Should We Still Use Red Flags in the Diagnosis of Low Back Pain?

Abstract

Commentary The landmark clinical practice guideline on acute low back problems in adults published by the Agency for Health Care Policy and Research (AHCPR) in 1994 drew attention to the identification of red flags1. These refer to a group of clinical symptoms and signs that may indicate the potential presence of a serious underlying spine pathology like fracture, malignancy, infection, and cauda equina syndrome or a rapidly progressive neurologic deficit. In the absence of these signs and symptoms, diagnostic testing is not clinically helpful in the first 4 weeks of symptoms. Since then, the identification of red flags has become part of our clinical routine. It has also been adopted in many clinical practice guidelines for low back pain all over the world2. However, the utility of red flags in detecting serious spine pathologies has often been questioned, because of the high false-positive rates, the lack of standardization in the wide variety of red flags, and an overall lack of evidence for the diagnostic accuracy of recommended red flags2,3. The current study by Premkumar et al., “Red Flags for Low Back Pain Are Not Always Really Red. A Prospective Evaluation of the Clinical Utility of Commonly Used Screening Questions for Low Back Pain,” provides further insight into the diagnostic accuracy of red flags in low back pain. This is possibly the largest single-center study on the utility of red flags in low back pain, with nearly 10,000 patients completing a standardized questionnaire of red flag questions before seeing their surgeon. The authors found that “while a positive response to a red flag question may indicate the presence of disease, a negative response to 1 or 2 red flag questions does not meaningfully decrease the likelihood of a red flag diagnosis.” They recommended that clinicians use caution when interpreting red flags as screening tools. Red flags are still red. This statement is supported by the increased likelihood of diagnoses of fracture, malignancy, infection, and cauda equina syndrome when there were positive responses to the red flag questions in this study, except for night pain. Combining 2 red flag symptoms increased the likelihood ratios further. Night pain as represented by the red flag question “Pain awakens from sleep” was nonspecific, with high false-positive rates of >85% for malignancy and >96% for infection when no other red flag symptom was present. It was not found to be a useful question to determine any diagnosis. However, it should be noted that night pain was not listed as a red flag in the 1994 AHCPR guideline. The medical history red flags that were related to pain for cancer detection in the AHCPR guideline were “Bed rest no relief” and “Duration of pain >1 month.”1 The clinical practice guideline published jointly by the American College of Physicians (ACP) and the American Pain Society (APS) in 2007 also did not include night pain as a key feature in the diagnostic workup for malignancy and infection, although “awakening due to back pain during the second part of the night…” was listed in the diagnostic workup for ankylosing spondylitis4. Interestingly, this 2007 guideline made no reference to “red flags,” but instead listed key features in the history and physical examination that were almost identical to red flags in the diagnostic workup for possible causes of low back pain. Both the AHCPR and the joint ACP/APS guidelines recommend performing a focused medical history and physical examination during which red flags or key features of serious underlying spine pathology are identified. These red flags or key features form part of the clinical assessment, quite unlike the current study design in which the red flag questions were in a self-administered questionnaire completed before the visit. The findings and recommendations of the current study should be interpreted in the context of its study design. Misinterpretation of the questions and incorrect filling out of the questionnaire could give rise to inaccurate data. The authors acknowledged that it is possible that patients may respond differently to the questions when asked verbally by their physician, who would be available for clarification of the question. It is also unrealistic to expect high sensitivity and specificity of diagnoses from just 2 to 4 questions in a self-administered questionnaire. This study clearly shows the inherent danger of identifying red flags using screening questionnaires alone, as a negative response to some of the red flag questions did not help to rule out red flag diagnoses. Clinical evaluation by medical history and physical examination is still mandatory.