Three Part Question
In [adult patients with minor head injury] how can [protein S-100B integrate with NICE guidelines or other CT clinical decision rules] in [detecting significant traumatic brain injuries]?
A 25-year-old healthy man presented to a semi-urban trauma unit with a minor head injury (MHI). After thorough assessment and following UK National Institute for Health and Clinical Excellence (NICE) guidelines, you ascertain that the patient does not require CT brain scan and can be safely discharged home. A neurosurgical colleague, having recently returned from a European conference, had adopted protein S100B sampling as a neuroimaging triage tool for the department. He suggests this is a more accurate measurement of the extent of brain injury. On following his departmental policy serum S100B sampling yielded a positive result, but subsequent neuroimaging reveals no intracranial pathology. You wonder what the clinical utility of protein S100B actually is in MHI and also whether integration of this blood test within clinical guidelines might provide improved, more cost-effective care?
Week beginning 1 May 2013.
A primary literature search was conducted using key databases: the National Library for Health, MEDLINE, CINAHL and EMBASE. Secondary searches were carried out using the Cochrane Library, King\'s Fund Library (OVID), the National Research Register, ClinicalTrials.gov, WHO International Clinical Trials registry platform, BestBETs (http://www.bestbets.org) and GoogleScholar (http://scholar.google.co.uk/). Databases provided were through NLH search 2.0 CINAHL (EBSCO), EMBASE (OVID), KINGS FUND (OVID) and MEDLINE (OVID). The search was limited to the past 13 years.
The following search terms were used in various combinations: [Protein OR serum] AND [S100 OR S-100B OR S100B OR *S100] AND [*Minor head injury OR minor craniocerebral trauma OR minor head trauma OR mild traumatic brain injury OR acute brain injury OR acute brain damage] AND [NICE OR CT OR CT triage OR Canadian CT head rules OR New Orleans criteria].
Results of the primary literature search revealed 35 relevant papers. These were manually filtered after reading the article abstracts and removing duplicates. Scanning the bibliographies of identified studies and undertaking a secondary database search complemented this exercise. Overall, 21 papers were identified: 17 prospective cohorts, one prospective validation study, one invited review and two systematic reviews with meta-analytical interpretation. All 21 papers ascertained the diagnostic accuracy of protein S100B by quoting conventional measures. Twenty manuscripts were retrieved and critically appraised utilising the quality assessment of diagnostic accuracy studies checklist. Eleven of these papers were included in one or both of the published meta-analyses. No published literature explored the integration of protein S100B with UK NICE guidelines. The eight remaining papers are shown in the table
|Author, date and country
||Study type (level of evidence)
|Zongo et al,|
|1559 Consecutive adults with MHI and 1/10 inclusion criteria
Sampled within 6 h
CT positive rate 7% ||Diagnostic study
||Clinical utility S100B <0.12 μg/l||Sensitivity 99% Specificity 12% NPV 99%||Single, tertiary, neurosurgical centre
QUADAS 13/14 |
|Cervellin et al,|
|60 Consecutive adults with MHI; S100B and CCT within 3 h
CT positive rate 33% ||Diagnostic study
||Clinical utility S100B <0.38 μg/||Sensitivity 100% Specificity 58% AUC 0.8||CT within 30 min of S100B may not be reproducible in practice
Small sample size,
Selection bias (high pre-test prevalence)
QUADAS score 11/14 |
|Kotlyar et al,|
|346 Adults with MHI sampled within 6 h of injury
CT positive rate 6.4% ||Diagnostic study
||Clinical utility S100B <0.24 μg/l||Sensitivity 96% Specificity 13% AUC 0.64||Utilised higher cut-off for S100B in comparison to other studies
QUADAS score 12/14 |
|Leidel et al,|
|Adults with MHI
two reviewers utilising QUADAS tools reviewed 76 studies of which eight are included in the final analysis ||Systematic literature review and meta-analysis
||Clinical utility||Sensitivity 94% (95% CI 88 to 98%)||Marked heterogeneity
Subgroup analysis showed significant differences |
|Muller et al,|
|233 Consecutive adults with MHI and clear inclusion criteria cut-off
Stratified three groups A, B, C, based on clinical, CT and S100B results
CT positive rate 9% ||Diagnostic study
