Red Flags for Low Back Pain- Research Update – July 2014

Sorry for the delay in posts.

I was pleased to recently be asked by the Western Australian branch of Musculoskeletal Physiotherapy Australia (MPA) to present a day of lecturing on their annual “Spinal 1” course, and an evening seminar on “Red Flags for Low Back Pain”. Hence my last couple of months have been taken up by prepping material for these events. The benefits for the blog being I now have a reasonable amount of up-to-date material on a number of topics that I plan to put up over the next few weeks. This first up date is in regards to screening for Red Flags in Low Back Pain.

I have previously written about this topic on the blog, in particularly in relation to this article by Henschke et al (2009). The authors of this study investigated the use of recommended signs in the identification of the following 5 Red Flags presentations in lower back pain:

1. Cancer/Malignancy
2. Infection
3. Spinal Fractures
4. Inflammatory Disorders/Inflammatory Back Pain (IBP)
5. Cauda Equina Compression

The symptoms they used in their screening process were as follows (based on clinical guidelines from the United States, Europe and Australia):

• Cancer
  • Age at onset less than 20 or over 55 years
  • Unexplained weight loss (of more than 10 pounds [4.5 kg] in 6 months)
  • Previous history of cancer
  • Tried bed rest, but no relief
  • Insidious onset
  • Systemically unwell
  • Constant, progressive, non-mechanical pain
  • Sensory level (Altered sensation from the trunk down)
• Infection
  • Systemically unwell
  • Constant, progressive, non-mechanical pain
  • Recent bacterial infection e.g. UTI or skin infection
  • IV drug use
  • Immune suppression from steroids, transplant or HIV
  • Altered sensation from the trunk down
• Spinal Fracture
  • Age over 70 years
  • Significant trauma (major in young, minor in elderly)
  • Prolonged use of corticosteroids
  • Sensory level (altered sensation from the trunk down)
• Inflammatory Disorder/Inflammatory Back Pain (IBP)
  • Gradual onset before age of 40
  • Tried bed rest, but no relief
  • Insidious onset
  • Systemically unwell
  • Constant, progressive, non-mechanical pain
  • Morning back stiffness (30 mins or more)
  • Peripheral joint involvement
  • Persisting limitation of spinal movements in all directions
  • Iritis, skin rashes (psoriasis), colitis, urethral discharge
  • Family history of arthritis
  • Improves with exercise
• Cauda Equina Syndrome
  • Acute onset of urinary retention or overflow incontinence.
  • Loss of anal sphincter tone or faecal incontinence.
  • Saddle anaesthesia about the anus, perineum or genitals
  • Widespread (greater than 1 nerve root) or progressive motor weakness in the legs or gait disturbance.
  • Sensory level (altered sensation from the trunk down)

Key elements from the study:

• It involved 1,172 consecutive patients presenting to primary care for acute low back pain.
• Practioners involved were General Practitioners, physiotherapists, chiropractors. They were trained in screening for Red Flag disorders as per the criteria above.
• A 12 month follow-up process was utilised to continue to screen patients, thereby picking up any disorders that were missed at initial assessment.

Results were:

• Out of the 1,172 patients there were only 11 cases (0.9%) of serious pathology:
  • 8 cases of fracture, 1 cauda equina, 2 inflammatory disorders.
  • No cases of cancer or infection.
• Clinicians only identified 5 of the 11 (45%) cases of serious pathology at the initial consultation.
• In addition they made 6 false-positive diagnoses (i.e. they incorrectly decided that a patient had a Red Flag disorder when in fact they didn’t).
• Despite the low prevalence of serious pathology, most patients (80.4%) had at least 1 red flag sign.
• Only 3 of the red flags for fracture recommended for use in clinical guidelines were informative:
  • Prolonged use of corticosteroids
  • Age >70 years
  • Significant trauma.

So I found those results interesting for a number of reasons:

1. Red flag disorders don’t appear to be common (only 11 out of 1,172 patients had one).
2. Interestingly the 25 question screening process only picked up 4 of the 11 red flags disorders initially, and made 6 incorrect (i.e. false positive) diagnosis’ that a patient had a Red Flag disorder. So basically:
   1.  They missed more than half.
   2. More false positive diagnosis’ were made initially than correct ones.
3. With over 80% of the patients reporting one Red Flag sign we probably shouldn’t be relying on having people investigated for a Red Flag disorder initially based on them having 1 Red Flag sign (remember the overall presence of a Red Flag disorder in this study was limited to only 0.9% 0f the patients).

