What Is the PROMIS Questionnaire and How Is It Scored?

PROMIS stands for Patient-Reported Outcomes Measurement Information System, a collection of standardized questionnaires developed by the National Institutes of Health to measure how people feel and function in their daily lives. Launched in 2004, the project was designed to replace the patchwork of older, less precise surveys that clinics and researchers had been using for decades with a single, flexible system that works across diseases and settings.¹ If you have been handed a PROMIS form at a doctor’s office or asked to complete one on a tablet before your appointment, what follows covers what the scores mean, how the system works, and why clinicians increasingly rely on it.

Why PROMIS Was Built

Before PROMIS, researchers and clinicians relied on dozens of different questionnaires to measure things like pain, fatigue, and depression. The problem was not a shortage of surveys but the opposite: too many instruments that could not be compared with one another. A pain score from one study using one questionnaire had no direct relationship to a pain score in another study using a different one. The NIH set out to fix this by creating a shared measurement system, designed from the ground up to be precise with fewer questions, flexible enough to let clinicians swap individual items in or out, and standardized so that a score of 60 on physical function means the same thing whether the patient has arthritis, cancer, or a broken ankle.² The initiative was structured as a five-year cooperative research program to develop, validate, and standardize banks of questions relevant across common medical conditions.³

What It Actually Measures

PROMIS covers three broad areas of health, each broken into specific domains. Physical health domains include physical function (can you climb stairs, carry groceries, get dressed), pain interference (how much pain gets in the way of daily life), fatigue, and sleep disturbance. Mental health domains include depression, anxiety, anger, and applied cognition, meaning how well you think you are thinking. People screening positive for depression or anxiety on standard clinical tools report significantly worse scores on the PROMIS cognition scales, which makes them useful for flagging cognitive difficulties that travel alongside mood problems.⁴ The depression, anxiety, and anger item banks were calibrated together as part of a broader emotional distress framework.⁵

Social health rounds out the picture. PROMIS measures both your ability to participate in social roles and activities and your satisfaction with that participation, across contexts like family, friends, work, and leisure. The social health items use a seven-day recall period, asking how often or how much something applied in the past week.⁶

How the Scoring Works

Every PROMIS measure produces a T-score. The general population average is set at 50, and every 10 points represents one standard deviation above or below that average.⁷ So a physical function score of 40 means you are functioning about one standard deviation worse than the typical American adult. A fatigue score of 60 means your fatigue is one standard deviation higher than average. The reference population is usually the U.S. general population, though country-specific reference values have been developed for pain, physical function, and participation measures in other nations.⁸

This is where it helps to know which direction is “better.” For physical function and social participation, higher scores are better: a 60 means you are doing well. For symptoms like pain, fatigue, anxiety, and depression, higher scores are worse: a 60 means you are experiencing more of that symptom than most people. The same metric is used in oncology populations as well, where a score of 50 still represents the general U.S. population mean, allowing clinicians to see immediately how a cancer patient’s symptoms compare to the broader public.⁹

Short Forms Versus Computerized Adaptive Testing

You will encounter PROMIS in two main formats. Short forms are fixed sets of questions, typically four to twenty items, that every patient answers the same way. Computerized adaptive tests, or CATs, are dynamic: an algorithm picks the next question based on how you answered the previous one, zeroing in on your level of function or symptom severity with fewer questions. In a study of patients undergoing hip replacement, CATs averaged about five items per domain, compared to eight to twenty items for short forms.¹⁰

CATs tend to perform somewhat better than short forms, especially at the extremes of health. Across all domains, the average number of items administered in CATs is about five, and they provide a wider accurate range, meaning they do a better job distinguishing among people who are very sick or very healthy.¹¹ That said, the gap is not enormous. When researchers compared CATs and short forms head-to-head for measuring depression, short forms performed only marginally worse, and using a branching format (where a short initial set of questions directs you to a second set) narrowed the difference further.¹² In pediatric sickle cell disease, CAT and short form scores correlated strongly, though the short forms showed more floor and ceiling effects in domains like fatigue and pain interference.¹³

In practice, CATs require a computer or tablet and a connection to the scoring software, so clinics without that infrastructure tend to use paper short forms. Both are valid. CATs are just faster and slightly sharper, particularly for patients whose symptoms are unusually mild or unusually severe.

