Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) of the Spine

Number: 0236

Policy

Aetna considers magnetic resonance imaging (MRI) and computed tomography (CT) of the spine medically necessary when any of the following criteria is met: 

• Clinical evidence of spinal stenosis; or
• Clinical suspicion of a spinal cord or cauda equina compression syndrome; or
• Congenital anomalies or deformities of the spine; or
• Evaluation of recurrent symptoms after spinal surgery; or
• Evaluation prior to epidural injection to rule out tumor or infection and to delineate the optimal anatomical location for performing the injection; or
• Follow-up of evaluation for spinal malignancy or spinal infection; or
• Known or suspected myelopathy (e.g., multiple sclerosis) for initial diagnosis when MRI of the brain is negative or symptoms mimic those of other spinal or brainstem lesions; or
• Known or suspected primary spinal cord tumors (malignant or non-malignant); or
• Persistent back or neck pain with radiculopathy as evidenced by pain plus objective findings of motor or reflex changes in the specific nerve root distribution, and no improvement after 6 weeks of conservative therapyFootnotes for conservative therapy^*; or
• Primary spinal bone tumors or suspected vertebral, paraspinal, or intraspinal metastases; or
• Progressively severe symptoms despite conservative management; or
• Rapidly progressing neurological deficit, or major motor weakness; or
• Severe back pain (e.g., requiring hospitalization); or
• Spondylolisthesis and degenerative disease of the spine that has not responded to 4 weeks of conservative therapyFootnotes for conservative therapy^*; or 
• Suspected infectious process (e.g., osteomyelitis epidural abscess of the spine or soft tissue); or
• Suspected spinal cord injury secondary to trauma; or
• Suspected spinal fracture and/or dislocation secondary to trauma (if plain films are not conclusive); or
• Suspected transverse myelitis.

Aetna considers MRI and CT of the spine experimental and investigational for all other indications because their clinical value for indications other than the ones listed above has not been established.  Clinical guidelines, including those from the Agency for Healthcare Policy and Research, have consistently recommended against routine imaging studies for acute low back pain (Cho et al, 2009).

Aetna considers the use of MRI for further evaluation of unstable injury in neurologically intact individuals with blunt trauma after a negative cervical spine CT result not medically necessary.

Aetna considers dynamic-kinetic MRI experimental and investigational for evaluation of the cervical spine because its effectiveness has not been established.

Aetna considers the use of routine MRI after a normal CT of the cervical spine in obtunded or comatose individuals experimental and investigational because the clinical value of this approach has not been established.

Footnotes for conservative therapy^* Conservative therapy = moderate activity, analgesics, non-steroidal anti-inflammatory drugs, muscle relaxants. 

Background

Because of its complexity, the spine is probably the most difficult part of the skeletal system to evaluate radiologically.  Improvement of computed tomography (CT) scanners and the advent of magnetic resonance imaging (MRI) have changed the approach to diagnostic imaging of the spine.  Previously, invasive modalities were required to obtain information that is now available with non-invasive technologies.

The appropriate use of these new technologies is still somewhat unsettled.  The focus is on which test will provide the most accurate and cost effective diagnostic information for each particular clinical situation.  Computed tomographic scan, CT myelography, MRI and plain radiography all have their place in the diagnostic work-up of problems related to the spine.

Bulging intervertebral discs have been found in over half of all otherwise asymptomatic adults.  It is therefore, important to perform MRI or CT at the right time and to interpret the results in the context of the clinical findings to ensure an accurate diagnosis and avoid unnecessary treatment of conditions that may not be the cause of a patient's symptoms.

1. suspected spinal fracture or dislocation due to trauma, where CT scan is the preferred method of imaging if plain films are inconclusive, and
2. evaluation of a patient with signs or symptoms of spinal stenosis, where MRI or CT are equally appropriate. For evaluation of recurrent symptoms after spinal surgery, MRI with and without gadolinium enhancement, is the preferred method of imaging.

