THE BACK PAGE RESEARCH REVIEW Value of psychosocial factors in predicting patient outcomes BY SHAWN THISTLE STUDY TITLE: Adding psychosocial factors does not improve predictive models in patients with spinal pain enough to warrant extensive screening for them at baseline AUTHORS: Ailliet L, Rubinstein S, Hoekstra T, et al. PUBLICATION INFORMATION: Physical Therapy 2016; 96: 1179-1189. I t is widely accepted that low back pain (LBP) and neck pain (NP) are best understood as biopsycho-social phenomena. Thus, baseline screening for psychosocial charac-teristics in spinal pain patients, along with the conventional medical exami-nation and history, has been empha-sized in recent years. The extent psychological or social factors are useful in predicting out-comes in NP and LBP patients beyond what can be gleaned from a focused history and examination is uncertain. The purpose of this study was to determine whether psychological or social factors add predictive value to biomedical factors known to be associ-ated with a better or worse outcome in NP or LBP patients treated by chiropractors. Results Neck pain patients were found to be 73.8 per cent, 70.1 per cent, and 63.9 per cent much improved or completely recovered at three, six, and 12 months, respectively, although only 50.6 per cent of them experienced lasting recovery. Patients with low back pain were 66.7 per cent, 73.2 per cent, and 67.9 per cent much improved or completely recovered at three, six, and 12 months, respectively, but only 51.2 per cent of them experienced lasting recovery. There were differences between NP patients in the “lasting recovery” and “nonlasting recovery” groups. In the “lasting recovery” group, there were: twice as many acute cases with a dura-tion of less than six weeks; less previous treatments and imaging; and half the number of markedly elevated somati-zation scores. Differences between LBP patients in the “lasting recovery” and “nonlasting recovery” groups were also observed. In the “lasting recovery” group: the duration of complaints was shorter (more than 50 per cent of acute cases with a duration of less than six weeks); more than twice as many patients were treated for a first episode of LBP; pa-tients had less previous treatment and imaging; and fewer patients had mark-edly elevated scores on the psycholog-ical subdomains distress, depression, fear and somatization. Univariable analyses revealed that occupational status, body mass index, duration of complaints, and previous treatment were predictors of NP. The percentage correctly predicted by the model was 65 per cent. “Participant expectations” was the only psychosocial variable that was a statistically significant predictor of lasting recovery. After adding partici-pant expectations to the model, the percentage correctly predicted im-proved slightly, from 65 per cent to 66.1 per cent. Predictors in the final model for LBP included country of residence, pain patterns, age, duration of complaints, and previous imaging. The percentage correctly predicted by the model was 68.6 per cent. “Somatization” was the only psycho-social variable that was a significant predictor of lasting recovery. When somatization was added to the model, the percentage correctly predicted improved slightly, from 68.6 per cent to 69.5 per cent. However, the variable “pain patterns” was no longer a signif-icant predictor in the model. Application, conclusions The addition of psychosocial variables did not substantially improve the model fit indexes for predicting NP and LBP outcomes. The authors thought the main reason for this was that only a small proportion of patients had high scores on psycho-social variables. Thus, this study showed that adding psychosocial factors hardly changed the participating chiropractors’ ability to predict recovery for NP and LBP patients, which is in agreement with previous studies. The study’s main conclusion was that psychosocial vari-ables provided very little benefit in predicting outcome in NP or LBP patients who sought chiropractic care, so chiropractors need not screen exten-sively for them at baseline. This study also demonstrated the differences that appear when an out-come is measured at a single point in time versus repeated measurements over time. For NP, 73.8 per cent of patients reported that they were much improved or completely recovered at three months, but only 50.6 per cent of them had lasting recovery. Likewise, 66.7 per cent of LBP patients reported they were much improved or com-pletely recovered at three months, but only 51.2 per cent experienced lasting recovery. www.canadianchiropractor.ca DR. SHAWN THISTLE is the founder and CEO of RRS Education (rrseducation.com), which helps busy clinicians integrate current research evidence rationally into practice. He also maintains a practice in Toronto, lectures at CMCC and provides chiropractic medicolegal consulting services. Reach him at: [email protected] 34 Canadian Chiropractor October 2016
Chiropractic + Naturopathic Doctor October 2016: Page 34
