Understanding the epidemiology of lower back pain could improve the process for selecting patients for treatment and further management.
While imaging is essential to improving our understanding of lower back pain, its role is limited in the outpatient setting due to a lack of robust prevalence estimates of MRI findings in these patients.
To that end, a recent study out of the University of Washington, Seattle, estimated the prevalence of sex- and age-specific findings in MRI reports of primary care patients presenting with lower back pain.
“By informing a physician’s pre-imaging probabilities, these estimates will not only increase the accuracy of a differential diagnosis without imaging, but also decrease the ordering of unnecessary imaging studies,” said Qian Zhang, an MD/PhD candidate at the University of Washington and the study’s lead author.
Co-authors on the study include Jeffrey Jarvik, MD, radiologist at the University of Washington Medical Center, Seattle, and Michael O'Reilly, MD, interventional fellow, Department of Radiology, University of Washington Medical Center.
Study Includes More than 40,000 Patients
In her study of 44,129 patients, Zhang explained how researchers used natural language processing to locate 26 imaging findings, which were then organized into three groups.
In group one (fracture, any degeneration), prevalence increased with age, with the prevalence between men and women of each age group (18 – 39 years old, 40 – 59 years old) being approximately equal.
In the second group (disc herniation), prevalence did not increase with age and the prevalence between men and women of each age group was approximately equal. In the third group (scoliosis, any stenosis, displaced or compressed nerve root), prevalence between men and women differed by at least 5% with age.
“This grouping can help health care providers remember how age and sex may influence a finding’s pre-test probability,” Zhang said. “For example, a patient who is older would increase a provider’s pre-test probability of a group one finding, while the patient’s sex would not change this probability.”
Informing Pre-Test Probabilities
Researchers then plotted these findings as estimates that can be used to inform pre-test probabilities.
For example, for an MRI finding of foraminal stenosis, the narrowing of the cervical disc space caused by enlargement of a joint in the spinal canal, the prevalence was higher for older patients and for men.
Before obtaining an MRI of the spine, a 60-year-old man with lower back pain would have an 88% probability of having foraminal stenosis, while a 30-year-old woman would have a 38% probability of having foraminal stenosis, Zhang said.
“Without knowing whether foraminal stenosis is associated with lower back pain in 60-year-old men or 30-year-old women, and following current guidelines against performing early imaging in patients without red flag findings on history and physical exam, a primary care provider could use these pre-test probabilities to inform their MRI ordering,” Zhang said.
For example, a primary care provider seeing the 60-year-old man with lower back pain might consider his pre-test probability of foraminal stenosis high enough to merit an MRI, which would likely find foraminal stenosis, and high enough to discuss a possible surgical referral.
But a physician seeing a 30-year-old woman with lower back pain might consider her pre-test probability low enough that an MRI would not be recommended, as it would likely not detect foraminal stenosis, reducing the need for surgery.
Zhang emphasized that history and physical exam further refine pre-test probabilities. And because imaging findings and symptoms rarely occur in isolation, even if the pre-test probability for one finding is low, it may be high for another finding.
Zhang concluded by stating that the prevalence of imaging findings should not impact current guidelines that do not support early imaging in patients without red flag findings on history and physical exam.
For More Information:
View the research Prevalence of Magnetic Resonance Imaging Findings by Age and Gender in the Lumbar Imaging with Reporting of Epidemiology (LIRE) Cohort — SMK07 at RSNA2020.RSNA.org.