Low Back Pain CPG (2012): Red Flags and Outcome Measures

How is the OCS going to approach low back pain differential diagnosis and red flags? Dr. David Smelser discusses low back pain red flags using the CPG as well as a recent JOSPT article by Finucane et al. You can bet your bottom dollar this will be on the exam.

Find the Finucane article at https://www.jospt.org/doi/10.2519/jospt.2020.9971.

Transcript

Today on the OCS Field Guide Podcast, we’re going to talk about low back pain differential diagnosis, red flags, and outcome measures from the 2012 Clinical Practice Guidelines.

As we begin our discussion of low back pain differential diagnosis, I think it is very important that we read the section of the CPG labeled “pathoanatomical features.”

The CPG reads, “Any innervated structure in the lumbar spine can cause symptoms of low back and referred pain into the extremity or extremities. This long list of potential structures includes the muscles, ligaments, dura mater and nerve roots, zygapophyseal joints, annulus fibrosis, thoracolumbar fascia, and vertebrae. One might expect that improvement in the resolution of imaging technology has increased the likelihood of detecting a link between pathology and pain in the lumbar spine. However, the determination of a pathoanatomic origin of low back pain is made difficult by the rate of false-positive findings on imaging studies, that is, individuals without low back pain showing abnormal findings. For example, evidence of herniated disc material is shown on computerized tomography (CT) scans, MRI, and myelography in 20% to 76% of persons with no sciatica. Furthermore, Savage et al reported that 32% of their asymptomatic subjects had “abnormal” lumbar spines…and only 47% of their subjects who were experiencing low back pain had an abnormality identified. 

“In longitudinal studies, low back pain can develop in the absence of any associated change in radiographic appearance of the spine. Boos et al followed asymptomatic patients with a herniated disc for 5 years and determined that physical job characteristics and psychological aspects of work were more powerful than MRI-identified disc abnormalities in predicting the need for low back pain–related medical consultation. Thus, the association between clinical complaints and concurrent pathological examination with radiological findings must be considered cautiously. Further, even when abnormalities are present, establishing a direct cause and effect between the pathological finding and the patient condition has proven to be elusive and most often does not assist greatly in patient management.”

In other words, in the vast majority of cases, imaging is next to useless at determining a clear cause of low back pain. A huge proportion of individuals with no back pain have significant degenerative findings on MRI, and many individuals with back pain have no significant MRI findings. So when we get into the differential diagnosis section of the CPG, the CPG reads, “The best available evidence supports a classification approach that de-emphasizes the importance of identifying specific anatomical lesions after red flag screening is completed.”

So when you read a low back pain case on the OCS, your differential diagnosis should not be formed with specific anatomical lesions in mind. You shouldn’t be thinking, “this case sounds like facet dysfunction!” Because right now we don’t have the medical technology to really distinguish between facet dysfunction or pain originating from any of the other structures in the lumbar spine. And as the CPG mentioned, our research indicates that psychosocial factors play a much bigger role in back pain than the pathoanatomical features we can identify. The OCS is generally not going to look for pathoanatomical explanations for generic low back pain, and it’s also not going to look for guru-based explanations and interventions. So, for example, if you’ve taken a lot of osteopathic-based courses, you shouldn’t be thinking about ERS-Left and FRS-Right on the OCS exam. Instead, let’s read that sentence from the CPG again:

“The best available evidence supports a classification approach that de-emphasizes the importance of identifying specific anatomical lesions after red flag screening is completed.”

So there are two components to differential diagnosis with low back pain. First is red flag screening, which we are going to discuss next. The second is using a classification approach that does not rely on identifying specific anatomical lesions. The authors here are talking about a treatment-based classification to low back pain, which we are going to dive into in our next episode. For now, let’s look at those red flags:

The main red flag conditions we need to be worried about in the lumbar spine are cancer, cauda equina syndrome, back-related infection, spinal compression fracture, and abdominal aneurism. I’m going to go through each one and discuss the red flags from the CPG, and I’m also going to pull in some information from an article published in JOSPT a couple months ago by Finucane  et al., called, “International Framework for Red Flags for Potential Serious Spinal Pathologies.”

