The 1HP Clinical Outcomes Report 2026

Every member gets an exercise program that raises the weight or reps on its own after sessions they handle well, and lowers them after a flare-up. Members log their activity minutes, pain and irritability each day, and repeat a short function survey over time.

In Power Leveling, each member is also paired with a 1HP expert who runs an onboarding consultation, holds regular 1-on-1 coaching check-ins, adjusts their exercise load, and teaches a pain science curriculum. Power Leveling is an education and coaching program. It is not physical therapy, occupational therapy or medical care, and it does not create a provider-patient relationship. In the Wrist Pain Troubleshooter, members use the same app on their own.

Three experts coached Power Leveling members: two physical therapists, Matthew Hwu, PT, DPT, OCS, CSCS, and Elliot Smithson, PT, DPT, MS, ATC, EMT, and an occupational therapist, Brett Becker, MS, OTR/L, ACE-CPT, ACE-GFI. We estimate they spent an average of 227 minutes with each member over the period studied, about 3.8 hours, counting 30 minutes per recorded visit. Members chose their own program. Nobody was assigned at random, which matters for how we read every comparison in this report.

We measure function with the QuickDASH, an 11-question survey about how hard everyday tasks are with your arm, wrist and hand [1]. It scores from 0 (no difficulty) to 100 (unable to do most tasks). Members started at an average of 32.6, and the two programs started at almost the same level: 32.9 in Power Leveling and 32.1 in the Troubleshooter.

The groups were not identical in other ways. Power Leveling members were more often men (90% vs 77%) and gamers (59% vs 37%), and fewer had pain in only one region (4% vs 16%). Our adjusted comparisons in Chapter 03 account for these differences where the data allows.

Among members who completed a follow-up survey at least 28 days after starting, Power Leveling members' QuickDASH fell from 34.4 to 18.0 on average: a 16.4-point improvement, or 46% (n = 108). Troubleshooter members' scores fell from 33.8 to 26.3: 7.6 points, or 13% (n = 39).

The comparison after adjusting for differences between the groups

Because members chose their program, we adjusted for starting score using a standard method for before-and-after data called ANCOVA [6]. Power Leveling members improved by 8.5 more points (95% CI 4.1 to 13.0; p < 0.001). Adding age, sex, symptom duration, nerve symptoms, number of pain regions, enrollment period and follow-up length gave a difference of 11.0 points.

At a fixed 12-week check-in, Power Leveling members had improved by 15.1 points (42%), and 69% had reached a meaningful improvement (n = 75). Only 13 Troubleshooter members completed that check-in: they had improved by 3.5 points (10%). That small number makes the 12-week Troubleshooter figure unreliable on its own.

Change came over weeks, not days. At two weeks the average Power Leveling member had improved by about 6 points. By the 4-week check-in, they had improved by 10.3 points on average, and 55% had already reached a meaningful improvement (n = 89). For Troubleshooter members at 4 weeks, the figures were 5.1 points and 30% (n = 23).

Pain changed less than function

Members' worst pain during activity fell from 4.0 to 3.1 out of 10 on average (n = 97), and 38% reduced it by 2 points or more. That is a smaller change than in function. In practice, most members were doing much more with their hands while still feeling some pain, which is what our approach expects: rebuilding capacity and confidence comes first, and pain tends to settle more slowly. A systematic review of pain education trials found it can reduce pain and disability in musculoskeletal pain [8].

A survey score shows how hard daily tasks feel. It does not show what someone went back to doing. For that, we reviewed all 1,157 call notes for the 106 Power Leveling members who had them. An AI tool generated each note from the transcript of a recorded call between the member and their expert. The experts did not add to or edit them. For each member we recorded what their symptoms had stopped or limited when they joined, and what they were doing at their latest note.

Fourteen members had fewer than three weeks of notes, so the counts in this chapter use the other 92. Their notes covered a median of 15 weeks (range 3 to 40).

Before and after, in the notes

Where the notes gave numbers, the change was often large. These are individual members, chosen to show the range of activities, not typical results.

In members' words

The notes often record what members said on their calls. These statements are lightly paraphrased to protect privacy.

What the notes add to the surveys

Function moved before pain. Many members were doing far more while their pain scores barely changed, which matches the survey results in Chapter 05. Judged on pain alone, several of the biggest functional gains would look like failures.

Fear dropped as people did more. The notes describe members taking off braces, stopping voice dictation, returning to the game or task they feared most, and handling a flare-up on their own instead of resting for days. The 79 of 92 members whose notes show less fear match the 87% who improved on the fear-avoidance survey.

Gaming came back first and fullest. Music, sport and full workdays came back last. Several members said their gaming was now limited by free time, not pain. The final step, such as a full day of nonstop typing, an instrument, or heavy gym work, was often still in progress at the latest note.