What Counts as a Meaningful Change in Your Score

A common question from patients and clinicians alike is: how much does a score need to change before it reflects a real, noticeable difference? Researchers call this the minimal clinically important difference, or MCID. The answer varies by condition and domain, but the evidence clusters around roughly three to six points on the T-score scale for most physical function and pain interference measures.

In foot and ankle surgery, the overall MCID for physical function was about 5.5 points and for pain interference about 5.8 points, though the specific surgery mattered a lot: hardware removals had much smaller thresholds than soft tissue surgeries.¹⁴ For patients recovering from blood vessel procedures for peripheral artery disease, the MCID for physical function converged around 4 to 5 points.¹⁵ In patients with distal femur fractures, the threshold was a bit lower, around 3.5 points overall, and it varied by how far out from injury the measurement was taken: about 2 points at the early post-injury stage and nearly 4 points by six months out.¹⁶

The practical takeaway is that a change of three to six T-score points generally signals something you would feel in your daily life, not just a statistical blip. Clinicians use these thresholds to decide whether a treatment is actually working for you or whether the numbers have moved on paper without moving in reality.

How PROMIS Compares to Older Questionnaires

If you have been through a clinical trial or a rheumatology clinic, you may have filled out the SF-36, the HAQ, or one of many condition-specific surveys. These “legacy” instruments have decades of data behind them, but PROMIS was designed to improve on them in specific ways. Researchers have built crosswalk tables showing that PROMIS physical function, the SF-36 physical functioning subscale, and the HAQ Disability Index all measure essentially the same concept, which means older data can be translated into the PROMIS metric and compared directly.¹⁷

The advantage of PROMIS shows up most clearly in responsiveness, meaning how well a questionnaire detects real changes over time. In a study of over 450 patients with rheumatoid arthritis followed for a year, the PROMIS 20-item physical function form was more responsive than both the SF-36 and the HAQ at detecting clinically meaningful shifts in function.¹⁸ Researchers have also added new items at the extreme high and low ends of physical function, extending the measurable range so that both elite athletes and severely impaired patients get meaningful scores rather than bumping up against a ceiling or floor.¹⁹

Where Clinicians Use It Most

Orthopedics has been one of the fastest-adopting specialties. PROMIS physical function measures are considered superior to legacy instruments in several orthopedic populations.²⁰ A large retrospective study of patients presenting to orthopedic providers found that about 75% reported at least mild limitations in physical function, and a similar proportion reported at least mild pain interference.²¹ One particularly useful finding is that preoperative PROMIS scores can predict surgical outcomes. In foot and ankle patients, those with a physical function T-score below about 30 before surgery had an 83% chance of achieving a meaningful improvement afterward, while those scoring above 42 had a 94% chance of failing to reach that threshold, likely because they did not have much room to improve.²²

In oncology, PROMIS CATs have been integrated into routine outpatient cancer care for symptom screening, with scores feeding directly into electronic health records and triggering referrals to supportive care when symptoms cross a threshold.²³ The system’s flexibility across diagnoses and treatment phases makes it well-suited to the cancer continuum, from diagnosis through active treatment into survivorship.²⁴

Rheumatology is another stronghold. In rheumatoid arthritis, PROMIS Fatigue short forms have been shown to capture fatigue across the full spectrum of disease activity in diverse patient groups.²⁵ In lupus, PROMIS fatigue scores correlated strongly with the Fatigue Severity Scale, an established measure in that field, with excellent internal consistency.²⁶ The PROMIS-29, a profile measure that covers seven domains plus a pain intensity item, offers “guarded support” across rheumatoid arthritis, osteoarthritis, fibromyalgia, and lupus, able to distinguish among different levels of health and measuring similar constructs as older questionnaires.²⁷

Pediatric Versions and Parent Proxy Reports

PROMIS was not built for adults alone. Pediatric item banks cover domains like pain interference, physical activity, stress, family relationships, and life satisfaction, with versions validated for children and adolescents.²⁸ A pediatric pain interference scale was developed using the same statistical framework as the adult versions, giving researchers calibrated items specifically for children.²⁹

For younger children or those unable to self-report, parent proxy versions are available. These forms are subsets of the pediatric self-report items, adjusted for items where parent responses tended to cluster unnaturally. An important caveat: parent proxy reports showed only moderate to low agreement with children’s own self-reports.³⁰ Parents tend to rate their child’s symptoms somewhat differently than the child does, which is a known challenge across all pediatric patient-reported outcome measures. When a child can complete the form themselves, that is generally preferred.

Translations and Use Outside the United States

PROMIS was originally developed in English for U.S. populations, but there has been a concerted effort to translate and culturally adapt the item banks for other languages. The Dutch translation of the physical function bank, tested in arthritis patients, was found to be semantically and conceptually equivalent to the original.³¹ Arabic translations of PROMIS measures for breathlessness and sleep-related impairment similarly achieved equivalence of meaning while using more colloquial phrasing appropriate to the target culture.³² A large-scale translation of eight pediatric item banks into German and Spanish found that the vast majority of items posed no or only minor translation difficulties, and cognitive interviews confirmed that over 90% of the final items were appropriate for children aged eight to seventeen.³³

The translation process follows a rigorous methodology designed to preserve conceptual equivalence rather than just literal word-for-word accuracy. This matters because a question about physical function phrased naturally in English might carry different connotations in Arabic or German. The goal is for a patient in Amsterdam and a patient in Philadelphia with the same level of physical function to produce the same score.

Getting It Into the Clinic

A questionnaire is only useful if it gets into patients’ hands at the right time and its results reach clinicians in a format they can act on. One of the more practical developments in PROMIS adoption has been integration with electronic health records. A middleware system adopted across 27 clinics at one hospital system collected over 70,000 patient-reported outcomes from nearly 800 providers in its first two years, with scores appearing directly in the clinical record.³⁴

Integration matters because it changes what clinicians do with the data. In oncology, for instance, real-time symptom scores embedded in the chart have been used to trigger psychosocial referrals automatically. But integration alone is not enough. A randomized trial in primary care found that simply showing clinicians a visual display of their patients’ PROMIS symptom scores, without additional support or incentives, did not improve symptom outcomes compared to usual care.³⁵ Giving a doctor a number is not the same as giving them a workflow for acting on it. The systems that seem to work best pair the score with a decision point: if pain interference exceeds 60, offer a palliative care consult; if depression exceeds 65, schedule a mental health follow-up.

Health Literacy and Who Gets Left Behind

One of the less discussed challenges with any patient-reported questionnaire is whether patients can actually understand and complete it. In an orthopedic setting, patients with low health literacy had a 23% higher rate of not completing their questionnaires at all, and Black patients had a 25% higher non-completion rate compared to white patients. Among those who did complete forms, patients with lower health literacy were much less likely to use online patient portals, preferring in-clinic completion methods.³⁶

Even among patients who fully completed PROMIS CATs, about 11% did not fully understand all of the questions. Patients with inadequate or marginal health literacy were roughly three times more likely to not fully understand the content compared to patients with adequate health literacy.³⁷ This does not invalidate PROMIS, but it does mean that clinics need to think carefully about how they deliver the questionnaire. Offering in-person assistance, using simplified language where possible, and not relying exclusively on patient portals are all strategies for making sure scores represent real patient experience rather than just the experience of patients who happen to be comfortable with health questionnaires.

Screening for Cognitive Decline in Older Adults

One of the newer applications of PROMIS is cognitive screening during Medicare Annual Wellness Visits. The PROMIS Cognitive Screener, or PRO-CS, is a brief tool designed to flag older adults who may be at risk for mild cognitive impairment or subjective cognitive decline, potentially before a formal diagnosis would otherwise be made.³⁸ A large health system in Pennsylvania implemented an electronic version of the four-item PRO-CS into their wellness visit workflow, with over 5,700 patients completing it in about seven months. Providers noted it was useful for opening conversations about cognition, though some raised the limitation inherent in all self-report measures: people with early cognitive decline may not reliably assess their own thinking.³⁹ It is a screening tool, not a diagnostic one, but it represents a meaningful expansion of PROMIS beyond its original territory of pain and physical function.