Osteopathic Manual Treatment and Ultrasound Therapy for Chronic Low Back Pain

A Randomized Controlled Trial

John C. Licciardone, DO, MS, MBA; Dennis E. Minotti, DO; Robert J. Gatchel, PhD; Cathleen M. Kearns, BA; Karan P. Singh, PhD


Ann Fam Med. 2013;11(2):122-129. 

In This Article


Design Overview

The OSTEOPATHIC Trial used a randomized, double-blind, sham-controlled, 2 × 2 factorial design to study OMT and UST.[15] The study was designed to compare main effects (OMT vs sham OMT, and UST vs sham UST) (Figure 1). It was approved by the Institutional Review Board at the University of North Texas Health Science Center.

Figure 1.

Allocation of patients to osteopathic manual treatment and ultrasound therapy interventions using a 2 × 2 factorial design.

Setting and Patients

We recruited patients in Dallas-Fort Worth, Texas, from August 2006 to September 2010 through newspaper advertisements, community agencies, and medical clinics, including those affiliated with the group practice of the University of North Texas Health Science Center, but excluding clinics that provided OMT specialty services. Adult (aged 21 to 69 years) nonpregnant individuals with low back pain for at least 3 months were telephonically screened for the following exclusion criteria: red flag conditions (cancer, spinal osteomyelitis, spinal fracture, herniated disc, ankylosing spondylitis, or cauda equina syndrome); low back surgery in the past year; workers' compensation benefits in the past 3 months; ongoing litigation involving back problems; angina or congestive heart failure symptoms with minimal activity, history of a stroke, or transient ischemic attack in the past year; implanted biomedical devices (such as cardiac pacemakers or artificial joints); active bleeding or infection in the lower back, or other conditions impeding protocol implementation; use of corticosteroids in the past month; or use of manual treatment (OMT or manual therapies delivered by chiropractors or physical therapists) or UST in the past 3 months or more than 3 times in the past year. Candidates whose screening was successful by telephone received a clinical screening to exclude those with a high probability of lumbar radiculopathy, a specific cause of low back pain and a relative contraindication to OMT. Clinical screening involved testing for ankle dorsiflexion weakness, great toe extensor weakness, impaired ankle reflexes, loss of light touch sensation in the medial, dorsal, and lateral aspects of the foot, ipsilateral straight leg raising, and crossed straight leg raising.[16]

Randomization, Allocation Concealment, and Blinding

Patients were allocated to OMT + UST, OMT + sham UST, sham OMT + UST, or sham OMT + sham UST at the central randomization site based on a computer program that generated pseudorandom numbers. Patients were secondarily allocated to type of physician (faculty physician, predoctoral fellow, or resident) using stratified randomization. Assignments were then conveyed directly to the physicians using numbered, opaque sealed envelopes, which were subsequently placed in secured, segregated treatment files. Patients and outcome assessors remained unaware of group assignments at randomization.


Treatments were scheduled at weeks 0, 1, 2, 4, 6, and 8 using 15 different physicians. We maintained the same physician at recurring treatment sessions for a given patient unless there was a scheduling conflict. Patients could self-initiate low back pain co-treatments, such as nonprescription drugs and complementary and alternative medicine therapies. Patients could also independently receive low back pain usual care (any co-treatments except OMT, other manual therapies, or UST) at any time from physicians not associated with the study. Co-treatments were documented at 4-week intervals throughout the study.

Active and Sham Osteopathic Manual Treatment

The OMT techniques were delivered after a standard diagnostic evaluation[17] at each treatment session. The lumbosacral, iliac, and pubic regions were targeted using high-velocity, low-amplitude thrusts; moderate-velocity, moderate-amplitude thrusts; soft tissue stretching, kneading, and pressure; myofascial stretching and release; positional treatment of myofascial tender points; and patient's isometric muscle activation against the physician's unyielding and equal counter-force. Time permitting, optional techniques[18] could be used if the physician judged 1 or more of the 6 designated techniques to be contraindicated or ineffective.

Sham OMT was aimed at the same anatomical regions as active OMT. Sham OMT involved hand contact, active and passive range of motion, and techniques that simulated OMT but that used such maneuvers as light touch, improper patient positioning, purposely misdirected movements, and diminished physician force. Similar methods achieved a robust placebo response[13] when compared with other placebo treatments for pain.[19] Our methods have been adopted by others to deliver sham manipulation.[20]

Active and Sham Ultrasound Therapy

The UST intervention was delivered after the OMT intervention, using the Sonicator 730 (Mettler Electronics Corp), with a 10 cm2 applicator at an intensity of 1.2 W/cm2 and frequency of 1 MHz in continuous mode. Conductivity gel was used to enhance absorption and produce deep muscle thermal effects.[21] About 150 to 200 cm2 of the lower back were treated. Sham UST was delivered in the same manner at a subtherapeutic intensity (0.1 W/cm2).[22]

Treatment Fidelity and Adherence

Training for physicians delivering the treatment was conducted at regular intervals using strategies to enhance protocol implementation and treatment fidelity, including provision of sham treatments.[23] A standard form was used at each treatment session to ensure consistency in delivery of active or sham OMT for 15 minutes, and of active or sham UST for 10 minutes. Both OMT and UST interventions were delivered by the same physician during a treatment session. Treatment adherence and reported pain levels were used as surrogate measures of patient blinding.