Prospective cohort||Clinical utility S100B <0.105 μg/l||Sensitivity 86% Specificity 12% PPV 12.8%||No mention of independent reference/index test
QUADAS score 12/14 |
|Egea-Guerrero et al,|
|143 Adult patients with MHI and 1/5 inclusion criteria. GCS 15/15 only and excluded patients with LOC
Sampled within 6 h of MHI and CT within 24 h
CT positive rate 10.5% ||Diagnostic study
Prospective cohort||Clinical utility S100B <0.13 μg/l||Sensitivity 100% Specificity 32.8%||Small sample size
Unclear on reporting of withdrawals
QUADAS score 12/14 |
|Unden and Romner,|
|12 Eligible articles involving 2466 adult patients with MHI
Quality assessment using CEBM criteria and QUADAS criteria
Various cut-offs for S100B explored ||Systematic review and meta-analysis||Clinical utility||Sensitivity 97% (95% CI 91 to 99%)||Marked heterogeneity in studies
Search strategy not entirely explicit
No inter-rater reliability measured. |
|Calcagnile et al,|
|512 Consecutive adults with MHI
GCS 14/15 with LoC and/or amnesia and no additional risk factors
S100B sampling within 3 h
Incorporated into modified SNC guidelines
Composite outcome determined by CT/questionnaire/telephonic follow-up/review of medical records and review of Swedish National database of patients up to 3 months post-trauma
CT positive rate 4.7% ||Prospective cohort
Clinical validation study ||Clinical utility S100B <0.1||Sensitivity 100% Specificity 28% PPV 6% NPV 100%||Reference test/gold standard not applied to all study participants
Low compliance by physicians with guideline
Only 32% had neuroimage
No blinding of radiologists
No comment on inter-rater reliability
QUADAS score 9/14 |
The published work in this field consists mostly of studies with small samples, limited to single centres and with diverse methodological quality. Patient numbers vary greatly between studies (range 501559) with only five studies recruiting more than 200 patients. In addition, there were marked variations in sample analysis techniques and the thresholds used to define a positive S100B result. The pooled estimate for the prevalence of intracranial injury detectable on CT was 12.6% (range 4.733%); this observed heterogeneity in a positive neuroimaging rate is important as the range is much wider than that reported in the MHI population.
Overall, it seems that S100B has a high sensitivity for intracranial pathology, but a poor specificity. It may have a role to play as a rule-out tool, but its poor specificity is a limiting factor because use in a low prevalence population will result in a low positive predictive value.
There is no real evidence to assess the use of S100B as an integral part of a clinical decision rule.
AUC, area under the curve; CCT, cranial CT; GCS, Glasgow coma scale; LOC, loss of consciousness; MHI, minor head injury; NPV, negative predictive value; PPV, positive predictive value; QUADAS, quality assessment of diagnostic accuracy studies; SNC, Scandinavian Neurotrauma Committee.
Clinical Bottom Line
S100B is a sensitive but non-specific test for traumatic brain injury. However, there is no evidence to determine whether it adds value to any current clinical guidelines.
- Zongo D, Ribereau-Gayon R, Masson F, et al. S-100B Protein as a screening tool for the early assessment of minor head injury Ann Emerg Med. 2012;59.
- Cervellin G, Benatti M, Carbucicchio A, et al. Serum levels of protein S100B predict intracranial lesions in mild head injury. Clin Biochem 2012;45:40811.
- Kotlyar S, Larkin G, Moore C, et al. S100B Immunoassay: An assessment of diagnostic utility in minor head trauma. Clin Lab Emerg Med 2010;40:28591.
- Leidel BA, Bogner V, Zock M, et al. Serological determination of Protein S-100B: significance in emergency diagnosis of adults with mild cranio-cerebral trauma. A meta-analysis. Der Unfallchirurg 2012;115:90312.
- Muller B, Evangelopoulos DS, Bias K, et al. Can S-100B serum protein help to save cranial CT resources in a peripheral trauma centre? A study and consensus paper. Emerg Med J 2010;28: 93840.
- Egea-Guerrero JJ, Revuerto-Rey J, et al. Accuracy of the S100 beta protein as a marker of brain damage in traumatic brain injury. Brain Inj 2012;26:7682.
- Unden J, Romner B. Can low serum levels of S100B predict normal CT findings after minor head injury in adults?: An evidence based review and meta-analysis. J Head Trauma Rehabil 2010;25:22840.
- Calcagnile O, Unden L, Unden J. Clinical validation of S100B use in management of mild head injury. BMC Emerg Med 2012;12:13.