In the positive, this study was able to devise a diagnostic rule for vertebral fractures, and this has been cited and incorporated into other guidelines since:

• Presence of 3 out of 4 red flags for fractures were highly predictive of fracture [100% specificity, positive likelihood ratio (+ve LR) of 218]:
  • Prolonged use of corticosteroids
  • Significant trauma (major in young, minor in elderly)
  • Age > 70 years
  • Female.

A quick reminder on statistics, in particular likelihood ratios (LR’s):

Findings with LRs greater than 1 argue for the diagnosis of interest; the bigger the number, the more convincingly the finding suggests that disease/presentation.

So the above rule has significant specificity and and very high +ve LR. Key thing to point out being that these values mean that it is good as a diagnostic rule (i.e. the presence of these findings means a very high likelihood that a vertebral fracture is present), but when you look further into the findings this cluster of signs does not have good sensitivity and therefore cannot be used as a screening process (i.e. if the patient doesn’t have the cluster of signs it doesn’t mean they don’t have a fracture). Please see this post for information regarding specificity, sensitivity and diagnostic vs screening if you require any clarification.

Just out of interest, how much is too much in relation to corticosteroids?

Romas (2008) reports that daily doses of more than 2.5 mg prednisolone or equivalent are associated with a higher fracture risk. Intermittent oral corticosteroids and inhaled corticosteroids increase vertebral fracture risk, but patients taking intermittent corticosteroids are less likely to sustain fractures than those taking continuous therapy.

So apart from the vertebral fracture information above, the lack of results from Henschke et al (2009) on red flags for identifying the other disorders suggested to me that I had better move on to look into each topic individually.

Cancer/Malignancy

This recent Cochrane Database Systematic Review by Henschke et al (2013) investigated the following signs in the usefulness of screening for malignancy in LBP:

1. Age > 50
2. Age > 70
3. Constant progressive pain
4. Duration of this episode > 1 month
5. Gradual onset before age 40
6. Is the low back pain familiar?
7. Insidious Onset
8. Not improved after 1 month
9. Previous history of cancer
10. Recent back injury
11. Severe pain
12. Systemically unwell
13. Thoracic pain
14. Tried bed rest with no relief
15. Unexplained weight loss
16. Altered sensation from the trunk down
17. Fever
18. Muscle spasm
19. Neurological symptoms
20. Spine tenderness

So lots and lots of them! Did the findings shed any additional light on the subject that Henschke et al (2009) hadn’t?:

• In six primary care studies (6622 patients), the prevalence of spinal malignancy ranged from 0% to 0.66%.
• There was some evidence from individual studies that having a previous history of cancer meaningfully increases the probability of malignancy.
• Most “red flags” such as insidious onset, age > 50, and failure to improve after one month have high false positive rates.
• All of the tests were evaluated in isolation and no study presented data on a combination of positive tests to identify spinal malignancy.
• For most “red flags,” there is insufficient evidence to provide recommendations regarding their diagnostic accuracy or usefulness for detecting spinal malignancy.
• The available evidence indicates that in patients with LBP, an indication of spinal malignancy should not be based on the results of one single “red flag” question.
• Uncritical use of “red flags” as a trigger to order further tests may lead to unnecessary investigations that are themselves harmful.
• The authors suggest considering the possibility of spinal malignancy (in light of its low prevalence in primary care) when several “red flags” are present.
• Further research to evaluate the performance of different combinations of tests is recommended.

Once again, low numbers of metastatic low back pain in primary care, and the usefulness of “single flags” is once again questioned. Using “flags” in combination sounds like a sensible idea, unfortunately there is little or no information on this at present.