First, cancer, or back-related tumor.

Red flags associated with a back-related tumor are: constant pain not affected by position or activity, age over 50, history of cancer, failure to improve with 30 days of conservative management, unexplained weight loss, and no relief with bed rest.

The estimated prevalence of cancer among people presenting to a clinic with low back pain is 0.7%. So if someone walks in your clinic doors with low back pain, there is a 0.7% chance that pain is due to cancer. It is going to take some very large positive likelihood ratios to increase that probability up to a point where you refer the patient out of your clinic.

With that in mind, the red flag with the largest positive likelihood ratio is a previous history of cancer. That positive likelihood ratio is 23.7, which, if you remember from our research and statistics episode, is very large. There are a few types of cancers that are most likely to metastasize to the spine: prostate, thyroid, breast, lung, kidney. There are a couple mnemonics to remember these, but the mnemonic I was taught is, “PT Barnum Loves Kids.” P: prostate, T: thyroid, B/Barnum: breast, L/Loves: lung, K/Kids: kidney. If you see a history of these cancers in a back pain case, your suspicion of cancer should increase; however, remember that if you burst out into a song from The Greatest Showman, you might get kicked out of the OCS testing center.

The next red flags with the largest positive likelihood ratios are unexplained weight loss and failure to improve in 30 days with conservative management, followed by age over 50. The most sensitive finding is no relief with bed rest.

You will notice that I haven’t used the term “night pain” as a red flag. This is because night pain is so common in musculoskeletal conditions that it is a very poor red flag. Instead of thinking about “night pain,” think about constant pain that does not change with position or activity and does not improve with bed rest. In other words, if no position or movement or activity or rest can change the level of the pain—either better or worse—then you’re probably not looking at a musculoskeletal source of pain.

So what do we do with this information? No individual red flag is enough to warrant an immediate referral. However, when you start to see at least two or three of these clustered together, your suspicion might increase enough to refer out. So age over 50 and a history of cancer may not make you refer immediately, but it’s enough to watch the patient closely. Then if the patient fails to improve at all after a month of therapy, or if the pain is constant and doesn’t change at all with position or activity, now you have reason to refer.

And where do you refer? MRI is the gold standard for ruling out cancer. However, referring for an erythrocyte sedimentation rate blood test (or and “ESR”), is typically faster and much more affordable than an MRI. An ESR value of above 20 should increase your suspicion of cancer slightly, but you should be really concerned if the ESR is above 50. An ESR over 50 has a positive likelihood ratio of 19.2 for presence of cancer. So if you get a patient case on the OCS, and the patient is more than 50 years old, has a history of cancer, and has an ESR above 50, the probability that this patient has cancer is 88.9%. Refer this patient for imaging immediately.

Before we wrap up our discussion of cancer, I want to jump ahead to the CPG section on outcome measures—and I’ll explain why in a moment. The CPG recommends that clinicians use two main region-specific outcome measures. The first is the Oswestry Disability Index, sometimes abbreviated ODI. This is by far the most common. The second is the Roland-Morris Disability Questionnaire, and we will briefly talk about it first. The Roland-Morris Disability Questionnaire is a list of 24 statements about back pain, and the patient marks all the statements that apply to them. The test is scored by adding up all the positive responses, so a score of 24 indicates very high disability, and a score of zero is very low. A change of 5 points or a 30% improvement is considered to be the minimal clinically important difference. If you’re unfamiliar with the term “minimal clinically important difference,” sometimes abbreviated “MCID,” here’s what it is: the MCID is the smallest change you need to see on your outcome form to be confident that the change represents an actual, meaningful improvement in the patient’s clinical status and isn’t just the result of test-retest variability or the patient being in a good mood that day.

The first test I mentioned, the Oswestry Disability Index (or ODI), is by far the most common low back pain outcome measure. This test is scored as a percentage, with 100% representing complete disability, and 0% representing no disability. Interpretation of the ODI uses cutoffs every 20 percentage points, so that 0-20% indicates minimal disability, 21-40% indicates moderate disability, 41-60% indicates severe disability, and it just gets worse from there. The MCID on the Oswestry is 10 percentage points. So you need to see an improvement of 10 points on the ODI to know that your patient is improving.

