Excessive levels of diagnostic imaging remain a frustrating and persistent problem in the management of low back pain in the United States and elsewhere. Indeed, the lack of success in reducing the overall prevalence of spinal imaging has become one of the signal failures of the modern evidence-based back care movement. Inappropriate imaging has exacerbated the back pain crisis, inflated costs, and harmed patients for decades.
There have been numerous attempts to address this problem through interventions involving patient and provider education, decision support systems, hybrid interventions, and even algorithms based on machine learning and artificial intelligence. But no single intervention has achieved broad and consistent success. (See Jenkins et al., 2015)
And, unfortunately, there are no easy, formulaic answers to this problem on the near horizon. Many physicians and other healthcare providers pride themselves on adhering to the recommendations of evidence-based guidelines regarding imaging decisions. Those guidelines rely heavily on the assessment of red flags—i.e., symptoms and signs of serious abnormalities and conditions—to guide those decisions.
Yet research over the past two decades has largely discredited individual red flags as a roadmap to imaging decisions. “Nearly all recommended individual red flags are uninformative and do not substantially change post-test probabilities of a serious abnormality,” noted the landmark Lancet Series on low back pain. Healthcare providers shouldn't ignore the possibility of serious conditions. But they need to look at the larger clinical picture rather than individual red flags. And they need to look beyond the initial visit.
“Clinicians...need to consider if the overall clinical picture might indicate a serious cause for the pain, remembering that the picture can develop over time,” according to the Lancet Series authors. (See Hartvigsen et al., 2018)
Need for a New Research Agenda
Clearly, there is a need for a major research agenda in this area, to redefine the indications for imaging and develop interventions to promote appropriate use.
Aron Downie, MChiro, of the University of Sydney in Australia has authored or coauthored several recent studies on imaging for low back pain, including a recent international systematic review on the prevalence of imaging for this common symptom. (See Downie et al., 2019) He said in a recent email that he favors a multi-pronged research approach, starting with developing new criteria for appropriate imaging.
“The majority of red flags are uninformative for raising suspicion of pathology when patients present with LBP, and appear to be unhelpful when used to guide imaging decisions despite their inclusion in appropriateness criteria. Prospective studies that explore the performance of a reduced set of red flags alone or in combination, and in the context of the clinical presentation, will advance knowledge of how red flags may contribute to improved imaging decisions,” according to Downie.
He would like to see reform of clinical practice guidelines and imaging ordering systems. “The majority of clinical practice guidelines continue to endorse imaging based on the presence of a single red flag, many of which do not increase the probability of disease. In contrast, the recommendations of Roger Chou et al in 2011, and Chris Maher et al in 2017 to consider the level of clinical suspicion and urgency, with or without a trial of therapy prior to imaging referral, should be included in future guidelines, then tested within institutional imaging ordering systems,” said Downie. (See Chou et al., 2011; Maher et al. 2017.)
And Downie said he also favors studying electronic medical records with data analytic techniques to better understand the behavior of clinicians regarding imaging. “Real-time monitoring—in the emergency setting, for instance—may be a novel way to test strategies aimed to minimize low-value care in the management of back pain.”
What About the Underuse of Imaging?
While this BackLetter article primarily addresses the overuse of imaging for low back pain, there is also a significant problem with the underuse of imaging among patients who need it. A 2018 review by Hazel J. Jenkins, MChir, of Macquarie University and colleagues, documented problems on both sides of this equation. (See Jenkins et al., 2018.)
“In patients presenting for care, imaging was not performed where appropriately indicated in 65.6% (95% CI: 51.8, 77.2) of patients who presented with red flags, and 60.8% (95% CI: 42.0, 76.8) with clinical suspicion of serious pathology,” according to Jenkins et al. [See study for details.]
Researcher Mark Hancock, PhD, of Macquarie University, was senior author on the systematic review by Jenkins et al. and a coauthor of the Downie review. The two studies were performed in parallel, he explained in a recent email.
“[In the review by Jenkins et al.], we attempted to rate the proportion of appropriate and inappropriate imaging and non-imaging. This paper found high levels of inappropriate non-imaging as well as high rates of inappropriate imaging,” Hancock added.
“A key issue is [that] the criteria for appropriate imaging remain unclear and vary substantially between guidelines, so it is not surprising that clinicians are confused. If you use the presence of red flags as the criteria for appropriate imaging, then you will continue to have high rates of imaging—as we know most patients present with at least one red flag,” Hancock explained.
So these are deceptively complex problems affecting the evaluation of patients with low back pain.
“For imaging in people with LBP we have two problems: medical overuse and medical underuse,” said researcher Chris Maher, PhD, recently. The second problem has not really been well recognized.”
There is a need for research on both sides of this problem, a better understanding of the drivers of inappropriate imaging, better definitions of appropriate and inappropriate imaging, and the development of sophisticated interventions that might promote their implementation.
