More treatment. Better recovery? What the evidence actually shows.
Walking robots, extra spinal cord training, back-pain treatment and exercise for muscle loss—read through the lens of function.
Radu Physiatry editorial desk · About 12 minutes · AI-assisted, source-checked; not physician-reviewed
A rehabilitation program can be longer, more expensive and more technologically impressive without helping a person walk farther or live more independently. Four trials from 2026 bring that distinction into focus. They ask different questions, in different populations, but reward the same reading habit: look past the treatment’s name and ask what changed in daily life.
This inaugural edition of The Physiatry Daily pairs selected 2026 research with enduring clinical teaching. The studies below were published between March and September; they are a launch reading list, not a claim that all four appeared today. Future editions will separate newly discovered material from background review. The source desk records the journals, organizations and reference collections in our current coverage.
Today’s reading lens
A better treatment needs a better outcome.
Ask five questions of every headline: Who was studied? What was the comparison? Which outcome mattered most? How large was the difference? What remains uncertain?
A walking robot is an option, not a promise of better walking.
In a trial of 100 people after stroke, four weeks of robot-assisted walking plus conventional physiotherapy did not clearly outperform conventional physiotherapy on walking independence or basic daily activities. Both groups improved. The robot group reported greater perceived quality-of-life improvement, a different outcome from better walking.
For the patient: ask what the machine adds to your particular goal. A robotic session may be worthwhile, but its presence alone does not establish superiority. Read the trial record ↗
Clinical & board study notes
The co-primary measures were Functional Ambulation Categories and the Barthel Index: walking assistance and daily activity independence. Participants had subacute or chronic stroke; follow-up reached week 12. A secondary quality-of-life finding should not replace the primary result. Lack of demonstrated superiority is not proof of equivalence, and one device protocol cannot settle the value of all rehabilitation robots.
Evidence boundary A 100-person trial; one inpatient protocol. The primary outcomes do not establish a universal advantage for robots.
Adding twelve hours a week did not improve the main motor outcome.
A trial randomized 220 people with recent spinal cord injury and preserved motor function below the injury. One group received usual inpatient rehabilitation; the other also received 12 hours of task-specific and strength training weekly for ten weeks. The additional program did not significantly improve the primary motor score: the reported between-group difference was 0.93 points, with a 95% confidence interval from −1.63 to 3.48.
For the patient: this tests an added program, not whether rehabilitation itself helps. It is a reason to discuss the goal and burden of extra sessions. Read the trial record ↗
Clinical & board study notes
The endpoint was the 100-point International Standards motor score. Assessors were blinded. Four serious adverse events occurred, three in the intervention arm, including two deaths; the abstract does not establish causation.
Evidence boundary The finding concerns an added program in a specific recent-injury population.
03 / Chronic low back pain · adaptive randomized trial · June 2026
A statistically detectable advantage can still be small.
In 749 adults with chronic low back pain, physical therapy produced a greater improvement in function than cognitive behavioral therapy at ten weeks. The adjusted difference was 2.8 points on the 100-point Oswestry Disability Index. That was below the trial’s six-point minimum important difference. Pain intensity did not differ clearly between groups.
Nonresponders were randomized again to switch treatment or receive mindfulness. At 52 weeks, neither second-stage approach clearly won. For the patient: discuss a feasible treatment and a meaningful functional goal. Read the trial record ↗
Clinical & board study notes
This adaptive design tested pathways after assessing response. The function difference had a 96% confidence interval of 0.38–5.1. Statistical significance and clinical importance answer different questions.
Evidence boundary Participants were not blinded. Low treatment initiation, especially for CBT and nonresponders, limits interpretation.
04 / Aging & function · small randomized trial · September 8, 2026
For muscle loss, measure what a person can do.
A small study assigned 30 older adults with sarcopenia—loss of muscle strength and capacity—to an exercise program or usual routines. It combined resistance exercise and social activities, three 50-minute sessions weekly for 12 weeks. Between-group improvements favored exercise on the Short Physical Performance Battery, Berg Balance Scale and Mini Nutritional Assessment.
For the patient: strength, balance and everyday performance deserve attention alongside muscle size. The study is encouraging, but too small to identify the best program. Read the trial record ↗
Clinical & board study notes
Improvement from baseline does not prove superiority over a comparator. The abstract reported more within-group improvements than between-group advantages. Between-group effect sizes and confidence intervals were not supplied. This trial does not establish fewer falls or hospitalizations.
Evidence boundary Preliminary evidence. The group program’s exercise and social components were bundled, so their separate effects cannot be isolated.
The clinician’s notebook / established teaching
Two bedside lessons that remain worth revisiting.
When a nerve test is normal, what has actually been ruled out?