Primary Low Back Pain Outcomes. The current level of low back pain was measured before each treatment and at week 12 using a 100-mm visual analog scale. Primary outcomes were based on the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) consensus statement recommendations for moderate (30% or greater pain reduction) and substantial (50% or greater pain reduction) improvement.[24] Such reductions are highly sensitive and specific in predicting global impression of change in chronic pain patients[25] and provide tangible evidence for clinical applications.[26]

Secondary Outcomes. Patient-based secondary outcomes[27] were measured at baseline and at weeks 4, 8, and 12, using the Roland-Morris Disability Questionnaire (RMDQ),[28] Medical Outcomes Study Short Form-36 Health Survey general health scale (SF-36 GH),[29] number of lost work days in the past 4 weeks because of low back pain, and satisfaction with back care on a 5-point Likert scale. We measured 8 low back pain co-treatments reported by patients as being self-initiated or received from independent, community-based clinicians.

Safety Monitoring

Blinded research personnel assessed patients for contraindications to continued participation or adverse events at each encounter. An independent safety officer reviewed all reported contraindications and adverse events to identify any serious adverse events, defined as deaths, life-threatening situations, hospitalizations, severe or permanent disability, or other important medical events. The safety officer also assessed causality of serious adverse events in relation to study interventions.

Sample Size

The planned sample size of 488 was designed for a statistical power of 82% or greater in testing OMT vs sham OMT main effects (standardized mean difference of 0.264 for OMT vs control treatments,[30] corresponding to a between-group difference of 6.6 mm on a 100-mm visual analog scale, with a standard deviation of 25 mm[31]). In September 2010, under supervision of the Data and Safety Monitoring Board, an unplanned interim analysis was performed using visual analog scale change scores over 12 weeks that were measured through the provisional study end date of June 2010. This analysis was undertaken to determine whether the study could be terminated with 455 (93%) of the planned number of patients having enrolled in the study. Alternatively, additional sources of funding would have been required for further recruitment and study completion. The results of this analysis indicated that the change scores for OMT patients were significantly better than for sham OMT patients (P = .003). This finding crossed the O'Brien-Fleming stopping boundary (P = .005).[32] Consequently, study recruitment was terminated.

Statistical Analysis

Low Back Pain Outcomes. We observed a bimodal distribution of visual analog scale change scores and corresponding marginal test for normality (P = .08). Consequently, we used the median and interquartile range (IQR) as descriptive measures, and the Mann-Whitney test and contingency table methods for analysis of low back pain outcomes. Responder analysis was used to assess treatment effects at week 12, with a focus on moderate and substantial improvements in low back pain.[24] Response ratios (RRs) and 95% confidence intervals for active vs sham treatments were used to interpret treatment effects. Significant results were considered clinically relevant if they met the Cochrane Back Review Group criteria for medium (1.25 ≤RR ≤2) or large (RR >2) effect sizes.[33]

To test for statistical interaction between OMT and UST, we performed repeated measures analysis of variance on the ranked visual analog scale pain scores for each protocol visit, with the ranked baseline pain score as a covariate. Rothman's T statistic[34] was used to test for statistical interaction between OMT and UST based on moderate and substantial improvements in low back pain with each intervention. Any significant departure from T = 0 was indicative of interaction.

Secondary Outcomes. The distributions of secondary outcome measures also mandated use of nonparametric methods. The Mann-Whitney test was used to analyze RMDQ and SF-36 GH scores at weeks 4, 8, and 12. We dichotomized disability specific to low back pain (0 vs ≥1 lost work days in the past 4 weeks), satisfaction with back care (very satisfied vs any other response), and use of low back pain co-treatments (no use vs any use during the study).

Other Statistical Methods. Patient flow, treatment adherence, and safety were assessed by contingency table methods. Hypothesis testing was by intention to treat, with a 2-sided α = .05. Missing data were generally imputed using the last observation carried forward. For the multi-item RMDQ and SF-36 GH outcomes, however, we preferentially used responses acquired during a given encounter whenever possible to impute missing data for that encounter.[29] Sensitivity analysis was conducted to assess the impact of missing data and robustness of our imputation methods. First, we conducted per-protocol analysis using only the 362 patients with complete data. Second, we conducted an alternate analysis in which any patient with 1 or more missed treatments was considered to be a nonresponder. We did not specify a priori subgroup analyses because of concerns about statistical power and confounding in such analyses.[35] Statistical analyses were performed with SPSS 17.0.3 (SPSS Inc), using Epi Info 6.04d (Centers for Disease Control and Prevention) for low back pain treatment effects.