many PDMPs through grants from the Centers for Disease Control and Prevention and the Justice Department, but lacks consistent reporting from states back to the U.S. government. And states struggle with consistent reporting from local health agencies because they often have their own systems, own software and suffer from data quality issues.
The RX Summit was replete with examples of individual data sources showing benefits, but none were able to show reductions in deaths. Predictive analytics produces the most useful results when applied to multiple data sources.
President Trump has declared the opioid epidemic to be a national emergency. Attorney General Jeff Sessions announced a new Opioid Fraud and Abuse Detection unit that will use health care data and analytics to try and address the problem. A similar system using law enforcement data is the logical next step.
The opioid crisis has shocked us all, but it has also motivated us. We have the will, and the way, to tell different stories at next year’s summit. With intelligent policies, data sharing and advanced analytics, we can increase and improve interventions that finally reverse opioid death trends in America.
Dr. Steve Kearney is the medical director and senior manager for health care in the U.S. Government Practice of Cary, North Carolina-based analytics company, SAS. Formerly of Duke University and Pfizer Global Medical, Kearney now leads a team that helps solve complex health care challenges with advanced analytic solutions that can be applied across state, federal and local governments.