Magnetic resonance imaging or CT evaluation of chronic mechanical low back pain (LBP) without radiculopathy or neurologic deficit, trauma, or clinical suspicion of systemic disorder (e.g., infectious process, metastatic disease) is not necessary unless back pain is severe (e.g., requiring hospitalization) or where symptoms are progressing despite conservative management (ICSI, 2002).

The American College of Physicians (2012) has recommended against obtaining imaging studies in patients with non-specific low back pain. In patients with back pain that cannot be attributed to a specific disease or spinal abnormality following a history and physical examination (e.g., non-specific low back pain), imaging with plain radiography, computed tomography (CT) scan, or magnetic resonance imaging (MRI) does not improve patient outcomes. The American Academy of Family Physicians (2012) recommends against do imaging for low back pain within the first six weeks, unless red flags are present. Red flags include, but are not limited to, severe or progressive neurological deficits or when serious underlying conditions such as osteomyelitis are suspected. Imaging of the lower spine before six weeks does not improve outcomes, but does increase costs. Low back pain is the fifth most common reason for all physician visits. The North American Spine Society (2013) has issued similar recommendations.

Cho et al (2009) reported the results of a systematic review and meta-analysis of imaging strategies for LBP without indications of serious underlying conditions.  Inclusion criteria were randomized controlled trials that compared immediate, routine lumbar imaging (or routine provision of imaging findings) versus usual clinical care without immediate lumbar imaging (or not routinely providing results of imaging) for LBP without indications of serious underlying conditions.  Primary outcomes were improvement in pain or function.  Secondary outcomes were improvement in mental health, quality of life, patient satisfaction, and overall improvement.  Outcomes were categorized as short-term (less than or equal to 3 months), long-term (greater than 6 months to less than or equal to 1 year), or extended (greater than 1 year).  A total of 6 trials met the inclusion criteria: 4 assessed lumbar radiography and 2 assessed MRI or CT.  Duration of follow-up ranged from 3 weeks to 2 years.  One trial excluded patients with sciatica or other symptoms of radiculopathy, and 1 did not report the proportion of patients with such symptoms.  In the other 4 trials, the proportion of patients with sciatica or radiculopathy ranged from 24 % to 44 %.  Three trials compared immediate lumbar radiography with usual clinical care without immediate lumbar radiography, and 1 compared immediate lumbar radiography with a brief education intervention plus lumbar radiography, if no improvement was seen by 3 weeks.  Patients (n = 1,804) enrolled in these trials had mainly acute or subacute (less than 12 weeks) LBP, and all trials were done in primary-care or urgent-care settings. Two studies assessed advanced imaging modalities.  One study compared immediate MRI or CT with usual clinical care without advanced imaging in patients with mainly chronic LBP (82 % had LBP for greater than 3 months) referred to a surgeon, whereas in the other study all patients with LBP for less than 3 weeks underwent MRI, with randomization to routine notification of results within 48 hours versus notification of results only if clinically indicated.  Patients were recruited from various settings (primary care, spine clinic, or emergency room).  In both trials, the proportion of patients who underwent lumbar radiography before enrollment was not reported.  The most frequent methodological shortcoming was lack of (or unclear use of) blinded outcome assessment (5 of 6 trials), followed by inadequate description of randomization method (4 of 6 trials).  All trials excluded patients with features suggestive of a serious underlying condition, but exclusion criteria varied and trials did not indicate the number of patients excluded because of such factors.  The authors found no significant difference between routine, immediate lumbar imaging and usual clinical care without immediate imaging for improvement in pain or function at short-term or long-term follow-up.  In the trial that reported extended (2-year) follow-up data, immediate MRI or CT was not better than usual clinical care without immediate imaging on either the EuroQol-5D (mean difference 0.02, 95 % confidence interval: -0.02 to 0.07, 0 to 1 scale) or the SF-36 mental health score (-1.50, -4.09 to 1.09, 0 to 100 scale) in unadjusted analyses.  The authors concluded that lumbar imaging forLBP without indications of serious underlying conditions does not improve clinical outcomes and that clinicians should refrain from routine, immediate lumbar imaging in patients with acute or subacute LBP and without features suggesting a serious underlying condition.