Past history of cancer is something that seems to keep popping up as a reason to get immediate medical investigation. Whilst I don’t disagree with this, I have to say I have two queries regarding it:

1. How long are we talking about? 20 years ago, 10 years ago, 5 years ago, less than 5 years ago? I couldn’t find any information but will keep looking and update you if I do. Certainly I would expect that the more recent, the more important. But how recent?
2. When we look at the Henschke et al (2009) study, 46 of their patients had a history of cancer yet none actually had metastatic cancer. So really how important is someone having had a past history of cancer. The information seems conflicting. I agree it is worthwhile to be aware of, but I would think taking into account other factors (e.g. other cancer red flags) in conjunction with this, and the general clinical picture, is probably most critical.

Spinal Infection

I could find very little information investigating the signs of infection. The Henschke et al (2009) article did not identify any patients with infection so could not comment on the usefulness of the signs they had incorporated in their screening process. This is not surprising as its frequency is reported to be very low. Underwood (2009) reports the prevalence of spinal infection to be 0.01%.

So in relation to Infection, the following articles by Go et al (2012) and Hanck et al (2012) provide us with some excellent guidance:

1. Types of infections:

• Spondylodiscitis is an uncommon infection of the nucleus pulposus.
• Vertebral Osteomyelitis is when infection involves the adjacent cartilaginous endplates and vertebral bodies.

2. Signs and symptoms:

• Often has a slow and insidious onset.
• Patients typically present with back pain, tenderness, and rigidity at the site of involvement.
• Symptoms are usually not relieved with conservative therapy such as analgesics and bed rest.
• There may be a low-grade fever. Accompanying fever should alert clinicians to the possibility of infection.
• Paravertebral involvement and involvement of the neural foramina and exiting nerve roots may present as a radiculopathy or polyradiculopathy.
• Delay in the diagnosis can lead to development of, and subsequent worsening, neurologic deficits.
• The disks of the lumbar region are most frequently involved, followed by cervical spine and thoracic spine.

3. Risk Factors:

• Systemic infection (current or recent), immunocompromise, and recent penetrating trauma.

4. Investigations:

• Laboratory studies are of limited diagnostic utility.
• MRI is the optimal study, as it will demonstrate disk enhancement as well as adjacent soft tissue abscess or inflammation and can differentiate conditions in the differential diagnosis, such as neoplasia or tuberculosis of the spine

Granted I couldn’t find any systematic review regarding this topic but I feel the above provides a good level of up-to-date evidence based guidance. Please enlighten me if you know of any systematic reviews on this topic!

Inflammatory Back Pain (IBP)

Once again the Henschke et al (2009) study was unable to provide any guidance on their suggested signs/symptoms since only 2 patients presented with IBP. So lets look into some more recent research regarding this topic. Firstly, what is IBP?

Braun et al (2011) and Burgos-vargas and Braun (2012) summarise IBP via the following explanations and key points:

• The term spondyloarthritis (SpA) covers a group of rheumatic diseases characterised by common clinical symptoms such as inflammatory back pain (IBP). IBP has long been a central part of the classification criteria for AS  and SpA.
• Patients with SpA have been divided into two subgroups. Symptoms are either localised in the spine (axial SpA) or in the peripheral joints (peripheral SpA).
• Ankylosing spondylitis (AS), is the prototype of axial SpA.
• The other common differentiations used to diagnose or classify patients with SpA are the presence of a disease-defining feature such as psoriasis (psoriatic arthritis), inflammatory bowel disease (enteropathic arthritis) or the history of a triggering infection in the enteral or urogenital tract (reactive arthritis). Hence why Henschke et al (2009) have questions in their screening criteria relating to these disease defining features.
• In the absence of these features, the term ‘undifferentiated SpA’ (uSpA) has been used. While for patients with IBP but without structural changes in the sacroiliac joints and the spine, the term ‘non-radiographic axial SpA’ (nrSpA) is used.
• Not all patients with SpA will develop AS.
• Prevalence of 5% among patients with chronic back pain.
• Knowledge of the signs and symptoms that distinguish the inflammatory involvement of the spine from that of noninflammatory (mainly mechanical) conditions is fundamental for the identification of patients with AS and other SpAs in clinics.

Interestingly IBP it is not always listed as a Red Flag disorder [Duffy (2010)]. From my perspective I would consider it a potential Red Flag disorder given the possible complications that can develop. The following, from the American College of Rheumatology, is a list of some of the possible complications:

1. Inflammation of the eye (uveitis).
2. Inflammation of the aortic valve.
3. Intestinal inflammation (can be severe and require medical management).
4. Significant ankylosis of the spine.

