How Do You Measure Patient Activation?

Patient activation is formally measured with the Patient Activation Measure (PAM), a licensed 13-item survey (a 10-item short form exists) that scores patients 0 to 100 and groups them into four levels. The PAM is proprietary, and most specialty practices will never license it. They do not need to: you can measure activation with behavioral proxies your practice management system already records, chiefly booking rate, recall return rate, referral completion rate, and time-to-rebook.

What the PAM actually is

The Patient Activation Measure was developed by Judith Hibbard and colleagues at the University of Oregon in 2004. Patients rate their agreement with a series of statements on a four-point scale, from disagree strongly to agree strongly, and Rasch calibration converts the responses into a score from 0 to 100. The original instrument had 22 items; the PAM-13 short form is what most studies and health systems use today, with a 10-item version available where intake time is tight.

The items are copyrighted, so we will not reprint them here, but the themes are consistent. They probe whether the patient believes their role in their own care matters, whether they know what their medications do and when to seek help, whether they feel confident handling new problems at home, and whether they can keep healthy behaviors going under stress. The score is a snapshot of knowledge, skill, and confidence, not of satisfaction or loyalty.

This post stays narrowly on measurement. For the category definition, and how the activation motions fit around it, start with our cornerstone guide to what patient activation is.

The four PAM levels and what each should change about your outreach

PAM scores band into four levels. The academic literature describes what patients at each level believe; almost none of it tells you what to do differently at the desk. Here is the operational translation:

• Level 1, disengaged and overwhelmed: these patients do not yet see themselves as participants in their own care. Go phone-first and high-touch: a live conversation, one small ask at a time, and scheduling handled for them rather than a portal link.
• Level 2, becoming aware but still struggling: willingness is forming but knowledge and confidence lag. Pair short education with a direct booking path, and plan on more reminder touches at shorter intervals.
• Level 3, taking action: these patients respond well to digital self-service. Text-first outreach with a self-scheduling link usually converts without staff involvement.
• Level 4, maintaining behaviors and pushing further: light touch only. An annual recall reminder and a one-tap booking path are enough, and over-contacting this group burns goodwill and staff time.

The pattern is simple: the lower the level, the more human effort each contact needs. The higher the level, the more automation can carry.

Why activation scores matter

The evidence base is why the PAM became a fixture in payer and health-system contracts. Three studies carry most of the weight:

• Cost: among 33,163 patients, those at the lowest activation levels had predicted average costs 8% higher in the base year and 21% higher in the first half of the next year than the most activated patients (Hibbard, Greene & Overton, Health Affairs, 2013).
• Outcomes: higher activation was associated with better performance on 9 of 13 health outcomes and with lower costs two years later, and when activation changed, outcomes and costs moved in the expected direction (Greene et al., Health Affairs, 2015).
• Utilization: hospitalized patients at PAM Level 1 had 1.75 times the rate of 30-day post-discharge hospital utilization (ED visits plus readmissions) compared with highly activated patients (Mitchell et al., Journal of General Internal Medicine, 2014).

For a specialty practice, the takeaway is practical rather than academic. Your least activated patients are where no-shows, dormant charts, and incomplete referrals concentrate, so knowing where activation is low tells you where outreach effort pays.

The licensing reality

The PAM is proprietary. The items are copyrighted, scoring requires the licensed algorithm, and commercial terms are negotiated with the rights holder rather than published on a price list. Health plans and large systems license it because activation scores feed risk stratification and care-management triage at population scale, and in value-based contracts that math works.

For a private specialty practice, the math rarely does. You would add a survey to intake, pay for a license, train staff, and store one more score in the chart, and the score itself never books an appointment. The PAM is a well-validated instrument; the issue is fit, not quality.