Now I was told that I needed to memorize all of the MCIDs for all of the outcome measures for every part of the body before I took my OCS. And that’s crazy. There are just too many outcome measures, and the MCID numbers are nearly meaningless when you’re just memorizing them. So will MCIDs be on the exam? Yes. But do you need to memorize all of them? I don’t think so. The MCIDs that are most likely to be on the test are the MCIDs from the most common outcome measures that have the most research behind them. The Oswestry fits the bill. So will you need to know the Oswestry’s MCID on the OCS exam? My guess would be almost certainly yes.

But the test won’t say something like, “What’s the ODI’s MCID?” Instead, it’s going to give you a patient case, give you some outcome scores, and then you’re going to have to interpret them to decide what to do next. So let’s make up a case:

A 61-year-old male patient presents to your clinic with central low back pain without radiating pain that began insidiously and has been bothering him for a couple months now. His pain is a 4/10, and he has been using a heating pad for pain relief. The patient has a history of type II diabetes, HTN, occasional alcohol use, and prostate cancer. His ODI at his initial evaluation is a 39%. You have been treating him with a combination of manual therapy, motor control training, and overall strengthening and work hardening. After a month of treatment, you re-evaluate the patient, and his ODI is a 31%. What do you do next?
A. Discharge to an HEP.
B. Change your current treatment plan to include lumbar traction.
C. Progress plan of care to focus more on work hardening and less manual therapy.
D. Refer to a physician for an ESR.

You will notice in this case that you need to know a few different things. You need to pick up on the red flags: that the patient is above 50 years old and has a history of cancer that tends to metastasize to the lumbar spine. You also need to know what the ODI is, that a 21%-40% score indicates moderate disability, and you need to know that the MCID is 10% to know if you patient has actually improved. In this case, the score improved by 8 percentage points, but that doesn’t meed the ODI’s minimal clinically important difference. So now we have a patient who is above 50, has a history of cancer, and has failed to improve significantly after 30 days of conservative management. It is time to call the physician and rule out cancer.

So you don’t need to know every imaginable outcome measure and MCID. But you definitely need to know the big ones. And the Oswestry is a big one.

Let’s go back to our red flags discussion and move on to cauda equina syndrome.

The most significant red flags for cauda equina syndrome are bowel and bladder changes, saddle anesthesia, and sensory or motor deficits in the L4, L5, and S1 areas.

Out of these red flags, urinary retention appears to be the most sensitive and specific. Although urinary incontinence is sometimes considered a red flag for cauda equina, urinary retention appears to be the bigger deal. Urinary retention can mean an inability to urinate, and inability to urinate much, or an inability to completely empty the bladder.

For these findings to be most significant, the onset of these symptoms need to coincide with the onset of the back pain. So if a patient reports urinary retention, fecal incontinence, or saddle anesthesias that started around the same time their back pain started, your suspicion of cauda equina syndrome should be high. So what next? Remember that cauda equina syndrome is a time-dependent medical emergency, so you refer straight to the emergency room for this one. If there is cauda equina syndrome present, the faster those nerves get decompressed, the greater a chance of full recovery the patient has, so refer straight to the ER.

Next is back-related infection.

We don’t have as much hard data on the sensitivity and specificity for red flags suggesting spinal infection, so this diagnosis is a little trickier. Here are the flags from the CPG: recent infection, intravenous drug user, concurrent immunosuppressive disorder, deep constant pain that increases with weight bearing, fever/malaise/swelling, and spine rigidity. Out of all of those red flags, spine rigidity seems to be the least useful. Fever and swelling seem to me to be the most useful. So one more time, those red flags are: recent infection, intravenous drug user, concurrent immunosuppressive disorder, deep constant pain that increases with weight bearing, fever/malaise/swelling, and spine rigidity.

According to the JOSPT article I mentioned above by Finucane et al., back infections eventually progress to cause neurological involvement, and the “classic triad” of clinical findings for spinal infection are back pain, fever, and neurological dysfunction. However, the authors mention that many people don’t exhibit all three of these findings, and only 50% of people with spinal infections report fevers. So a fever should make you suspicious of a spinal infection, but absence of a fever can’t rule out infection. If you start seeing several of these red flags together, like recent infection, immunosuppression, malaise, and constant pain, you should still be suspicious of a back-related infection.

