Physician Liaison KPIs: How Health Systems Measure Program Performance

I usually hear "How should we measure the physician liaison program?" about three minutes before someone asks the real question: "Are we getting enough return to justify this budget?" That second question is the one making the room tense, and pretending otherwise is where most bad measurement frameworks start.

Those questions overlap, but they are not the same job. One is operational management. The other is budget defense. When teams mash them together, they usually end up with dashboards full of activity counts that look busy and say very little about whether referral behavior changed.

What I will lay out here is what measurement actually looks like in programs that are functioning well, based on three full program evaluations and supporting work on several others over the past two years. Program structure, size, and clinical context vary considerably, and some of what I describe applies more naturally to health systems with dedicated PL teams of four or more than to smaller programs with one or two liaisons.

Disclaimer: all pricing, timing, and performance figures in this article are directional estimates based on limited samples and should not be treated as guarantees.

The Two Categories of PL Metrics

Before going into specific metrics, it is worth distinguishing between two categories that tend to get conflated in practice.

Activity metrics track what the liaison is doing: visits per week, contacts made, events attended, educational materials delivered. These are easy to capture and tend to be what gets reported upward to leadership because they are consistent and legible. They are also nearly useless as a measure of program impact on their own. A liaison can hit every activity target and produce no meaningful change in referral behavior.

Outcome metrics track what changes as a result of liaison activity: referral volume from visited practices, referral capture rate within defined service lines, time from referral to appointment, leakage reduction. These are harder to measure because they require clean referral data, which many health systems do not have, but they are the only metrics that actually answer the budget question.

A well-constructed measurement framework uses activity metrics as a management tool, to ensure the liaison is in the field covering the right accounts, and outcome metrics as the program evaluation tool, to determine whether the program is driving the intended behavior change.

What Good Measurement Looks Like at Six Months

At the six-month mark, most PL programs should not yet be evaluated on referral outcomes. The relationship-building phase, particularly for liaisons entering a market with no existing provider relationships, takes time to produce measurable behavior change. Expecting referral volume shifts by month three is unrealistic in most contexts and leads to premature program judgments.

What should be measurable at six months:

Coverage completeness. Has the liaison made contact with the target practice list, and how frequently are they returning to established contacts? A liaison who has visited the top fifty practices in their territory at least twice in six months is generally on track. A liaison who has visited thirty-five practices once is behind.

Access quality. Are visits resulting in face time with physicians and clinical staff, or primarily with front desk staff? This requires tracking that most programs do not do rigorously but should. Meetings that stop at the front desk produce no relationship value and no referral behavior change.

Problem identification. Has the liaison surfaced referral friction points, access complaints, or barriers from referring practices? This is consistently the most undervalued early-stage metric. A liaison gathering referral barrier intelligence in the first six months is building the foundation for the outcome improvements that will appear later. If your six-month reporting contains no record of referral barriers identified, that is itself a signal worth investigating.

What Good Measurement Looks Like at Eighteen Months

By eighteen months, a program that is working should show measurable referral behavior change in the practices the liaison has been visiting consistently. Not every practice, and not dramatic numbers in every case, but directional improvement in the visited-versus-unvisited group.

The comparison that matters most is not the whole-program referral volume number. It is the difference in referral behavior between practices with regular PL contact and practices without contact. If that difference is not present at eighteen months, the program has a problem that activity metrics will not surface and that budget conversations will eventually expose.

The three metrics I look at first at the eighteen-month evaluation point, in this order: referral volume delta in high-touch accounts relative to the baseline period, leakage rate trend in the liaison's primary service line, and time-to-appointment trend for referred patients. Related: Physician Liaison Program ROI.

The Most Common Measurement Mistakes

The first, and most common, is measuring the wrong population. Whole-facility referral volume is noise when the liaison covers twelve practices. The signal is in the liaison's specific account list. I've reviewed program reports that looked reassuring on the summary metrics and had no movement at all in the practices the liaison was actually visiting. The aggregate number was being driven by market trends, not the program.

Isolating liaison-touched accounts from the general referral base sounds basic. Fewer programs do it cleanly than you'd expect, because it requires referral data at the practice level and a consistent way of tagging which practices are "active" in the liaison's portfolio. Most health systems have the data. The attribution infrastructure often isn't set up.

The second problem is the absence of a baseline. I've worked with programs that couldn't tell me what referral volume looked like before the liaison started, which means they also couldn't tell me whether it improved. If you're implementing a program or adding a liaison to a new territory, pull the prior twelve months of referral data for that geography before the liaison's first day in the field. It takes about a week to set up properly. Without it, you're operating on impressions rather than evidence, which eventually surfaces as a vulnerability in budget conversations.

The third issue I'd describe more as a framing problem than a pure measurement error. When health systems use contractor liaisons, they sometimes apply different performance timelines without being able to articulate why. The assumption that an internal hire would have ramped faster, or would be producing differently at month nine, is usually not grounded in anything specific. In the programs I've evaluated, individual quality and management quality are much stronger predictors of outcome metrics than employment model. A contractor liaison who gets real onboarding and consistent manager engagement performs comparably to an internal hire on the metrics that matter. MDliaison's marketplace includes physician liaison professionals vetted specifically for program experience, but the management infrastructure on the health system side has to be there regardless of who sourced the candidate. Related: The Complete Guide to Hiring Physician Liaisons.

Connecting Measurement to Budget Justification

A word on something that comes up in nearly every program evaluation I have been part of.

If you implement this framework seriously, it will occasionally tell you something uncomfortable. Some liaison programs are underperforming and no amount of prettier reporting fixes that. You are better off seeing the problem early than carrying it for another budget cycle.

The programs that lose budget are almost always ones that have been measuring activity for two or three years and then, when someone asks whether the spend is justified, have no outcome data to produce. The programs that hold and grow budget are the ones that can show, with reasonable precision, how much referral volume has moved and what revenue that represents.

The AAPPR publishes benchmarking guidance on PL program measurement that is worth reviewing if you are building a measurement system from scratch. Their benchmarking data gives useful comparative context that internal data alone cannot: https://www.aappr.org. SHSMD also publishes strategic guidance on business development program evaluation that complements the operational metrics focus: https://www.shsmd.org.