What to measure instead: the behavioral proxy stack

Activation shows up in behavior, and the behavior is already sitting in your systems. Four ratios cover it, each computable monthly from your PM system, phone logs, and referral records:

• Booking rate: appointments booked divided by total inbound contacts (calls, web forms, chats) in the same period. This is the activation number for your website and phones; across Clinekt deployments, 98% of website visitors leave without making contact, so the true denominator is far larger than form submissions suggest.
• Recall return rate: dormant patients who completed a visit divided by dormant patients contacted. This is the best single proxy for activation inside your existing panel; our playbook on how to reactivate dormant patients covers the campaign mechanics.
• Referral completion rate: referrals that ended in a completed visit divided by referrals received. Track it by referring practice so you can see exactly where handoffs break.
• Time-to-rebook: median days from a missed or cancelled appointment to the next completed visit. A rising number here is often the earliest signal that patients are quietly disengaging.

The fair objection is that these metrics might just reflect the patients you happen to have. The evidence says the behavior moves when you act on it. When a provider encouraged portal use, 87% of patients accessed the portal within the year, versus 57% of patients who got no encouragement, a 30-point gap from outreach alone (ASTP/ONC Data Brief 77, 2024). A Cochrane review of 55 studies with 138,625 participants found reminder and recall outreach raised preventive-care completion by 8 percentage points on average (risk ratio 1.28), with telephone outreach the most effective channel (Jacobson Vann et al., Cochrane, 2018).

Setting benchmarks honestly

Published benchmarks for booking rate, recall return rate, and time-to-rebook largely do not exist. The PAM literature is academic, and most vendor benchmarks are self-reported marketing, so treat any figure you cannot trace to a named source with suspicion. That standard applies to us too, which is why every number on this page carries its source.

The workable approach is to baseline yourself: compute all four ratios for one full quarter, then manage to your own trend. The adjacent metric that does have published benchmarks is the no-show rate, worth tracking alongside the proxy stack; we break those numbers down in our guide to the average no-show rate and what it costs.

One reference point we can stand behind because we ran it: Baldwin Bone & Joint in Daphne, Alabama booked 159 appointments from 263 qualified surgical leads in a single quarter, a 60% lead-to-appointment booking rate.

From measurement to action

Measurement finds the gap; it does not close it. Whatever your numbers show, the fix is one of three motions: engage the website visitors who would otherwise leave without booking, bring dormant patients back from the EHR, and generate net-new demand you can trace through to completed care. Across Clinekt deployments, 82% of patients try to book outside office hours, which is why the first motion cannot be solved by adding phone staff.

If you are still untangling how activation differs from engagement (the difference determines what kind of software you actually need), read our comparison of patient engagement vs patient activation. Measurement tells you which motion is underperforming; the comparison tells you what closing the gap should look like.

Common questions

What is a good PAM score?

PAM scores run from 0 to 100 and higher is better, but there is no universal passing grade. Scores band into four levels, and movement matters more than the raw number: a patient who climbs one level is the win the research rewards. If you do not license the PAM, a rising booking rate and recall return rate tell you the same story.

Is the PAM free to use?

No. The PAM is a proprietary, licensed instrument: the items are copyrighted, scoring requires the licensed algorithm, and commercial terms are negotiated with the rights holder rather than published. Researchers can apply for academic licenses, but a practice cannot simply copy the questionnaire out of a journal article.

What is the PAM-13?

The PAM-13 is the 13-item short form of the original 22-item Patient Activation Measure. Patients rate agreement with each statement on a four-point scale, Rasch calibration converts the answers into a 0 to 100 score, and the score is grouped into one of four activation levels.

Can you measure patient activation without the PAM?

Yes. Track the behavioral proxies your systems already record: booking rate (appointments per inbound contact), recall return rate (dormant patients who return after outreach), referral completion rate, and time-to-rebook. They measure what activated patients actually do, and they respond to outreach, which is the point of measuring in the first place.

If your proxy metrics say patients are leaking away between interest and appointment, the fix is an action layer, not another survey. Clinekt's AI agents engage website visitors, bring dormant patients back, and trace new demand through to completed care, on a flat subscription with no long-term contracts, live in under a week. Estimate what the gap costs your practice with our leakage calculator, check the FAQs, or book a demo.