Next is a spinal compression fracture.

The CPG lists the risk factors of spinal compression fracture as a history of major trauma, such as vehicular accident, fall from a height, or direct blow to the spine; age over 50, with age over 75 being an even higher risk factor; prolonged use of corticosteroids, point tenderness over the side of the fracture, and increased pain with weight bearing.

The Finucane article sheds a little more light on the factors and adds a couple: a history of osteoporosis and history of spinal fractures are also very high risk factors for a subsequent fracture. Trauma is an even greater risk factor if it involves flexion. Female sex is a significant risk factor, as 12% of women aged 50-70 have had a spinal fracture, and 20% of women over 70 have had a spinal fracture—and 70% of these will not know about it. Prolonged corticosteroid use is defined as greater than 3 months—so it doesn’t have to be years or lifelong, it could just be three months of taking a corticosteroid. And finally, thoracic pain in a population with some of these other risk factors should make you suspect a thoracic compression fracture, since 70% of non traumatic spinal fractures occur in the thoracic spine.

Finally, abdominal aneurysm.

Again, we don’t have a lot of hard data here, so I will give you the list of risk factors and then point out the ones I think are most significant. The CPG’s list is as follows: back, abdominal, or grain pain; presence of peripheral vascular disease or coronary artery disease and associated cardiovascular risk factors; smoking history; family history; age over 70; non-caucasian; female; symptoms not related to movement stresses associated with somatic low back pain; presence of a bruit (or bru-ee if you prefer the French) in the central epigastric area upon auscultation, palpation of aortic pulse, and aortic pulse 4 cm or greater with an even higher risk if the pulse is 5 cm or greater.

Out of these risk factors, the arterial disease and cardiovascular risk factors are important to increase your suspicion of a vascular issue. But if you see things like a bruit or a palpable aortic pulse—especially if it’s large enough to measure at 4 or 5 cm—you should definitely be concerned. And if you’re seeing a 5 cm aortic pulse with throbbing low back pain that appears unrelated to movement stresses, that’s another emergency room referral.

I’ll also note that the CPG includes abdominal girth <100 cm in this list, and that seems counterintuitive. So I looked up the article they’re citing, and the article is saying that in individuals with a girth <100 cm it is easiest to palpate an aneurism. But in individuals with a girth >100 cm, palpation of an abdominal aneurism gets more difficult. So it’s not that large people don’t get abdominal aneurisms, but you might have to rely less on palpation and more on the other signs when you’re reasoning through your differential diagnosis here.

I want to conclude this discussion of red flags by reading a paragraph from the CPG on imaging. Since we have been talking about when to refer out, it’s also important to know when not to refer for imaging. So the CPG reads, “Imaging modalities have frequent false positive and negative results, limiting their utility in identification of active anatomic pain generators. Therefore, the primary utility of imaging lies in interventional and/or surgical planning or in determining the presence of serious medical conditions. For these purposes, lumbar MRI represents the most useful tool. However, routine ordering of imaging for low back pain should be discouraged. In particular, imaging in acute low back pain has not been shown to yield significant new findings or alter outcomes. In chronic low back pain, the role of routine diagnostic imaging is even less established. Current recommendations from the American College of Physicians are that (1) imaging is only indicated for severe progressive neurological deficits or when red flags are suspected, and (2) routine imaging does not result in clinical benefit and may lead to harm. 

So when do we refer for imaging? When we see clusters of the red flags we just described, or when there are severe, progressive neurological deficits and surgery needs to be considered. Otherwise, unnecessary imaging may lead to harm.

That wraps up this discussion on red flags. If you want to dive deeper, I highly recommend the JOSPT article by Laura Finucane I’ve been referencing. It includes case vignettes that go along with each condition so you can practice your red flag screening. I will include a link to it in this episode’s details. Next week, we'll get into the treatment-based classification approach, which is bound to be a huge portion of the OCS exam.