How effective are PDMPs?

05/24/2017

All but one state has established a prescription drug monitoring program (PDMP), with California being the first in 1939, and Missouri remaining the lone holdout. Such databases have indeed made it more difficult for individuals to go doctor shopping for prescription medication, but research is still ongoing as to what type of impact PDMPs are having on reducing opioid overdose deaths and improving interventions.

There is evidence that they are valuable. Overall, PDMPs have helped reduce instances of doctor shopping, reduced the overall number of opioid prescriptions being written and helped physicians intervene earlier when it appears patients may be abusing or diverting pain medication. And there is some tentative evidence that they are reducing overdose deaths as well.

However, without access to comprehensive treatment services or standard protocol for what physicians should do once they identify a patient that may have a substance use problem, states have not adequately addressed the crux of the opioid crisis: Denied access to prescription opioids, many individuals often turn to diverted sources or illicit opioids like heroin.

While prescription-opioid-related deaths are beginning to level off or even decline in some regions, heroin overdose deaths increased by 39% between 2012 and 2013, and rose 20.6% between 2014 and 2015, according to the Centers for Disease Control and Prevention (CDC). A 2014 study published in JAMA Psychiatry found that 75% of more recent heroin users began opioid use via prescription drugs.

However, there is still significant work to be done when it comes to effectively using PDMPs as clinical tools to help identify and treat those with addiction.

“We have to be realistic,” says Sherry L. Green, president of the National Alliance for Model State Drug Laws (NAMSDL). “The PDMP helps identify people whose pattern of behavior suggests an abuse or addiction problem. It’s an identification tool for prescribers. But we can’t just stop there. If we are going to refer someone, we need to make sure we follow through that process systematically and not only refer them but have the resources to make sure we’re getting them the help they need. That piece is not getting as much focus in this country as it should be.”

Physician objections

While doctors have often voiced complaints about using PDMPs—particularly when mandates are in effect—resistance to them is not nearly what it was earlier in the decade.

“I don’t think there are really any more general objections to the concept of the program, but there are still legitimate concerns being raised about transforming these PDMPs into better healthcare tools,” Green says.

More states have shifted to automated registration to make the databases easier for physicians in their daily workflow. There are also efforts to better integrate PDMP queries with existing clinical systems and to allow other staff to make those queries, to help make the process more efficient.

“We’re seeing efforts to fit the PDMP into the workflow,” says Cynthia Reilly, director of the Pew Charitable Trusts’ substance use prevention and treatment initiative. “In some cases, they can integrate the PDMP into other health IT systems. In New Jersey, there is a mobile app. There are a lot of strategies to decrease the burden on prescribers.”

There’s also a second challenge facing physicians in that most are not trained to have discussions about addiction with their patients.

“Having the data from the PDMP can actually help them have that conversation because they can point to that data,” says Peter Kreiner, principal investigator at the Prescription Drug Monitoring Program Training and Technical Assistance Center (PDMP TTAC) at Brandeis University. “But often there is a lack of treatment resources, or the doctor or practice may not have the information about what sort of treatment resources are available.”

Measuring PDMP effectiveness

PDMPs are being studied both at the state and federal level using a number of different measures to gauge effectiveness. In general, the most common data involve the opioid prescribing rate and the number of multi-provider incidents (doctor shopping) that occur. There are some limits to these outcomes, given that prior to PDMP, implementation there is little or no data to compare.

More research is coming, experts say.

“I can tell you based on the calls we get that there is an increase in the number of researchers who are trying to assess the impacts of these programs,” Green says.

According Reilly, assessments vary from state to state. In some cases, there are large formal projects in conjunction with local universities; in others, the data is collected by state health departments and might not be as widely publicized.

“They are looking at intermediate or proxy measures, such as the impact on prescriptions dispensed or multi-provider episodes,” Reilly says. “Generally, analyses are showing that if prescribers see this information, it increases their awareness that other providers are engaged in patient care, and it helps them make more informed prescription decisions.”

PDMP data from 12 states is also being gathered in the Prescription Behavior Surveillance System, a collaboration among PDMP TTAC, the CDC, the Food and Drug Administration, and the Bureau of Justice Assistance. The de-identified data has helped provide a longitudinal early warning system that not only helps spot trends, but also helps participating states gauge their own performance against other states.

Measuring overdose deaths in relation to PDMPs alone has been more challenging because there are so many other programs in place in states that likely would affect those numbers.

“When you look at things the PDMP directly affects, you’re looking at things like doctor shopping,” Kreiner says. “Some have looked at treatment admission rates. Findings have been mixed, but it’s a moving target because some of these studies were done when only half or two-thirds of theses PDMPs were in place. Also, you would really need prescription data from before the PDMP, and most states typically don’t have that type of data.”

Positive indicators

PDMPs are having a clear impact on opioid prescribing nationally. In a study published this spring, for example, University of Kentucky College of Public Health researcher Hefei Wen evaluated data from the Centers for Medicare and Medicaid Services in 46 states and found that PDMP mandates reduced Schedule II opioid prescriptions and spending by 9% to 10% from 2011 to 2014.

Additionally, a study published in Health Affairs in 2016 and led by researchers at Vanderbilt University found that implementation of PDMPs was associated with the prevention of approximately one overdose death every two hours on average.