Radiculopathy means a problem involving a spinal nerve root. Sensory nerve conduction responses are often preserved because the lesion is usually proximal to the dorsal root ganglion, where sensory cell bodies sit. Needle electromyography evaluates motor axon involvement; a purely sensory problem may escape detection.
Open the localization pearl
Abnormalities in at least two muscles sharing a root but supplied by different peripheral nerves support root localization. Sampling, timing and the clinical examination matter. An isolated paraspinal abnormality deserves caution, and a normal needle examination does not exclude every radicular disorder. Use the full review before making a clinical inference.
A “normal” absolute pressure can hide a dangerous rise.
Autonomic dysreflexia is a potentially life-threatening blood pressure surge after spinal cord injury, typically at T6 or above. Bladder or bowel problems commonly trigger it. A substantial rise above a person’s baseline matters, even if the number does not initially look alarming.
Open the safety pearl
A sudden headache, flushing or sweating in a susceptible person should prompt immediate blood pressure assessment and urgent evaluation. Initial clinical steps include sitting the person upright, loosening restrictive clothing and seeking the trigger, often urinary. Bradycardia is not required; tachycardia can occur. Patients need a personalized emergency plan. This brief is not a medication or catheter-management protocol.
A stronger muscle, a faster walk and a return to work are related, but they are not interchangeable. The World Health Organization’s functioning framework separates body functions and structures from activities and participation, while recognizing the influence of the environment. In practical terms, a treatment can change an impairment without yet demonstrating a change in a person’s daily life. WHO ICF framework ↗
The outcome map in today’s four studies
Question
Measure
Interpretation to preserve
How much motor function is present after SCI?
International Standards motor score
Motor impairment; do not substitute it for independence.
How independently can someone walk after stroke?
Functional Ambulation Categories
Walking assistance; perceived quality of life answers another question.
How much does back pain limit function?
Oswestry Disability Index
A disability score; read magnitude alongside statistical uncertainty.
How well does an older adult perform?
SPPB and Berg Balance Scale
Performance and balance; an improvement alone does not prove fewer falls.
Measure names and trial interpretations: the four linked trial records above. For administration and measurement properties, consult the Rehabilitation Measures Database ↗.
Study question: a trial reports p < 0.05. Is the treatment change necessarily meaningful?
Answer: no. Statistical evidence against a null hypothesis does not determine how important the effect is to a patient. Read the effect, confidence interval, population and outcome together. A threshold helps interpret a group result; it is not a guarantee about each individual.
Learning desk / how the sources fit together
Use a journal for new evidence, a guideline for a recommendation and the board for the exam.
The ABPMR Part I page links the official examination outline. The continuing-certification pages govern longitudinal assessment and other requirements. Our 15 study domains are navigation categories; they do not replace the board’s current outline or establish examination weights.
PM&R KnowledgeNow supplies specialty topic reviews. AANEM’s clinical resources collect neuromuscular and electrodiagnostic guidance. The PVA collection adds SCI-specific practice guidelines. Each serves a different purpose from an individual trial; check the document’s date, scope and evidence before applying it.
Courses, journal clubs, podcasts and conference presentations can organize learning. They do not automatically confer continuing medical education credit, and a conference abstract is not the same as a completed peer-reviewed study. Textbooks and point-of-care subscriptions remain useful for background, but this issue does not claim access to their licensed chapters. The 88-source directory gives each source a purpose and an access boundary.
A few terms, translated
Read the fine print without losing the story.
Primary outcome
The main result a study was designed to test. Other findings help explain it but should not quietly replace it.
Confidence interval
A range expressing uncertainty around an estimate. Its width and whether it includes no difference both matter.
Comparator
The treatment or care against which the intervention was tested. “Usual care” can already be substantial.
Intention to treat
Analyzing people in the groups to which they were assigned, helping preserve the benefit of randomization.
Minimum important difference
A threshold used to interpret whether a change is meaningful. It depends on the measure, population and method.
Study summaries were checked against PubMed records and rendered abstracts. The four trial summaries use abstract-level evidence; full-text appraisal was not completed. The two clinical notebook reviews and the linked official education pages were also consulted. This is independent education, not an accredited CME activity or an official ABPMR publication.
The discovery window was 2026/09/01–2026/10/01. A journal index check can miss work not yet indexed. A retrieved landing page does not verify every linked publication, paid course, podcast or textbook chapter. Candidate counts are discovery records, not validated practice changes.
Educational, not medical advice. Discuss treatment choices with your clinician. Sources and evidence evolve; each dated edition preserves what was checked for that issue.
Correction: October 1, 2026: the stroke trial DOI was corrected to 10.3389/fneur.2026.1898475.