Statistics from Altmetric.com
Low back pain is the leading cause of years lived with disability and the second most common reason for a physician visit.1 Although its prevalence has not changed, imaging for low back pain and the cost of its management have continued to increase. For example, in a US study 9.6% of all patients with low back pain had a CT or MRI in 2006 compared with 3.2% in 2002.1
Lumbar imaging is of low yield in the general population presenting with low back pain, as the majority of cases are self-limited and benign. In a previous editorial,2 we discussed the negative effects of inappropriate imaging such as subsequent additional downstream testing, additional imaging, invasive diagnostic procedures and unwarranted patient fear and anxiety leading to poorer outcomes, with concomitant cost and radiation exposure.
Lumbar imaging is indicated when serious pathology is suspected and red flags are an accepted method for identifying when such pathology may exist (eg, malignancy, infection, fracture). However, developing a better understanding of which red flags are predictive is needed in order to reduce unnecessary imaging tests and patient distress.
The good, the bad and the guidelines
Red flags were initially developed based on expert opinion; however, several red flags previously endorsed by guidelines have poor or untested diagnostic accuracy.3
A systematic review3 highlighted those red flags which are most indicative of serious pathology. For example, older age, prolonged steroid use, severe trauma and contusions or abrasions increase the probability of fracture between 10% and 33% and the presence of multiple red flags increased the probability of fracture between 42% and 90%. Similarly, a history of malignancy meaningfully increased the probability of malignancy to between 7% and 33%. In contrast, red flags proposed to predict higher risk of malignancy such as older age, unexplained weight loss and failure to improve after 1 month did not meaningfully affect risk, with post-test probabilities of below 3%.3
One reason for the low diagnostic utility of some red flags lies in the low prevalence of some conditions for which red flags have been developed. For example, spinal infection and cauda equina syndrome, while critical to identify, are so rare that many primary care practitioners may never encounter them.4 With such low prevalence, it is not surprising that the positive predictive value of a red flag for such a condition would be low.
Referring patients to imaging based on the presence of a single red flag, as suggested by some guidelines,5 is unwise.3 Up to 80% of patients presenting to primary care with acute low back pain may have at least one red flag6 despite less than 1% having serious pathology. High-quality research is needed to evaluate the accuracy of red flags, or combinations of these in identifying those most in need for imaging.
The real challenge
Despite their availability, updated clinical practice guidelines are frequently not integrated into practice. Active and individualised methods are more effective for changing clinicians’ behaviours than passive guideline dissemination to clinicians through strictly educational interventions.7 Innovative strategies such as using clinical decision support have also shown some promise. Clinical decision support employs a series of questions and checklists added to the existing computerised order-entry forms for medical imaging to help physicians make appropriate imaging decisions.
Recent studies have found that clinical decision support is effective both in a primary care setting and in the emergency department. In a systematic review by Jenkins et al,7 the use of imaging for low back pain was significantly reduced by 36.8% when using clinical decision support in a hospital setting. Similarly, Min et al 8 found that the use of clinical decision support in the emergency department significantly reduced the proportion of patients with low back pain with an imaging order from 22% to 17%. Finally, an observational study by Blackmore et al 9 also found clinical decision support to decrease hospital-wide low back pain MRI rates by 23%. Clinical decision support likely modifies clinician behaviour via the ‘gatekeeper effect’ by making the clinician more accountable and critical in ordering imaging studies, in addition to the ‘education effect.’8
It takes a village
Red flags for imaging low back pain are a useful basis to inform clinical decisions; however, these need to be constantly updated to reflect evidence of utility. Successful implementation of appropriate imaging guidelines takes a village and champions within it. All individuals in the care chain, including guideline developers, primary care and specialist providers, nurses, radiologists, administrators, and of course patients, need to be fully engaged and supportive, if there is to be any impact on reducing inappropriate imaging for low back pain. Less can be more.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.