“Implementation of Prescription Drug Monitoring Programs Associated with Reductions in Opioid-Related Death Rates” found that a state’s implementation of a program was associated with an average reduction of 1.12 opioid-related overdose deaths per 100,000 population in the year after implementation. The study authors further estimated that if lone PDMP-holdout Missouri implemented a monitoring program and other states enhanced existing programs, there would have been in excess of 600 fewer overdose deaths nationwide in 2016—or two per day.

Stephen Patrick, assistant professor of health policy at Vanderbilt and lead researcher on the study, has called for adoption of a nationwide PDMP.

PDMP best practices

Pew gathered a number of state-level studies for its 2016 “Prescription Drug Monitoring Programs: Evidence-based practices to optimize prescriber use” report. The Pew report and work by the PDMP TTAC have highlighted some of the most effective features of state PDMPs in terms of participation and actual affect on prescription rates and overdose incidents.

1. Mandates

Comprehensive mandates can result in dramatic reductions in doctor shopping and opioid prescription levels, although physicians’ groups often oppose mandated programs. Mandates vary—some states only require registration, while others require PDMP queries in specific circumstances. Currently, only 13 states require opioid prescribers to check PDMPs.

“The studies that have been done show that comprehensive mandatory use laws have clear effects on doctor shopping rates,” Kreiner says.

Prescriber enrollment varies widely among states. Mandates (comprehensive or otherwise) increase enrollment and use of the data. A study in Kentucky (which has one of the most robust PDMP programs and mandates) found that the mandates reduced multiple provider episodes by more than half and produced a more than fivefold increase in PDMP utilization.

2. Delegation

Most PDMPs allow prescribers to authorize other staff to access the PDMP on their behalf. This type of delegation can increase use of the PDMP. In Oregon, for example, delegation led to a 30% increase in the use of the PDMP by healthcare personnel.

3. Integration

Ease of use can be greatly improved if PDMP look-ups are integrated into pharmacy management, electronic medical record, and other clinical systems that prescribers and pharmacists are already used to working with. If doctors have to log into a separate system, that eats up valuable time and creates extra steps.

So far, efforts in this type of integration have been on a relatively piecemeal basis. SAMHSA has funded projects through its PDMP Electronic Health Records Integration and Interoperability Expansion (PEHRIIE) program in nine states.

Each state took a different approach, but results showed that integration generally improves outcomes. For example, Anderson Hospital in Illinois was able to integrate PDMP data with EHRs, which resulted in a dramatic increase in solicited reports per registered prescriber at the hospital, as well as a 22% decrease in opioid prescriptions over two years, and a 41% decrease in the number of patients who receive at least one opioid prescription from Anderson prescribers during that period.

In Virginia, Purdue Pharma and the state’s department of heath and human resources have formed a partnership to connect the state PDMP with electronic health records used by prescribers and pharmacists. Other companies have also been working on this problem, including EHR/EMR vendors and large pharmacy chains like Kroger.

4. Unsolicited Reports

In some states, the PDMP data is evaluated and analyzed to spot potential problem behavior on the part of patients, prescribers, or both. Those patients can then be flagged so that their physicians will receive unsolicited alerts or reports.

This takes the onus of evaluating the prescription information off the doctor, which saves time and potentially makes it easier to spot a developing dependency problem earlier. According to Pew, roughly two-thirds of state programs provide these types of notifications.

“The reports identify problems up front and allow for earlier interventions,” Reilly says. “Each state sets it own thresholds in terms of identifying patients that may be at risk.”

Taking the next step

PDMPs have made it easier to identify patients with emerging substance use problems, but there has been no uniform approach to helping those patients once doctors have that information. Meanwhile, addiction treatment is underfunded and difficult to access.

“There is still a gap between treatment need and treatment capacity,” Reilly says. “That gap still has to be addressed.”

It’s not always clear how physicians are responding. They might be “firing” patients, or referring them to treatment. Some states have attempted to measure physician responses, but Green says there is a scarcity of that type of data, too.

"Just because you gave a patient a referral to treatment doesn’t mean they are going,” she says.

A small focus group study led by the Oregon Health and Science University in 2014 found that physicians take widely varying approaches to discussing PDMP findings with patients (from opening up a dialogue, to quietly getting them to leave the building). Most reported receiving no training beyond how to access the PDMP.

Addiction treatment professionals have a role to play, but are somewhat limited in terms of what they can do because only a few states allow non-prescribing clinicians access to PDMP data. While methadone and buprenorphine prescribers typically don’t check PDMPs for opioid prescriptions, advocates believe they should. And in most cases, such prescribers are not adding the methadone or buprenorphine dispensing to the PDMP for other providers to see.

“There are efforts to better link PDMPs and the treatment community, but we have to be careful because there are patient confidentiality concerns here,” Green says.

Regulations in 42 CFR Part 2 typically prevent broad access to patient information among providers, which would include PDMPs.

“There has been some reluctance to incorporate those types of queries because of confidentiality concerns,” Kreiner says. “But there are some potentially harmful effects when it comes to patients having prescriptions that providers don’t know about.”

Both SAMHSA and CDC have been awarding state grants for improvements in prevention and treatment services.

“Some grantees have proposed doing projects that would detail how physicians can better deal with patient referrals, and how to improve information systems when it comes to tracking those patients,” Kreiner says. “Those are still in the early stages.”

The resource shortage is still the biggest obstacle.

“Only 10% of people who need treatment are getting it,” Reilly says. “More clinician engagement can go a long way in addressing that gap. The PDMP can be a tool to have discussions with those patients so that treatment can be more successful.”

Brian Albright is a freelance writer based in Ohio.