In a meta-analysis, Schoenfeld et al (2010) examined if adding an MRI would provide useful information that alters treatment when a CT scan reveals no evidence of injury in obtunded blunt trauma patients.  Published studies from 2000 to 2008 involving patients undergoing MRI for the purposes of further cervical spine evaluation after a "negative" CT scan were identified via a literature search of online databases.  Data from eligible studies were pooled and original scale meta-analyses were performed to calculate overall sensitivity, specificity, positive and negative predictive values, likelihood ratios, and relative risk.  The Q-statistic p value was used to evaluate heterogeneity.  A total of 11 studies met the inclusion criteria, yielding data on 1,550 patients with a negative CT scan after blunt trauma subsequently evaluated with a MRI.  The MRI detected abnormalities in 182 patients (12 %).  Ninety traumatic injuries were identified, including ligamentous injuries (86/182), fractures and dislocations (4/182).  In 96 cases (6 % of the cohort), the MRI identified an injury that altered management.  Eighty-four patients (5 %) required continued collar immobilization and 12 (1 %) required surgical stabilization.  The Q-statistic p value for heterogeneity was 0.99, indicating the absence of heterogeneity among the individual study populations.  The authors concluded that reliance on CT imaging alone to "clear the cervical spine" after blunt trauma can lead to missed injuries.  The findings of this study supported the addition of MRI in evaluating patients who are obtunded, or unexaminable, despite a negative CT scan.

Callaghan et al (2012) examined diagnostic practice patterns as an early step in identifying opportunities to improve efficiency of care of patients with peripheral neuropathy.  The 1996 to 2007 Health and Retirement Study Medicare claims-linked database was used to identify individuals with an incident diagnosis of peripheral neuropathy using International Classification of Diseases, Ninth Revision, codes and required no previous neuropathy diagnosis during the preceding 30 months.  Focusing on 15 relevant tests, these investigators examined the number and patterns of tests and specific test utilization 6 months before and after the incident neuropathy diagnosis.  Medicare expenditures were assessed during the baseline, diagnostic, and follow-up periods.  Of the 12,673 patients, 1,031 (8.1 %) received a new International Classification of Diseases, Ninth Revision, diagnosis of neuropathy and met the study inclusion criteria.  Of the 15 tests considered, a median of 4 (interquartile range, 2 to 5) tests were performed, with more than 400 patterns of testing.  Magnetic resonance imaging of the brain or spine was ordered in 23.2 % of patients, whereas a glucose tolerance test was rarely obtained (1.0 %).  Mean Medicare expenditures were significantly higher in the diagnostic period than in the baseline period ($14,362 versus $8,067, p < 0.001).  The authors concluded that patients diagnosed as having peripheral neuropathy typically undergo many tests, but testing patterns are highly variable.  Almost 25 % of patients receiving neuropathy diagnoses undergo high-cost, low-yield MRI, whereas few receive low-cost, high-yield glucose tolerance tests.  Expenditures increase substantially in the diagnostic period.  The authors stated that more research is needed to define effective and efficient strategies for the diagnostic evaluation of peripheral neuropathy.

Also, an UpToDate review on "Overview of polyneuropathy" (Rutkove, 2012) does not mention the use of MRI or CT in the diagnostic evaluation of individuals with polyneuropathy.