Given Henschke et al (2009) could not comment on the usefulness of their IBP criteria I looked into some other criteria in the literature for identification of IBP. The following two articles provided me, and hopefully you, with some direction:

Sieper et al (2009) suggest the following criteria:

1. Age at onset 40 years or less.
2. Insidious onset.
3. Improvement with exercise.
4. No improvement with rest.
5. Pain at night (with improvement on getting up).

If at least four out of these five parameters were fulfilled, the criteria had a sensitivity of 77.0% and specificity of 91.7%.

Burgos-vargas and Braun (2012) suggest the following criteria for recognition of axial spondyloarthritis in primary care:

1. Age at onset less than (or equal) 35 years.
2. Waking up in the second half of the night.
3. Alternating buttock pain.
4. Improvement by NSAID’s within 48 hours.
5. Improvement by movement, not rest.

No single item was predictive, but ≥3 items proved useful for good sensitivity and specificity.

Cauda Equina Syndrome (CES)

Once again Henschke et al (2009) could not comment on the usefulness of the screening criteria in their study given only 1 patient presented with this diagnosis (0.08% of the patients). Underwood (2009) writes that:

Incidence of cauda equina syndrome is so low that most general (family) practitioners in the UK will not see a true case in their practicing lifetime.

Many of us who work in primary care will never see a case of cauda equina syndrome.

So whilst it is not common, have their been any recent developments in screening for CES in clinical practice? Pleasingly I tracked down a somewhat recent systematic review by Fairbank et al (2011) that investigated if patient history and physical examination predict MRI proven cauda equina syndrome.

In addition to similar symptoms the Henschke et al (2009) study used the authors also looked into the usefulness of symptoms and examination findings such as:

1. Rectal tone
2. Saddle numbness
3. Neurological Examination: Loss of power/sensation Increased or reduced reflexes.

Only 4 studies met the inclusion criteria, the key findings and conclusions were:

• All studies evaluated patients with symptoms suggestive of CES and compared symptoms and/or signs with findings at MRI.
• Insufficient data is present in the literature to identify signs or symptoms that correlated with either negative or positive MRI studies.
• The literature did not define objective, reliable clinical criteria for the diagnosis of CES.
• Prospective studies evaluating larger cohorts of patients are needed to more definitively determine whether any individual or combination of signs and/ or symptoms is associated with CES.

So it would appear, like a number of the other conditions above, we are lacking conclusive information about what to screen for to identify potential CES. Personally I would always err on the side of caution with CES. In my limited clinical experience with this presentation, and from speaking with other clinicians with experience in assessment of this presentation, one of the recurring/common signs seems to be saddle anaesthesia.

Overall Conclusions

Apart from IBP and Vertebral fractures, when we look deeply into the literature regarding screening for Red Flags we appear to be lacking some solid information about how to proceed in screening for these presentations. As many of the authors suggest, further research needs to look at the usefulness of a number of screening “flags” being grouped/clustered together.

We also need to be mindful that there are many reported case studies demonstrating red flag disorders that do not show any red flag signs, so we need to also use time as a guidance for investigating seemingly innocuous spinal pains that might appear to be mechanical in nature but are not improving with physiotherapy management. This then raises the question of “How long should be treat them for before investigating further?”. Most guidelines suggest no more than 4 to 6 weeks without demonstrable improvements. Clinical judgement/experience would not doubt come into this, as to would an earlier referral if symptoms were deteriorating. Personally I think 4 to 6 weeks is a long time to be treating someone without any improvement.

