SPECT more effective than MRI at detecting mechanical back pain in the short term

When symptoms last over 6 weeks, however, MRI may be the most efficient test.

CHICAGO – When children present with back pain, physicians want to expose them to the least amount of medical testing without missing a diagnosis.

Researchers at Children’s Hospital of Philadelphia (CHOP) found that bone scanning using single-photon emission computed tomography (SPECT) – not MRI – is superior in clearing those children with non-neurologic back pain at less than 6 weeks of symptom duration.

They also found that “absent hyperextension pain, negative X-ray and negative MRI at any point in time is 100% predictive of mechanical back pain,” said Joshua D. Auerbach, MD, a resident in orthopedic surgery at the University of Pennsylvania.

The researchers, including senior author John M. Flynn, MD, of CHOP, used their findings to develop an evidence-based imaging algorithm for diagnosing mechanical back pain. When further clinical testing is necessary, the algorithm recommends that surgeons test with bone scan if symptom duration is less than 6 weeks, and MRI if symptom duration is longer than 6 weeks.

“In either case, the predictive value is 1 to get you a diagnosis of mechanical back pain,” Auerbach said at the American Academy of Orthopaedic Surgeons 73rd Annual Meeting.

The researchers found 873 children who presented with back pain over 17 months at CHOP. Because the study was based on the “non-complicated back pain patient,” they eliminated any patients with previous back pain diagnoses, scoliosis, previous spine surgery and any “classic red flags,” including constitutional and neurologic symptoms and night pain, Auerbach said.

Researchers isolated 100 patients who met their criteria – those with the deluxe back pain workup, including X-ray, bone scan with SPECT and MRI.

This patient group included 38 boys and 62 girls with an average age of 13 years. All had an average 44 weeks duration of pain. Diagnoses included: 57 children with mechanical back pain, 21 with spondylolysis, six with spondylolisthesis, four with disc herniation and 12 patients with other pathologies.

The researchers split the patients into two groups. Group I, the mechanical back pain group, included children with all negative tests. Group II, the diagnosis group, included children with a positive diagnosis, indicated by a positive imaging exam and symptoms. Next, the researchers performed statistical analysis using multiple linear regression and chi-squared tests to determine distinguishing factors between the two groups.

Predictive value findings

Auerbach said they focused on the negative predictive value to determine the best “clearance” test. The researchers’ main findings included the following:

• Patients with a negative hyperextension test and negative X-ray had a predictive value of 0.81 (P=.0085).
• Patients with a negative hyperextension test, a negative X-ray and a negative bone scan had a negative predictive value of 0.94 (P<.0001).
• Patients with no painful hyperextension, a negative X-ray and a negative MRI had a negative predictive value of 1.
• A negative bone scan had a significantly less predictive value of 0.72 when symptom duration was more than six weeks (P=.015).
• If symptom duration was less than 6 weeks, bone scan with SPECT had a negative predictive value of 1.

Negative predictive values of physical exam combined with radiographic findings. Source: Auerbach JD

Because this bone scan finding was only based on 6 patients, the researchers validated it in their original 873-patient population.

“We found 29 patients who had a bone scan with a symptom duration of less than 6 weeks,” Auerbach said. “And again, we found that SPECT was 100% sensitive and specific in determining a difference between Group I and Group II.”

Bone scan (71%) was also superior in detecting spondylolysis, the most common Group II diagnosis. MRI and X-ray findings were 57% and 38% sensitive, respectively, according to the study abstract.

The researchers tested their provisional algorithm on the 100 patients to retrospectively determine if any tests were performed unnecessarily.

“We found that [among] our 100 patients … there were 88 unnecessary SPECT scans performed and 29 unnecessary MRI scans for a total of approaching $200,000 in extraneous imaging,” Auerbach said.

He was also surprised to find that one in nine CHOP patients who presented with back pain received all three studies during the course of the work-up.

Auerbach believes these recommendations will help surgeons diagnose patients without unnecessary testing that consumes time, radiation and cost.

“At CHOP, a SPECT scan costs $600 less than an MRI,” Auerbach said. “Fewer patients undergoing SPECT will require sedation compared with MRI at our institution.”

Plus, a bone scan with SPECT can be ordered and obtained within 1 to 2 days, compared with the average 2 to 2.5 weeks it takes to obtain an MRI at CHOP, Auerbach said.

The researchers’ next steps are to create and prospectively test back pain “rules.”

“[These] decision rules [will] guide further imaging of the spine, with the goal being to reduce unnecessary testing without missing a single diagnosis of spondylolysis,” Auerbach said.

For more information:

• Auerbach J, Zgonis M, Reddy S, et al. Towards an evidence-based approach for imaging in evaluation of back pain in children. #419. Presented at the American Academy of Orthopaedic Surgeons 73rd Annual Meeting. March 22-26, 2006. Chicago.
• Dr. Auerbach has no direct financial interest in the products discussed in this article, nor is he a paid consultant for any companies mentioned.