The Institute for Clinical Systems Improvement clinical practice guideline on “Adult acute and subacute low back pain” (ICSI, 2012) stated that  imaging (CT, MRI, or x-ray) is not recommended for non-specific low-back pain [strong recommendation, moderate quality evidence].

el Barzouhi et al (2013) noted that MRI is frequently performed during follow-up in patients with known lumbar-disk herniation and persistent symptoms of sciatica.  The association between findings on MRI and clinical outcome is controversial.  These investigators studied 283 patients in a randomized trial comparing surgery and prolonged conservative care for sciatica and lumbar-disk herniation.  Patients underwent MRI at baseline and after 1 year.  These researchers used a 4-point scale to assess disk herniation on MRI, ranging from 1 for "definitely present" to 4 for "definitely absent".  A favorable clinical outcome was defined as complete or nearly complete disappearance of symptoms at 1 year.  These investigators compared proportions of patients with a favorable outcome among those with a definite absence of disk herniation and those with a definite, probable, or possible presence of disk herniation at 1 year.  The area under the receiver-operating-characteristic (ROC) curve was used to assess the prognostic accuracy of the 4-point scores regarding a favorable or unfavorable outcome, with 1 indicating “perfect discriminatory value” and 0.5 or less indicating “no discriminatory value”.  At 1 year, 84 % of the patients reported having a favorable outcome.  Disk herniation was visible in 35 % with a favorable outcome and in 33 % with an unfavorable outcome (p = 0.70).  A favorable outcome was reported in 85 % of patients with disk herniation and 83 % without disk herniation (p = 0.70).  Assessment of disk herniation by means of MRI did not distinguish between patients with a favorable outcome and those with an unfavorable outcome (area under ROC curve, 0.48).  The authors concluded that MRI performed at 1-year follow-up in patients who had been treated for sciatica and lumbar-disk herniation did not distinguish between those with a favorable outcome and those with an unfavorable outcome.  Moreover, they stated that further research is needed to evaluate the value of MRI in clinical decision-making for patients with persistent or recurrent sciatica.

Steffens et al (2014) systematically reviewed whether MRI findings of the lumbar spine predict future LBP in different samples with and without LBP.  MEDLINE, CINAHL and EMBASE databases were searched.  Included were prospective cohort studies investigating the relationship between baseline MRI abnormalities of the lumbar spine and clinically important LBP outcome at follow-up.  These researchers excluded cohorts with specific diseases as the cause of their LBP.  Associations between MRI findings and LBP pain outcomes were extracted from eligible studies.  A total of 12 studies met the inclusion criteria; 6 studies presented data on participants with current LBP; 1 included a sample with no current LBP, 3 included a sample with no history of LBP and 2 included mixed samples.  Due to small sample size, poor overall quality and the heterogeneity between studies in terms of participants, MRI findings and clinical outcomes investigated, it was not possible to pool findings.  No consistent associations between MRI findings and outcomes were identified.  Single studies reported significant associations for Modic changes type 1 with pain, disc degeneration with disability in samples with current LBP and disc herniation with pain in a mixed sample.  The authors concluded that the limited number, heterogeneity and overall quality of the studies do not permit definite conclusions on the association of MRI findings of the lumbar spine with future LBP

Weber et al (2015) evaluated the incremental diagnostic value of spine MRI evaluated separately from and combined with sacroiliac joint (SIJ) MRI in non-radiographic axial spondyloarthritis (nr-axSpA) compared with SIJ MRI alone.  The study sample comprised 2 independent cohorts A/B of 130 consecutive patients aged less than or equal to 50 years with back pain, newly referred to 2 university clinics, and 20 healthy controls.  Patients were classified according to clinical examination and pelvic radiographs as having nr-axSpA (n = 50), ankylosing spondylitis (n = 33), or non-specific back pain (n = 47).  Four readers assessed SIJ and spine MRI separately 6 months apart, and 1 to 12 months later both scans simultaneously using standardized modules.  Readers recorded presence/absence of SpA and their level of confidence in this conclusion on a 0 to 10 scale (0 = definitely not; 10 = definite).  These researchers analyzed differences between SIJ MRI versus spine MRI alone, and SIJ MRI alone versus combined MRI, descriptively by the number/percentage of subjects according to the mean of 4 readers.  In cohorts A/B, 15.8 %/24.2 % of patients with nr-axSpA having a negative SIJ MRI were re-classified as being positive for SpA by global evaluation of combined scans.  However, 26.8 %/11.4 % of non-specific back pain controls and 17.5 % of healthy volunteers with a negative SIJ MRI were falsely re-classified as having SpA by combined MRI.  Low confidence in a diagnosis of SpA by SIJ MRI increased to high confidence by combined MRI in 6.6 %/7.3 % of patients with nr-axSpA.  The authors concluded that combined spine and SIJ MRI added little incremental value compared with SIJ MRI alone for diagnosing patients with nr-axSpA and enhancing confidence in this diagnosis.

