Common 'Red Flags' Are Poor Indicators of Spinal Cancer

Roxanne Nelson

February 27, 2013

The identification of spinal cancer in patients with low back pain is not helped much by the "red flag" signs and symptoms that indicate a higher risk for a serious underlying condition, according to a review published online today in the Cochrane Database of Systematic Reviews.

Ideally, these features would "raise the index of suspicion of having the disease to a level that would suggest further diagnostic work-up," the researchers write. However, in the few studies published in this area, they found insufficient evidence of diagnostic accuracy or usefulness in detecting spinal malignancy for most red flags.

"In clinical practice, physicians arguably use a combination of several elements of their assessment to determine whether a patient requires further diagnostic testing," said lead author Nicholas Henschke, PhD, a research fellow at the George Institute for Global Health in Sydney, Australia. "In our review, we attempted to provide data on which of the red flags is most important," he explained.

However, "apart from a previous history of cancer, there are insufficient data on most red flags to support their use in alerting physicians to the need for further testing," Dr. Henschke told Medscape Medical News.

The red flags for spinal cancer commonly cited in clinical practice guidelines include being older than 50 years, no improvement in symptoms after 1 month, insidious onset, a history of cancer, no relief with bed rest, unexplained weight loss, fever, thoracic pain, and being systematically unwell.

Data on the diagnostic accuracy of clinical indicators and recommendations on what to do if they are present should be re-evaluated in the guidelines, Dr. Henschke said. "The key is finding which clinical indicators are of use in a clinical setting when evaluating patients with low back pain. The low prevalence of spinal malignancy in this group makes this a difficult task," he explained.

One Flag Only

Of the 4 red flags that were recently endorsed by American Pain Society guidelines (unexplained weight loss, being older than 50 years, failure to improve after 1 month, history of cancer), only a history of cancer had a sufficiently high positive likelihood ratio to "meaningfully increase the probability of malignancy," the authors write. In fact, it increased the post-test probability of malignancy beyond 2%.

The other 3 red flags, when taken in isolation, only had a modestly high positive likelihood ratio, the authors note. For being older than 50 years and for failure to improve after 1 month, there were "substantial false-positive rates," which argue against their recommended use in clinical practice.

Other red flags, including thoracic pain, severe pain, and insidious onset, have both positive and negative likelihood ratios that come close to 1. This suggests that these red flags are of no real value in determining the likelihood of cancer.

Low Prevalence of Spinal Cancer

Identifying serious pathologic conditions, such as spinal cancer, is one of the primary purposes of the clinical assessment of patients who report low back pain.

In their study, Dr. Henschke and colleagues note that empiric data on the diagnostic accuracy of these features is limited; there is also very little information on how best to apply them in the clinical setting. Their goal was to evaluate the diagnostic performance of clinical characteristics identified during a physical examination and clinical history.

A total of 8 studies met the review criteria and were included in the meta-analysis. Of these, 6 were conducted in the primary-care setting (n = 6622), 1 was conducted in an accident and emergency setting (n = 482), and 1 was conducted in a secondary-care setting (n = 257).

The prevalence of spinal malignancy (21 cases) in the 6 primary-care studies ranged from 0.00% to 0.66%. In the accident and emergency setting, the prevalence was 1.45%; in the secondary-care setting, it was 7.00%.

All of the reports assessed individual tests from the clinical history or physical examination, but none provided any data on using a combination of tests to screen for spinal malignancy.

In all, the studies analyzed evaluated 20 red flags: being older than 50 years, being older than 70 years, constant progressive pain, duration longer than 1 month, gradual onset before the age of 40, familiarity with low back pain, insidious onset, no improvement after 1 month, history of cancer, recent back injury, severe pain, being systematically unwell, thoracic pain, bedrest with no relief, unexplained weight loss, altered sensation from the trunk down, fever, muscle spasm, neurologic symptoms, and spine tenderness.

Insufficient Evidence

The team found that only 7 of the 20 red flags were evaluated by more than 1 study, so there is "insufficient evidence to support or refute the clinical usefulness of most red flags to screen for spinal malignancy" in patients reporting low back pain.

They point out that the large number of patients with false-positive red-flag symptoms is of concern, because the presence of these conditions does not help clinicians decide whether further investigation or treatment is needed.

A large number of patients who presented with some of these red flags showed no signs of spinal malignancy when imaging was used.

"While the lack of evidence to support or refute the use of 'red flags' is recognized, a more pragmatic solution is to consider the possibility of spinal malignancy (in light of its low prevalence in primary care) when a combination of recommended 'red flags' are found to be positive," the authors write.

Because of the low prevalence of spinal malignancy in the primary-care setting, future studies would need to be very large to be sufficiently powered to produce precise estimates of the sensitivity and specificity of red flags. "Potentially, the quality of the evidence around diagnostic tests for such a rare condition could be improved through the use of well designed case–control studies or mathematical modeling to identify appropriate diagnostic strategies," they conclude.

The study was supported internally by Vrije Universiteit, EMGO Institute for Health and Care Research in the Netherlands, and the George Institute for Global Health in Australia. The authors have disclosed no relevant financial relationships.

Cochrane Database Syst Rev. 2013;2:CD008686. Abstract