By way of background, evidence-based guidelines have recommended against the routine use of spinal imaging in the early management of low back pain for more than a quarter century. Major professional societies across the back pain and spinal medicine fields have been nearly unanimous in their recommendations against routine imaging.
Numerous “Choosing Wisely” lists from major societies warn against excessive imaging. “In patients with back pain that cannot be attributed to a specific disease or spinal abnormality following a history and physical examination (e.g., non-specific low back pain), imaging with plain radiography, computed tomography (CT) scan, or magnetic resonance imaging (MRI) does not improve patient outcomes,” according to the American College of Physicians' Choosing Wisely list (http://www.choosingwisely.org/clinician-lists/american-college-physicians-imaging-for-non-specific-low-back-pain/).
Yet the medical field has seen scant, if any, progress in reducing the level of imaging. Some researchers suggest that imaging for low back pain is still a standard feature of back pain management in a third or more of patients. And that a substantial proportion of imaging referrals are inappropriate. Disturbingly, some forms of imaging may actually be on the increase in parts of the United States and several other countries.
Gauging the Prevalence of Imaging
Gauging the overall prevalence of imaging is no simple matter. Most studies in this area, small and large, are retrospective snapshots of selected groups of patients culled from patient records and insurance company data—and are not necessarily representative of the millions of patients with back pain who seek medical care. But the available studies certainly suggest that overutilization of imaging is common.
There is no established benchmark to determine what proportion of patients in primary care settings warrant imaging for their low back pain. But that proportion is likely to be fairly low.
In a 2018 systematic review and meta-analysis, Jenkins et al. offered the following estimate regarding appropriate imaging utilization. “Serious pathology (e.g., infection or tumor) or specific pathology (e.g., spinal canal stenosis) is estimated to be the cause of LBP in less than 1% and 10% of presentations, respectively. Therefore, commonly reported imaging rates for acute LBP of about 35% are considered too high, resulting in increased health-care costs, increased radiation exposure, and potentially inappropriate treatment.” (See Jenkins et al., 2018.) And after learning about abnormalities on their imaging scans patients sometimes become convinced they have serious anatomic problems requiring downstream diagnostic and treatment interventions.
Major Systematic Review on Imaging for Back Pain
The largest systematic review of back pain imaging to date—which looked at 45 studies involving more almost 20 million back pain consultations in nine countries—found that imaging is still common and that the prevalence of advanced imaging rose over a 21-year period. (See Downie et al., 2019)
Downie and colleagues concluded that 16.3% of consultations in primary care resulted in simple imaging (x-ray and/or ultrasound) and that 9.2% of consultations resulted in complex imaging (MRI, CT, or nuclear bone scan). A glaring 24.8% of consultations resulted in some form of imaging.
“Complex imaging increased from 7.4% (95% CI 5.7% to 9.6%) for imaging requested in 1995 to 11.4% (95% CI 9.6% to 13.5%) in 2015 (relative increase of 53.5%).”
“Our study showed that about one-quarter of patients in primary care and one-third of patients in emergency care will be imaged when they present with low back pain, and that complex imaging (CT, MRI) has increased despite clinical practice guideline-based advice and media campaigns such as Choosing Wisely,” said Downie in a recent email.
Imaging Among Primary Care Providers in the United States
A recent retrospective study of a large population of primary care providers in the United States found that x-ray use actually increased from 2011 to 2015, in defiance of guideline and Choosing Wisely recommendations.
Jina Pakpoor, MD, of Johns Hopkins University and colleagues studied the initial evaluation of low back pain in 627,118 outpatient clinical encounters involving a family medicine or internal medicine practitioner over a five-year period from 2011 to 2015.
Pakpoor studied patients making a first visit associated with a diagnosis of low back pain. The mean age of patients was roughly 45. A little over half—53%—were female. Fifty-four percent were active, full-time employees. The researchers looked at the use of radiographs, CT scans, MRI, or myelography.
“The percentage of patients undergoing any imaging increased from 14.3% in 2011 to 16.0% in 2016 (P < .01),” according to Pakpoor et al. X-rays represented the vast majority of scans—96%. MRI accounted for 3% and CT scanning for 2%.
They found significant geographic variation in imaging rates, for reasons that aren't clear. “The likelihood of having any imaging for low back pain varied significantly by US census region and by US state (P < .001). The greatest use of imaging was in the Midwest (13.9%) and the South (18.5%), and lowest in the Northeast and West (6.2% and 13.6%).”
This study documented a reasonably moderate imaging rate compared to other studies. But it is important to note that the authors only looked at imaging referrals made during the first visit. And even the 16% rate of imaging reported in this study is likely to be excessive.