On behalf of the Tufts Medical Center Evidence-based Practice Center, Dahabreh and colleagues (2011) performed a systematic review of emerging MRI technologies for musculoskeletal imaging under loading stress for the Agency for Healthcare Research and Quality (AHRQ).  The review included 57 studies about MRI under physiologic loading stress performed in an upright or sitting position or under axial loading by using a compression device.  The most commonly imaged regions were the spine (33 studies) and knee (13 studies).  Most studies had a cross-sectional (n = 37) or case-control (n = 13) design and reported on anatomical measurements rather than patient-relevant end points.  Studies were generally small: The median (25th, 75th percentile) number of case patients was 26 (17, 45), and the median (25th, 75th percentile) number of control participants was 13 (12, 20 for case-control studies).  Fifteen of 57 studies used at least 2 imaging tests and reported on diagnostic or patient-relevant outcomes, but did not report meaningful information on the relative performance of the tests.  In 10 studies that included information on adverse effects, 5 % to 15 % of participants reported new-onset or worsening pain and neuropathy during MRI under loading stress.  The authors concluded that available evidence is insufficient to support the clinical utility of MRI under loading stress for musculoskeletal conditions.

Lord et al (2014) reviewed the body of literature related to kinetic MRI (kMRI) of the cervical spine.  A review of literature related to kMRI was performed using PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines.  These researchers included 16 prospective and retrospective studies of symptomatic and asymptomatic patients who underwent kMRI of the cervical spine.  The authors concluded that data suggested that kMRI is able to provide meaningful information regarding changes in the cervical spine in both normal and pathologic segments.  Moreover, they stated that a prospective study comparing MRI and kMRI is needed to confirm clinically utility of this technology.

Also, an UpToDate review on “Evaluation of the patient with neck pain and cervical spine disorders” (Isaac and Anderson, 2014) states that “Magnetic resonance imaging (MRI) should be the first-line imaging study performed in patients with progressive signs or symptoms of neurologic disease.  MRI should also be obtained if there is a suspicion for infection or malignancy and if there are moderate to severe neck symptoms beyond six weeks, even if plain films are negative ….  A non-contrast MRI is sufficient in the majority of cases.  The addition of gadolinium contrast intravenously allows better diagnosis of infection, tumor, or post-surgical epidural fibrosis, and can be ordered subsequently if the non-contrast study is inconclusive”.  It does not mention the use of dynamic-kinetic MRI.

