For years, “deprescribing” has been the promise on every pharmacy vendor’s slide deck: identify the overmedicated members, get physicians to intervene, watch the savings roll in. New research published this spring puts a number on how hard that promise actually is to keep, and it changes what a smart client conversation about polypharmacy should sound like.
A systematic review and meta-analysis published in JAMA Network Open this May confirmed that deprescribing interventions do reduce polypharmacy among community-dwelling older adults. But the size of that effect is the real headline: on average, roughly seven patients need to go through an intervention to eliminate just one prescribed medication. Generic, low-touch deprescribing efforts move the needle, but barely, unless they’re deployed at real scale.
That modest per-patient effect isn’t the only gap. A separate systematic review of deprescribing implementation trials found that even after a therapy evaluation reaches a prescriber, the recommended change gets implemented only 24% to 67% of the time. A flagged risk and a resolved risk are two very different numbers, and most vendor pitches only ever show the first one.
There’s a second thread worth bringing into renewal conversations: delivery model matters as much as the clinical logic. A 2026 study covered by Medscape found that deprescribing education aimed at both patients and physicians together outperformed outreach to physicians alone. A flag sent into a portal and forgotten produces a very different outcome than a flag routed to someone positioned, and prompted, to act on it.
Here’s why this matters at the table with a client. Deprescribing has become a checkbox term, something almost every pharmacy vendor claims to offer. This research gives you a sharper set of questions to separate a real program from a slide. Ask for the completion rate on flagged cases, not just the number of flags raised. Ask how many members the intervention reached, not the size of the at-risk population it was aimed at. Ask whether physicians are engaged directly, or whether the recommendation stops at a report nobody reads.
This is exactly the gap RazorMetrics was built to close. Every flagged case routes directly to the prescriber, and RazorMetrics sustains a 76% physician response rate—well above the implementation range the research above describes as typical. That response rate, applied across a client’s full population rather than a low-touch pilot, is what turns a modest per-patient effect into durable, population-level savings: part of the $300M and growing that RazorMetrics has helped clients avoid in unnecessary drug spend.
The science says deprescribing works. It also says most programs never generate enough patient-level lift, or enough completed interventions, to matter at a population scale. That distinction is exactly what a client needs to hear before they sign off on a vendor’s deprescribing claim at renewal.
Heading into a client conversation about polypharmacy or deprescribing? Ask your RazorMetrics team for the current physician response rate and completion data for your population—a forwardable, renewal-ready answer to “what’s actually getting fixed?”