Study of 2.5 Million Patients Without a Red Flag Diagnosis
Another recent study from the United States found much higher levels of imaging for low back pain. Lily H. Kim of Stanford University and colleagues performed a retrospective cohort study of nearly 2.5 million patients seeking medical care for low back (LBP) or lower extremity pain (LEP) between 2008 and 2015. “Participants were US adults who were newly diagnosed with LBP or LEP between 2008 and 2015, did not have a red-flag diagnosis, and were opiate naive prior to diagnosis,” according to Kim et al. (See Kim et al., 2019.)
They divided the study into two cohorts: the 1.2% of subjects treated surgically and those managed nonoperatively. They wanted to see, in particular, whether those treated nonoperatively underwent management that was in line with evidence-based recommendations. And they did not, with imaging levels a glaring example of inappropriate care.
“Approximately one-third of patients (32.3%) treated nonsurgically received imaging within 30 days of diagnosis (CT, 0.7%; MRI, 8.1%; and radiography, 26.7%),” according to Kim et al.
Guidelines have often warned that early imaging can highlight irrelevant spinal abnormalities and lead to a cascade of unnecessary diagnostic and treatment interventions and elevated downstream costs.
And this study provided some support for those warnings. “Adjusted mean 12-month costs among patients who received imaging within 30 days of diagnosis were more than two times greater than costs for patients who did not receive early imaging ($1194 [95% confidence interval (CI), $1190–$1199] vs. $566 [95% CI, $563–$569]; P < 0 .001),” according to these researchers. Imaging accounted for roughly 20% of overall nonsurgical costs.
Half a Billion Dollars Spent on Imaging
Senior author and neurosurgeon John Ratliff, MD, said he was surprised at the pattern of imaging in this study.
“We anticipated that many patients with new onset low back or leg pain would receive imaging. My team was very surprised at how many health care resources were consumed by these potentially unnecessary films: $500 million every year,” said Ratliff in a recent email.
“We expected that there would be some discordance between guideline recommendations and clinical practice, but we did not anticipate it would add up to that huge an annual expense.”
He said that in many cases imaging did not appear to have any impact on subsequent care.
“We were also surprised by how many patients had MRI examinations and then nothing else—they did not progress to PT, or epidurals, or other interventions. It is likely that the MRI contributed little to their care.”
“Early imaging in this patient population is a clear opportunity for process improvement,” said Ratliff. “This is an area where there is general agreement among the various spine societies.” In other words, consensus building among professional societies is not an obstacle to progress.
A BackLetter editor asked Ratliff if he had any general advice for providers and patients considering imaging for uncomplicated low back pain. He suggested they think twice about it.
“New onset of low back and lower extremity pain in patients without a red flag [condition] is almost always self-limited. In the patients in our study, fewer than 2% ever needed operative intervention. Most of these patients will get better on their own, and are better served with a referral to physical therapy than with a prescription for an MRI that may contribute little or nothing to their overall care,” Ratliff noted.
Disclosures: None declared.
Chou R, Diagnostic imaging for low back pain: advice for high-value health care from the American College of Physicians. Annals of Internal Medicine, 2011 Feb 1;154(3):181–9. doi: 10.7326/0003-4819-154-3-201102010-00008.
Downie A, et al How common is imaging for low back pain in primary and emergency care? Systematic review and meta-analysis of over 4 million imaging requests across 21 years. British Journal of Sports Medicine, 2019, pii: bjsports-2018-100087. doi: 10.1136/bjsports-2018-100087.
Downie A, Williams CM, Henschke N, Hancock MJ, Ostelo RW, de Vet HC, et al Red flags to screen for malignancy and fracture in patients with low back pain. Br J Sports Med 2014;48:1518.
Jenkins H, et al Effectiveness of interventions designed to reduce the use of imaging for low-back pain: a systematic review. CMAJ, 2015;187:401–8. doi:10.1503/cmaj.141183
Jenkins H, et al Using behaviour change theory and preliminary testing to develop an implementation intervention to reduce imaging for low back pain. BMC Health Services Research, (2018) 18:734.
Jenkins H, et al Imaging for low back pain: is clinical use consistent with guidelines? A systematic review and meta-analysis, Spine Journal, 2018; 18(12): 2266–77.
Kim LH, et al Expenditures and health care utilization among adults with newly diagnosed low back and lower extremity pain: 2011-2016, JAMA Network Open, 2019; 2(5):e193676. doi:10.1001/jamanetworkopen.2019.3676.
Maher C, et al Nonspecific low back pain, Lancet, 2017 Feb 18;389(10070):736–47. doi: 10.1016/S0140-6736(16)30970-9.
Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.
Pakpoor J, et al Diagnostic imaging use for the initial evaluation of low back pain by primary care providers in the United States: 2011-2016 [published online ahead of print May 21, 2019], Journal of the American College of Radiology. doi:10.1016/j.jacr.2019.04.015.