Nonoperative treatments for sciatica: A pilot study for a randomized clinical trial

Gert Bronfort, Roni L. Evans, Alfred V. Anderson, Kurt P. Schellhas, Timothy A. Garvey, Ruth Ann Marks, Sarah Bittell

Research output: Contribution to journalArticlepeer-review

24 Scopus citations


Objectives: To assess the feasibility of patient recruitment, the ability of patients and clinicians to comply with study protocols, and the use of data collection instruments to collect cost-effectiveness data, and to obtain variability estimates for samplesize calculations for a full-scale trial. Study Design: Prospective, observer-blinded, pilot randomized clinical trial. Setting: Primary contact chiropractic and medical clinics. Patients: Ages 20 to 65 years, with low back-related radiating leg pain (sciatica). Outcome Measures: Self-report questionnaires were administered at baseline and 3 and 12 weeks after randomization. The measures included leg and back pain severity, frequency and bothersomeness of symptoms, leg/back disability, medication use, global improvement, satisfaction, and health care utilization. Interventions: Medical care, chiropractic care, and epidural steroid injections. Results: A total of 706 persons were screened by phone to determine initial eligibility. Of these, over 90% of those persons contacted did not meet the entrance criteria. The most common reason for disqualification was that the duration of the complaint was longer than 3 months. Twenty patients were randomized into the study. All 3 groups showed substantial improvements in the main patient-rated outcomes at the end of the 12-week intervention phase. For leg pain, back pain, frequency and bothersomeness of leg symptoms, and Roland-Morris disability score, the percent improvement varied from 50% to 84%, and the corresponding effect sizes ranged from 0.8 to 2.2. Bothersomeness of leg symptoms was the most responsive outcome associated with the largest magnitude of effect size. All within-group changes from baseline were statistically significant (P <. 01). No between-group comparisons were planned or performed because of the insufficient sample size and high risk of committing type I and type II errors. Conclusions: Pilot studies such as these are important for the determination of the feasibility of conducting costly, larger scale trials. Recruitment for a full-scale study of sciatica of 2 to 12 weeks duration is not feasible, given the methods used in this pilot study. Our results do indicate, however, that there are substantial numbers of patients with sciatica more chronic in nature who would be interested in participating in a similar study. In addition, collaboration with a local health maintenance organization would likely facilitate clinician referrals and optimize the recruitment process. Patient and provider compliance was high in the pilot study, which indicates that most study protocols are feasible. Additionally, we were able to collect complete outcomes data, including those regarding health care use. A suggested modification by investigators and outside consultants has resulted in the replacement of the medication group with a minimal intervention control group (ie, self-care advice). As a result, a second pilot study of patients with sciatica of more than 4 weeks duration has been planned before a full-scale trial is attempted.

Original languageEnglish (US)
Pages (from-to)536-544
Number of pages9
JournalJournal of manipulative and physiological therapeutics
Issue number8
StatePublished - 2000

Bibliographical note

Funding Information:
Supported by the Foundation for Chiropractic Education and Research, the Consortial Center for Chiropractic Research through the National Center for Complementary and Alternative Medicine, and the National Institute of Arthritis, Musculoskeletal and Skin Diseases of the National Institutes of Health.


  • Chiropractic Manipulation
  • Drug Therapy
  • Lumbar Spine
  • Sciatica


Dive into the research topics of 'Nonoperative treatments for sciatica: A pilot study for a randomized clinical trial'. Together they form a unique fingerprint.

Cite this