In a systematic review, Suri and colleagues (2015) examined if lumbar muscle characteristics on MRI or CT can inform clinicians as to the course of future LBP, functional limitations, or physical performance, in adults with or without LBP.  These investigators searched PubMed, Embase, and CINAHL through October 2014 for articles published in English in which authors assessed lumbar muscle characteristics on conventional MRI/CT as predictors of future LBP, functional limitations, or physical performance in adults.  Studies with only post-surgical subjects were excluded.  The search identified 3,554 articles, of which 6 observational cohort studies were included in the final review.  These researchers used the Newcastle Ottawa Scale to evaluate potential bias.  Data were extracted on study design, study population, sample size, participant characteristics, details of MRI/CT assessments, interventions, study outcomes, analysis methods, and study results.  Because of heterogeneity between studies, these researchers conducted a qualitative evidence synthesis.  Among high-quality studies, there was limited evidence that, for individuals with or without LBP, greater MRI-detected multifidus cross-sectional area at L5 to S1 predicted greater LBP intensity at 1-year follow-up, lesser erector spinae fat infiltration (FI) at L5 to S1 predicted greater LBP intensity at 15-year follow-up, and greater erector spinae side-to-side FI asymmetry at L3 to L4 predicted lower LBP frequency at 15-year follow-up; however, there was also limited evidence that all other MRI-detected para-spinal muscle characteristics examined were not predictive of LBP incidence, prevalence, frequency, or intensity at follow-up durations ranging from 1 to 15 years.  There was limited evidence that greater CT-detected trunk muscle FI predicted worse physical performance in older adults at 3-year follow-up, but that trunk muscle cross-sectional area did not.  The authors concluded that few lumbar muscle characteristics have limited evidence for an association with future LBP and physical performance outcomes, and the vast majority have limited evidence for having no association with such outcomes.

Routine MRI after a Normal CT of the Cervical Spine in Obtunded or Comatose Individuals:

1. obtundation (Glasgow Coma Scale less than or equal to 13, with 94 of this group comatose [Glasgow Coma Scale less than or equal to 8]);
2. no obvious neurologic deficits; and
3. a normal cervical CT.

The effect of MRI on the clinical management of these patients was evaluated.  Among the 150 obtunded or comatose patients with a negative CT, the majority (51 %) had a normal MRI.  Among the patients with a positive MRI, the most common MRI-positive findings were ligamentous and soft tissue injury (81 %).  However, no MRI findings were deemed unstable, and no surgical intervention or change in the clinical management aside from collar immobilization of these individuals occurred after MRI.  The authors concluded that the addition of a cervical MRI to the evaluation protocol of obtunded or comatose patients with an otherwise normal neurologic examination and a normal cervical CT did not provide any additional useful information to change the management of these patients.

Raza and co-workers (2013) stated that a true gold standard to rule out a significant cervical spine injury in subset of blunt trauma patients with altered sensorium is still to be agreed upon.  These investigators examined if in obtunded adult patients with blunt trauma, a clinically significant injury to the cervical spine be ruled out on the basis of a normal multi-detector cervical spine CT.  Comprehensive database search was conducted to include all the prospective and retrospective studies on blunt trauma patients with altered sensorium undergoing cervical spine multi-detector CT scan as core imaging modality to "clear" the cervical spine.  The studies used 2 main gold standards, MRI of the cervical spine and/or prolonged clinical follow-up.  The data was extracted to report true positive, true negatives, false positives and false negatives.  Meta-analysis of sensitivity, specificity, negative and positive predictive values (NPV and PPV) was performed using Meta Analyst Beta 3.13 software.  These researchers also performed a retrospective investigation comparing a robust clinical follow-up and/or cervical spine MRI findings in 53 obtunded blunt trauma patients, who previously had undergone a normal multi-detector CT scan of the cervical spine reported by a radiologist.  A total of 10 studies involving 1,850 obtunded blunt trauma patients with initial cervical spine CT scan reported as normal were included in the final meta-analysis.  The cumulative NPV and specificity of cervical spine CT of the 10 studies was 99.7 % (95 % confidence interval [CI]: 99.4 to 99.9 %).  The PPV and sensitivity was 93.7 % (95 % CI: 84.0 to 97.7 %).  In the retrospective review of obtunded blunt trauma patients, none was later diagnosed to have significant cervical spine injury that required a change in clinical management.  The authors concluded that in a blunt trauma patient with altered sensorium, a normal cervical spine CT scan was conclusive to safely rule out a clinically significant cervical spine injury.  They stated that the findings of this meta-analysis strongly supported the removal of cervical precautions in obtunded blunt trauma patient after normal cervical spine CT; any further imaging like MRI of the cervical spine should be performed on case-to-case basis.