So how do we we proceed from here? Incorporating the above with the following guidelines, in my opinion, provides us with some reasonable direction. Whilst I realise these guidelines are specific to “what investigations to run/request” I think we can easily translate that across to “when we should refer for investigation”. The following guidelines, from the American College of Physicians, was summarised by Deyo, Jarvik and Chou (2014):

• Immediate Radiography and blood analysis of erythrocyte sedimentation rate (ESR) is recommended for patients with:
  • Major risk factors for cancer such as history of cancer, multiple risk factors (i.e. flags) for cancer, and/or a strong clinical suspicion of cancer. Should the radiography and ESR be negative, but a strong clinical suspicion of cancer remains, then proceeding to MRI is recommended.
• Immediate MRI is recommended for patients with:
  • Risk factors for spinal infection such as fever and history of injection drug use or recent infection.
  • Symptoms of caudal equina syndrome (new urinary retention, fecal incontinence, saddle anaesthesia)
  • Severe neurological deficits such as progressive motor weakness or motor deficits at multiple level neurologic levels)
• Deferment of radiography and +/- ESR until after a trial of treatment in patients with:
  • Weaker risk factors for cancer
  • Risk factors for inflammatory back pain
  • Risk factors for vertebral fracture
• Deferment of MRI until after a trial of treatment in patients with:
  • Signs of symptoms of radiculopathy or spinal stenosis
• No imaging is recommended for patients with:
  • Improving back pain or resolved back pain after 1 month.
  • Previous spinal imaging with no change in clinical status/presentation

In addition to the above Deyo, Jarvik and Chou (2014) also provide some input relating to recent guidelines from the UK based NICE (National Institute for Health and Care Excellence). The NICE guidelines suggest:

• Continually keeping a diagnosis under review.
• Do not offer x-ray of the lumbar spine for management of non-specific low back pain.
• Consider MRI when a diagnosis of spinal malignancy, infection, fracture, cauda equina syndrome or ankylosing spondylitis or another inflammatory disorder is suspected.
• Only offer an MRI scan for non-specific low back pain within the context of a referral for an opinion on spinal fusion.

I have to say of both “guidelines” the American College of Physicians approach seems more useful to me.

One interesting point is that neither the American College of Physicians and the NICE guidelines suggest immediate imaging of Vertebral fracture and IBP. Why might we not send patients with suspicion of vertebral fracture or IBP for immediate assessment? Granted we might send them for medical review for pain relief but why not immediate medical screening? I realise this is debatable topic and that given the nature of these disorder there is argument for immediate investigation. Having said that I thought I would share the thoughts of Underwood (2009) on these two disorders and why immediate imaging is, arguably, not essential:

In the absence of convincing evidence that vertebroplasty or kyphoplast are superior to medical management, there is little need for urgent investigation and referral to specialists.

Diagnosing and treating osteoporosis to prevent further fractures are important, and possibly more important than diagnosing the fracture itself, in persons at risk of a fragility fracture. In my opinion, there is little point in doing further investigations for most patients with known osteoporosis who present with a new episode of low back pain.

Now the above is a debatable statement, but when you look at systematic reviews, such as Robinson and Olerud (2012), regarding treatment of of vertebral fractures, we see that there is a lack of evidence that vertebroplasty or kyphoplasty are superior to medical (conservative) management. So Underwood makes a good point. Having said that lets be mindful that someone with a vertebral fracture could also have a cauda equina presentation subsequent to this injury and that would require immediate referral!

In relation to IBP Underwood (2009) suggests:

These are chronic disorders, the diagnosis of which is commonly delayed for several years. What is less clear is whether the majority of patients with mild disease are harmed by this delay, because many will be treated with non-steroidal anti-inflammatory drugs and advised to exercise irrespective of the diagnosis. The possibility of ankylosing spondylitis as a diagnosis needs to be considered, but only in patients who are not experiencing improvement after more than 3 months.

Probably another fair point. But again this is a debatable topic and one that I happy to hear more thoughts about.

Finally, in relation to radiculopathy and spinal stenosis, as O’Sullivan and Lin (2o14) wrote via this article from Pain-ed.com, many of these patients will improve with conservative treatment. Hence immediate imaging is not recommended until a trial of conservative management has been conducted.

Final Thoughts

I hope the above has been of some help. As can be seen the topic of screening for Red Flag presentations is not “cut and dry”. Far from it. There is however some good guidance in the literature.

Personally I feel that the guidelines summarised by Deyo, Jarvik and Chou (2014) are the best way to proceed at present. Having said that, clinical expertise and experience also comes strongly into the picture, as to does the patients thoughts and beliefs i.e. the complete “triad” of Evidence Based Practice.

Happy to hear some other thoughts.

Thanks for reading.