What is reasonably established

A 2020 systematic review of 29 studies published between 2009 and 2019 found that 11 of 16 studies reported reduced prescribing following PDMP implementation, and that mandatory-access provisions were the feature most consistently associated with benefit. A 2021 cross-sectional analysis combining national mortality data with claims for 23 million commercially insured patients found that mandatory PDMPs reduced high-dose prescribing and deaths involving natural opioids and methadone.

So: PDMPs change prescribing behavior, and mandatory-use designs change it more. That is a real finding and it is why this site supports monitoring programs.

What is contested

Whether they reduce overdose death is genuinely unresolved, and the two strongest reviews disagree in emphasis.

• A 2018 systematic review of 17 US observational studies found low-strength evidence from 10 studies suggesting reduced fatal overdose after implementation, insufficient evidence on nonfatal overdose, and concluded that evidence that PDMPs either increase or decrease overdose is “largely insufficient.”
• A 2019 systematic review of 24 studies covering data years 1993 to 2014 found no overall evidence supporting an association between PDMPs and decreased opioid prescribing or dispensing, with limited and inconsistent evidence for Schedule II reductions, and flagged inadequate covariate adjustment across the literature.

A significant caveat sits underneath all of this: a meta-analysis found that only 57% (95% CI 0.48–0.66) of providers had ever used PDMP data. Studies of PDMP effect are frequently studies of a policy that was substantially not implemented at the level of the individual clinician.

What CDC says about how the data should be used

Clinicians should not dismiss patients from their practice on the basis of PDMP information.

Dowell et al., 2022 — recommendation 9

The guideline notes that PDMP information has been used to dismiss patients from practices, which “might adversely affect patient safety and result in untreated or undertreated pain.”

The reading we take from this

Monitoring produces genuinely useful information and should continue. What the literature shows is that information alone, without clinical context and without anywhere for the patient to go, produces refusal — and refusal produces substitution. That pattern is visible in the heroin findings, in the pharmacist refusal data and in the qualitative work, and it is the same pattern as step three.

Which is the argument for supplying the missing context and for making sure there is somewhere for the patient to go, rather than for monitoring less.

Questions this raises

Do PDMPs reduce overdose deaths?

It is unresolved. A 2018 systematic review of 17 studies found low-strength evidence of reduced fatal overdose but concluded the evidence either way is largely insufficient, and three of six studies it reviewed found increased heroin overdoses after implementation. Anyone stating a confident answer is ahead of the literature. See what population trends can and cannot tell you.

Should an agency stop using its PDMP?

No, and nothing here argues that. Monitoring produces useful information. The finding worth acting on is that information without clinical context tends to produce refusal, and refusal tends to produce substitution toward the illicit supply. The remedy is to add context, not to remove monitoring. See ProviderSynch.