Smith (2014) addressed the question "Can CT alone provide adequate clinical information to clear the cervical spine in the obtunded patient"?  The author performed a search of the literature for studies that compared CT with other radiologic modalities utilized to clear the cervical spine in obtunded patients.  PubMed, TRIP database, SUMSearch, Cochrane library, and Google Scholar were the databases applied.  Additional sources included bibliographies of selected articles.  Studies that integrated CT scan with at least 1 other diagnostic examination were included.  A review of 11 studies and 1 meta-analysis encompassing 2,458 and 14,327 patients, respectively, met inclusion criteria.  The meta-analysis generated a NPV for CT scan of 100 % without evidence of acute injury with an overall sensitivity and specificity of 99.9 % each.  The author concluded that these findings suggested that CT alone is a reliable clinical indicator to clear the cervical spine in obtunded patients.

Patel and associated (2015) noted that with the use of the framework advocated by the Grading of Recommendations Assessment, Development and Evaluation (GRADE) Working Group, they performed a systematic review and to develop evidence-based recommendations that may be used to answer the following PICO [Population, Intervention, Comparator, Outcomes] question: In the obtunded adult blunt trauma patient, should cervical collar removal be performed after a negative high-quality cervical spine (C-spine) CT result alone or after a negative high-quality C-spine CT result combined with adjunct imaging, to reduce peri-clearance events, such as new neurologic change, unstable C-spine injury, stable C-spine injury, need for post-clearance imaging, false-negative CT imaging result on re-review, pressure ulcers, and time to cervical collar clearance?  The protocol was registered with the PROSPERO international prospective register of systematic reviews on August 23, 2013.  Eligibility criteria consisted of adult blunt trauma patients 16 years or older, who underwent C-spine CT with axial thickness of less than 3 mm and who were obtunded using any definition.  Quantitative synthesis via meta-analysis was not possible because of pre-post, partial-cohort, quasi-experimental study design limitations and the consequential incomplete diagnostic accuracy data.  Of 5 articles with a total follow-up of 1,017 included subjects, none reported new neurologic changes (paraplegia or quadriplegia) after cervical collar removal.  There was a worst-case 9 % (161 of 1,718 subjects in 11 studies) cumulative literature incidence of stable injuries and a 91 % NPV of no injury, after coupling a negative high-quality C-spine CT result with 1.5-T MRI, upright x-rays, flexion-extension CT, and/or clinical follow-up.  Similarly, there was a best-case 0 % (0 of 1,718 subjects in 11 studies) cumulative literature incidence of unstable injuries after negative initial imaging result with a high-quality C-spine CT.  The authors concluded that in obtunded adult blunt trauma patients, they conditionally recommended cervical collar removal after a negative high-quality C-spine CT scan result alone.

Plackett and colleagues (2016) noted that the role of cervical spine MRI in the evaluation of clinically unevaluable blunt trauma patients has been called into question by several recent studies.  These investigators performed a PubMed search for all studies comparing CT and MRI in the assessment of the cervical spine in patients who cannot be evaluated clinically.  The radiologic findings and clinical outcomes from each study were collated for analysis.  Data for 1,714 patients were available.  All patients had a negative CT scan and then underwent an MRI.  There were 271 (15.8 %) patients who had a previously undocumented finding on MRI with the majority (98.2 %) being a ligamentous injury.  Only 5 injuries (1.8 %) resulted in surgical intervention.  The authors concluded that MRI identified additional injuries; however, the vast majority were of minor clinical significance.  They stated that routine MRI after a negative CT of the cervical spine is not supported by the current literature.

