Hurwitz 2002.
| Methods | Study design: RCT Country: Not reported Number recruited: 1469 Number randomised: 681 | |
| Participants | Population: Patients were required to be health maintenance organisation members with the medical group chosen as their health care provider, sought medical care from a health provider on staff at one of the three study sites during the enrolment period, presented with a complaint of low‐back pain, had not received treatment for low‐back pain within the previous month, and were at least 18 years old. Settings: People presenting at three study (HMO) ambulatory care facilities Mean Age: Not reported. Participants were required to be over 18 years of age. 10% were under 30 years, 20% in 30 to 39 years, 20% in 40 to 49 years, 20% in 50 to 59 years, 15% in 60 to 69 years, 20% in 70 plus years. Very little difference between treatment groups. Work status: Almost 60% are employed full time. 8% employed part time. 25% are retired. Little difference across treatment groups. Pain duration: 47% of participants had been in pain for longer than 1 year, 26% had been in pain for less than 3 weeks, 17% had been in pain for 3 weeks to 3 months, and 12% had been in pain for 3 months to 1 year. Differences ranged by 8% across treatment groups. | |
| Interventions | Chiropractic 1: Patients assigned to chiropractic care only received spinal manipulation or another spinal‐adjusting technique, instruction in strengthening and flexibility exercises, instruction in proper back care. At 6 months average back pain‐related visits were 5.3 in the chiropractic only group. Chiropractors spent an average of 15 minutes with patients at each visit. Chiropractic 2: Patients assigned to the chiropractic care with physical modalities received care as described in ‘chiropractic care only’, as well as one or more of the following at the discretion of the chiropractor: heat or cold therapy, ultrasound and EMS. At 6 months average back pain‐related visits were 5.7 in the chiropractic plus physical modalities group. Chiropractors spent an average of 15 minutes with patients at each visit Comparison 1: Patients assigned to the medical care only group received one or more of the following at the discretion of the primary care provider: instruction in proper back care and strengthening and flexibility exercise, prescriptions for pain killers, muscle relaxants, anti‐inflammatory agents, other medications to reduce or eliminate pain or discomfort, and recommendations regarding bed rest, weight loss and physical activities. At 6 months average back pain‐related visits were 2.9 in the medical care only group. Medical providers spent an average of 15 minutes with patients at each visit. Comparison 2: Patients assigned to the medical care with physical therapy received medical care as described in ‘medical care only’, plus one or more of the following at the discretion of the physical therapist: heat therapy, cold therapy, ultrasound, electrical muscle stimulation, soft tissue and joint mobilisation, traction, supervised therapeutic exercise, strengthening and flexibility exercises. At 6 months average back pain‐related visits were 5.4 in the medical care plus physical modalities group. Physical therapy providers averaged 31 minutes per patient visit. 85% patients in the chiropractic groups received high velocity spinal manipulation. The physical modalities most often given to patients were heat therapy alone (28%), heat and EMS (25%), heat, EMS and ultrasound (23 %), and heat therapy and ultrasound (15%). 4% patients in the modalities group were not treated with any modalities and 13 % patients in the chiropractic‐only group received modalities. The most common intervention in the physical therapy group were heat or cold therapy (71%), supervised physical exercise (59.5%), ultrasound (45%), EMS(33.6%), and mobilisation (20%). Prescription pain medications (58.5%), muscle relaxants (48.5%) and non prescription pain medications (30%) were the most frequent interventions in the medical groups. Country of training: Not reported Years in practice: Not reported |
|
| Outcomes | Outcome measures at baseline and then follow‐up measurement after 2 weeks*, 6 weeks, 6 months, and 18 months:
|
|
| Notes | *At 2 weeks: low‐back pain severity, improvement, and related disability, cut‐down days and bed days attributed to low‐back pain, and use of over‐the‐counter and prescription medication for low‐back pain using questionnaires. Functional status was measured by Roland‐Morris Low‐Back Disability Questionnaires. Pain status was measured by repeat numerical rating scales and scales of global improvement. Health care use data were extracted from the organisations computerised health care systems. Telephone interview to determine patients' low‐back pain visits. Combined baseline SD values from control and intervention groups in this study were used to back transform treatment effects expressed as SMD into treatments effects on a 10 cm VAS and on the 24 point scale of the Roland Morris Disability Questionnaire. |
|
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Adequate sequence generation? | Low risk | "Study statistician ran computer program to generate randomised assignments in blocks of 12, stratified by site." |
| Allocation concealment? | Low risk | "Each treatment assignment was placed in a numbered security envelope. A separate series of sequentially numbered sealed envelopes was provided for each of the three sites." |
| Blinding? All outcomes ‐ patients? | High risk | Not explicitly stated but patients would have known whether they were visiting a chiropractor or medical practitioner |
| Blinding? All outcomes ‐ providers? | High risk | Not explicitly stated but practitioners would have known what treatment was being supplied |
| Blinding? All outcomes ‐ outcome assessors? | Unclear risk | Not reported |
| Incomplete outcome data addressed? All outcomes ‐ drop‐outs? | Low risk | "Completers were 99.7% in the first 6 weeks and 95.7% in the first 6 months" |
| Incomplete outcome data addressed? All outcomes ‐ ITT analysis? | Low risk | Intention‐to‐treat analyses were performed throughout |
| Free of selective reporting? | Unclear risk | Insufficient information provided |
| Free of other bias? | High risk | No sample size calculation. Interventions and treatment settings described well enough |
| Similarity of baseline characteristics? | Low risk | Relatively small differences between treatment groups in the baseline distributions of sociodemographic and health status variables. Minor differences with respect to low‐back pain severity and related disability, but these differences are clinically insignificant. |
| Co‐interventions avoided or similar? | High risk | Approximately 20% patients in the chiropractic group received concurrent medical care, and 7% of patients in the medical group received concurrent chiropractic care in the first 6 months. |
| Compliance acceptable in all groups? | High risk | Approximately 33% patients randomly assigned to medical care with physical therapy had no physical therapy visits |
| Timing of the outcome assessment similar in all groups? | Low risk | At two weeks, 6 weeks, 6 months and 18 months |