The Agency for Health Care Policy and Research (AHCPR) guidelines for the management of acute low back pain recommend obtaining radiographs in the evaluation of certain patients. Specifically, the recommendations for use of radiographs relate to the presence of potential “red flags” for spinal fracture, cancer or infection. Red flags for spinal fracture include a history of major trauma, minor trauma or strenuous lifting in an older patient or in a patient with possible osteoporosis, prolonged steroid use, the presence of osteoporosis and age greater than 70 years. Red flags for tumor or infection include a history of cancer, age under 20 or over 50 years, constitutional symptoms and back pain that is worse in the supine position or severe at night. Suarez-Almazor and colleagues performed a retrospective study to examine the potential impact of the recommendations for radiographs among patients with back pain who visit family physicians.
Four family practice clinics in Edmonton, Alberta, participated in the study. Adult patients were included if they presented with a complaint of low back pain and had had no similar complaint in the past year. Patients were excluded if the complaint was ascribed to a visceral condition or if the patient was pregnant or had ankylosing spondylitis. Medical records were reviewed to determine whether lumbar radiographs were obtained and to determine what, if any, red flags were present. Follow-up for at least two years established whether vertebral fracture, tumor or infection was subsequently diagnosed.
A total of 963 patients met AHCPR guideline criteria for acute low back pain. Lumbar radiographs were obtained at the first visit in 127 patients (13.2 percent). The radiographs revealed degenerative changes in 64 of the 127 patients (50.4 percent), and findings were normal in 44 of these patients (34.6 percent). Fracture was identified in four patients and possible metastatic disease was found in one patient.
AHCPR guidelines would have required 44 percent of the patients to have radiographs. At least one AHCPR criterion for radiographic examination was present in 346 patients who did not undergo radiographic examination, indicating underutilization according to the guidelines. Thirteen patients had active cancer during the follow-up period, but spinal infiltrates were found in only three of these patients.
The authors conclude that adherence to the AHCPR guidelines would cause overutilization of radiographs, with a concomitant increase in health care costs. Even if all of the patients with subsequent serious disease had been identified through the use of radiographs at the initial visit, it is not clear that the outcome of the disease would have changed. The authors state that adherence to the age criterion would have substantially increased utilization of radiographs and that the patient's age had a low specificity. They recommend that age not be used to screen patients in the absence of other red flags. Deferring radiographs for two to three weeks would probably not be harmful unless the likelihood of serious disease is high.