Dynamic Supine Magnetic Resonance Imaging:

Xu and colleagues (2017) analyzed the current evidence regarding the role of dynamic supine MRI (dsMRI) in the evaluation of cervical spondylotic myelopathy.  A total of 13 studies were identified through a comprehensive literature search performed in the PubMed, Embase, and ISI databases as fulfilling the inclusion criteria and were reviewed for subject characteristics, radiographic parameters, and salient findings.  Studies reviewed suggested that dsMRI was able to detect new appearance or increased grade of medullary compression in greater than or equal to 20 % of patients and to demonstrate an average narrowing of the cervical canal by 20 % (in comparison with the neutral position).  Several additional parameters were investigated, but their clinical significance remained unconfirmed; 2 studies examined how surgical decision-making could be affected by the additional findings of dsMRI.  The authors concluded that dsMRI represents an available modification of conventional static MRI and is potentially able to demonstrate pathologies that might be previously missed.  They stated that evidence suggested that dsMRI can elucidate spinal cord compression with higher sensitivity, resulting in improved diagnostic accuracy of cervical spondylotic myelopathy, which may impact surgical planning for these patients; however, more high-quality studies are needed to further establish its indications to avoid over-diagnosis with this powerful imaging technique.

Magnetic Resonance Imaging in Cervical Spine Clearance of Neurologically Intact Patients With Blunt Trauma:

In a meta-analysis, Malhotra and co-workers (2017) quantified the rate of unstable injuries detected by MRI missed on CT in blunt cervical spine (CS) trauma patients and evaluated the utility of MRI in CS clearance.  These researchers undertook a systematic review of worldwide evidence across 5 major medical databases.  Studies were included if they reported the number of unstable injuries or gave enough details for inference.  Variables assessed included severity, CT/MRI specifications, imaging timing, and outcome/follow-up.  Pooled incidences of unstable injury on follow-up weighted by inverse-of-variance among all included and obtunded or alert patients were reported.  Of 428 unique citations, 23 proved eligible, with 5,286 patients found, and 16 unstable injuries reported in 5 studies.  The overall pooled incidence was 0.0029 %.  Among studies reporting only obtunded patients, the pooled incidence was 0.017 %. In alert patients, the incidence was 0.011 %.  All reported positive findings were critically reviewed, and only 11 could be considered truly unstable.  The authors concluded that there was significant heterogeneity in the literature regarding the use of imaging after a negative CT.  The finding rate on MRI for unstable injury was extremely low in obtunded and alert patients.  They stated that although MRI is frequently performed, its utility and cost-effectiveness needs further study.  Key points of this meta-analysis included the following -- There were 16 unstable injuries on follow-up MRI among 5,286 patients.  The positive finding rate among obtunded patients was 0.12 %.  The positive finding rate among alert, awake patients was 0.72 %.  MRI has a high false-positive rate; its utility mandated further studies.  The use and role of "confirmatory" tests showed wide variations.

Wu and associates (2018) noted that use of MRI for cervical clearance after a negative cervical CT scan result in alert patients with blunt trauma who are neurologically intact is not infrequent, despite poor evidence in regard to its utility.  These investigators evaluated the utility and cost-effectiveness of using MRI versus no follow-up in this patient population.  A modeling-based decision analysis was performed during the lifetime of a 40-year old individual from a societal perspective.  The 2 strategies compared were no follow-up and MRI.  A Markov model with a 3 % discount rate was used with parameters from the literature.  Base cases and probabilistic and sensitivity analyses were performed to assess the cost-effectiveness of the strategies.  The cost of MRI follow-up was $11,477, with a health benefit of 24.03 quality-adjusted life-years (QALY); the cost of no follow-up was $6,432, with a health benefit of 24.08 QALY.  No follow-up was the dominant strategy, with a lower cost and a higher utility.  Probabilistic sensitivity analysis showed no follow-up to be the better strategy in all 10,000 iterations.  No follow-up was the better strategy irrespective of the NPV of initial CT result, and it remained the better strategy when the incidence of missed unstable injury resulting in permanent neurologic deficits was less than 64.2 % and the incidence of patients immobilized with a hard collar who still received cord injury was greater than 19.7 %.  Multiple 3-way sensitivity analyses were performed.  The authors concluded that MRI is not cost-effective for further evaluation of unstable injury in neurologically intact patients with blunt trauma after a negative cervical spine CT result.

The above policy is based on the following references: