Can experienced physiotherapists identify which patients are likely to succeed with physical therapy treatment?

Chad E Cook; Thomas J Moore; Kenneth Learman; Christopher Showalter; Suzanne J Snodgrass

doi:10.1186/s40945-015-0003-z

Authors

Chad E Cook Department of Orthopaedics, Duke University, 2200 Main Street, 27705 Durham, NC, USA
Thomas J Moore Department of Physical and Occupational Therapy, Duke University Medical Center, Clinic 1E, Trent Drive and Erwin Road, 27710 Durham, NC, USA
Kenneth Learman Department of Physical Therapy, One University Plaza, 44555 Youngstown, Ohio, USA
Christopher Showalter Maitland-Australian, Physiotherapy Seminars, PO Box 1244, 11935 Cutchogue, NY, USA
Suzanne J Snodgrass Department of Physiotherapy, University Drive, 2308 Callaghan, NSW, Australia

DOI:

https://doi.org/10.1186/s40945-015-0003-z

Keywords:

Prognosis, Prediction, Physiotherapist, Musculoskeletal, Spine

Abstract

Background: The purpose of the study was to determine if clinician predicted prognosis is associated with patient outcomes. Methods: The study was a secondary analysis of data that were collected in 8 physiotherapy outpatient clinics. Nine physiotherapists with post-graduate training in manual therapy (mean 20.3 years of experience) were asked at baseline to project the outcome of the patients evaluated. In total, 112 patients with low back (74 %) or neck (26 %) pain were treated pragmatically with interventions consisting of manual therapy, strengthening, and patient-specific education. Outcomes measures consisted of percent change in disability (Oswestry or Neck Disability Index), self-reported rate of recovery (0–100 %), and percent change in pain (numerical pain rating scale). Hierarchical logistic regression determined potential factors (clinician predicted prognosis score (1–10) at baseline, dichotomised as poor (1–6) and good (7–10); symptom duration categorised as acute, subacute or chronic; same previous injury (yes/no); baseline pain and disability scores; within-session improvement at initial visit (yes/no); and presence of ≥ one psychological factor) associated with meaningful changes in each of the three outcomes at discharge (disability and pain > 50 % improvement, rate of recovery ≥82.5 % improvement). Results: Clinician predicted prognosis (OR 4.15, 95%CI = 1.31, 13.19, p = 0.02) and duration of symptoms (OR subacute 0.24, 95%CI = 0.07, 0.89, p = 0.03; chronic 0.21, 95%CI = 0.05, 0.90, p = 0.04) were associated with rate of recovery, whereas only clinician predicted prognosis was associated with disability improvement (OR 4.28, 95 % CI 1.37, 13.37, p = 0.01). No variables were associated with pain improvement. Conclusions: Clinician predicted prognosis is potentially valuable for patients, as a good predicted prognosis is associated with improvements in disability and rate of recovery.