physical therapy does not always add to primary care

There have been 3 recent JAMA articles suggesting lack of efficacy of physical therapy in several settings.

1. in  patients with acute low back pain (see low back pain and early PT JAMA2015 indropbox, or JAMA. 2015;314(14):1459-1467). details:

​--220 patients without low back pain (LBP) treatment for past 6 months: median age 37,Oswestry Disability Index (ODI, a validated 10-item measure of function) of >=20, symptom duration <16 days and no symptoms distal to the knee in past 72 hours.

--randomized to early PT (spinal manipulation and exercise) with 4 PT sessions over 3 weeks, beginning within 72 hours of enrollment vs no additional interventions for 4 weeks (all participants were given a copy of The Back Book, which was reviewed with them). outcomes assessed at 1 year.

--results:

    --PT better than usual care at 3 month (ODI score decreased from 41.3 to 6.6, vs 40.9 to 9.8 in usual care group). But, a change of ODI scale of 6 points is considered clinically significant, so this slight difference was not clinically important.  Similarly at the 4 week assessment (ODI score decreased to 14.5 vs 11.1). the change at 1 year was also not statistically different

    --no difference in pain scores

so, this result is consistent with guidelines not to refer early to PT in the first few weeks since most people recover quickly anyway, though some observational studies do report greater risk of invasive procedures when PT is delayed beyond 2-4 weeks. But it is really important that this study not be interpreted as meaning that we in primary care should do nothing for those with acute low back pain. It is abundantly clear from the literature that people do better and more quickly (including returning to work sooner) if they return to their usual activities as quickly as they can, even with some pain (for example, it is important to reassure patients that ambulating with pain does not lead to permanent injury/chronic back pain).

2. in patients with hip arthritis (see hip arthritis not benefit from PT JAMA 2015​ indropbox, or JAMA. 2014;311(19):1987). details:

--102 community volunteers with hip pain (scoring >40 on visual analog scale, or VAS, which goes from 0-100) and xray-confirmed hip osteoarthritis.

--randomized to PT (10 treatment sessions over 12 weeks, with education/advice, manual therapy, home exercise, and gait aid if needed) vs sham treatment (inactive ultrasound and inert gel);  for the next  24 weeks after treatment, the active group continued unsupervised home exercise while the sham group self-applied the gel3x/week

--results:

    --after 13 weeks, VAS went from 58.8 to 40.1 with PT and from 58.0 to 35.2 with sham treatment (nonsignificantly favoring sham treatment), though both groups had statistically significant improvement in pain during the course of the study.

    --no difference in function scores.

    ​--similar results after 36 weeks

    --med usage was similar (mostly acetaminophen in 50%, NSAIDs in 20%, glucosamine/chondroitin in 40%), and no difference in co-interventions (other PT, exercise, hydrotherapy, myotherapy)

    --19 patients (41%) in the active group had 26 mild adverse effects and 7 (14%) of the sham group had 9 mild adverse effects

so, interesting that PT did not help much over sham treatment --  both really helped patients with pretty painful hip OA (placebo effect). In this case the sham/placebo treatment involved lots of empathy, listening, providing reassurance, encouragement even without obviously beneficial medical interventions. so, this study, to me, mostly reinforces the therapeutic effect of a caring primary care provider and perhaps additional supportive staff.

3. in patients needing rehab after immobilization for an ankle fracture (see JAMA 2015 Oct 6; 314:1376). details:

--214 patients (mean age 48 for men and 60 for women, 45% men, 65% of fractures from falls, 45% considered "more severe" fractures, 45% had open reduction/internal fixation) randomized on the day after immobilization removal to rehabilitation vs advice, and followed at 1 month, 3 months and 6 months, assessing activity limitation and quality of life

--interventions: rehab program was supervised exercise program (individually tailored program by PT, including ankle mobility/strengthening, stepping exercises and weight-bearing and balancing. suggested protocol was 2 sessions in first week, then a single session weeks 2,3,4) and advice (given to all participants by PT in a single session of self-management advice after removal of immobilization, emphasizing non-weight bearing ankle movements, use of ice/compression/elevation for pain and swelling relief, and encouragement to walk as tolerated, ) vs just the advice

--results:

    --mean activity score for advice-only group increased from 30.1 to 64.3; for rehab group increased from 30.2 to 64.3 at 3 months (scale 0-80, higher score better)

    --mean quality of life score for advice-only group increased from 0.51 to 0.85; for rehab group increased from 0.54 to 0.85 at 3 months (scale 0-1, higher score better)

    --similar lack of benefit was recorded at the 1 month and 6 month evaluations

    ​--treatment effects were not moderated by fracture severity, age, sex.
    --but 36% of the participants in the advice group did receive PT out-of-trial (as did 14% in the rehab group!!), though modeling with propensity score matching did not undercut the overall conclusion

so, it could be that there was no benefit for the PT group in part because there is typically such a rapid early improvement anyway after the immobilization that, as with the acute back pain, additional PT would not help. and this study found that even in older women, who more often have osteoporosis-related fractures, there was  no additional benefit from  a more rigorous program (comparing those >50yo vs <50yo). but perhaps PT might be very useful in those who do not get adequate recovery 6-9 months later...

The reason I bring up these 3 articles is not to undercut the general value of PT. My concern is that sometimes providers may not spend much time talking about the importance of exercise for stretching/strengthening for an array of musculoskeletal complaints and simply refer the patient to physical therapy. And my sense is that often patients do not go to PT (very time-consuming for them/ they are unable to get out of work or chores to go 2-3x/week for several weeks, hard to get there, they do not understand its utility or have fears it will hurt them to do the PT….). and, in the control groups in the above studies who all demonstrated significant clinical improvement, there was information from the study on non-PT care and/or empathy and reassurance. So, I think these studies actually reinforce the importance of the primary care relationship, both in therapeutic interventions and emotional support etc, and, as noted above, in some instances this might work as well as rigorous PT programs. (and spending the time with the patient to review exercises,etc, further reinforces the therapeutic primary care relationship).