Partial results usually had causes outside the hands. Neck, back and shoulder problems, new injuries, illness, and life events such as moving, job loss, caregiving and grief held back many of the 40 members with partial gains. Of the three members with little change, one had pain in many areas with cycles of overdoing it on good days, one met their strength targets without lasting change in symptoms, and one rarely did the program.

Long-standing pain is where the two programs differed most. Among members in pain for more than a year, Power Leveling members improved by 16.3 points and 69% reached a meaningful improvement (53 of 77). Troubleshooter members improved by 3.6 points, and 25% did (n = 28).

People who started with mild symptoms improved the least, partly because they had less room to improve. People who started with severe symptoms improved the most, but some of that is expected anyway: people tend to join when their symptoms are at their worst, and scores drift back toward normal over time whatever happens. This is called regression to the mean [9]. Nerve symptoms such as numbness and tingling did not predict a worse result.

We looked at many subgroups, so a few differences will look important by chance. Treat this chapter as a guide to questions, not answers.

Within Power Leveling, how often a member exercised did not predict how much they improved. Nearly everyone with an expert stayed consistent, so the useful difference was between the programs: support kept people going. What did stand out was progression. Members whose exercise load never progressed improved by 1.5 points on average, and 18% reached a meaningful improvement (n = 22). Members with six or more automatic progressions improved by 17.3 points, and 72% did (n = 113). The link was clear inside the self-guided program. Within Power Leveling it was weak, because most members progressed steadily.

Not everyone returned a follow-up survey: 74% of Power Leveling members did (108 of 146) and 51% of Troubleshooter members did (39 of 76). People who are improving may be more likely to keep reporting. The switch below shows how the main result changes depending on what we assume about the people we did not hear from.

Limits to keep in mind

Design

A retrospective observational cohort study of data collected routinely in the 1HP app, reported following the STROBE guidelines for observational studies [10]. Data were extracted on October 31, 2026, covering members who enrolled from August 2025 onward. Every member of either lifetime program was included; staff and test accounts were excluded.

Outcomes

The main outcome was the change in QuickDASH from baseline to the latest follow-up at least 28 days later, since shorter gaps mostly measure day-to-day noise. Fixed check-ins at 4, 8, 12 and 26 weeks used the follow-up closest to that week, within 14 days (28 days at 26 weeks). Secondary outcomes were the QuickDASH work and sport or performing-arts modules, the Fear-Avoidance Components Scale [7], and daily pain and irritability ratings from the app.

Analysis

Changes within each program used paired t-tests. Programs were compared with Welch's t-test and with ANCOVA, which adjusts the follow-up score for the starting score [6], then with regression models adding member characteristics. Time to a meaningful improvement used Kaplan–Meier estimates and the log-rank test. Missing follow-ups were explored by counting missing members as unchanged and with best- and worst-case assumptions. All tests were two-sided at α = 0.05 with no adjustment for multiple comparisons, so secondary and subgroup results are exploratory.

Review of call notes

For Chapter 06 we included every call note for Power Leveling members: onboarding consultations, coaching check-ins and other session notes, 1,157 notes for 106 members. Each note was generated by an AI tool from the transcript of a recorded call, with no content added by the experts. Billing notes, member-written notes, notes from the 1HP Clinical Brain chat tool and notes under 80 characters were excluded. Each note was de-identified with automated removal of names, contact details and dates. Because notes ran to about 17,000 characters on average, the review used the opening summary of each note in full and longer extracts of each member's first and latest notes.

An AI language model coded each member against a fixed codebook: activities limited at the start, activities regained, overall functional change, return to work or study, flare-ups, fear and confidence, and other gains. Overall change was coded from function, not pain: back to most or all key activities, clear gains but still limited, little change, worse, unclear, or too early (under three weeks of notes). Codes were checked against the full original notes for a sample of members. Member statements were paraphrased, and examples were chosen to show the range of activities.

Data handling, funding and conflicts of interest

Survey and app records were de-identified on the server before analysis. Call notes were de-identified before review, and no note text is published in this report. De-identified member-level data and analysis definitions are available from the authors on reasonable request. The analysis was funded internally by 1 Healthpoint Inc. Elliot Smithson and Matthew Hwu are co-founders of 1 Healthpoint Inc., which sells both programs, and both coach members in Power Leveling. Brett Becker, the third Power Leveling expert, works for 1 Healthpoint Inc. and coached some of the members whose calls are reviewed in Chapter 06.

Context

Repetitive strain injury covers many upper-limb conditions, and strong evidence for any single treatment is limited, though exercise is beneficial for non-specific upper-limb disorders [11]. Digital programs for musculoskeletal pain have shown benefits for function [12, 13], mostly for back and knee pain. This report adds data on upper-limb pain in people who use computers and games heavily.

How to cite this report

Smithson E, Hwu M. The 1HP Clinical Outcomes Report 2026. 1 Healthpoint Inc.; November 2026.

References