Evidence Table 1a. Diagnostic reliability of McDonald criteria

StudyStudy DesignPatientsClinical PresentationAdditional Data Used for DiagnosisResultsComments/Quality Scoring
Barkhof, Filippi, Miller, et al., 1997 Prospective cohort studyTotal no. at start: 91Clinically isolated syndrome suggestive of MS and not attributable to other diseases; among those completing study (n = 74), presenting symptom was optic neuritis in 40 patients, spinal cord syndrome in 22, and brainstem/ cerebellum syndrome in 12Baseline MRIs performed at a median of 4 wk (range, 1–20 wk) after onset of symptomsThis study examined various MRI lesion characteristics and used regression analysis to determine the utility of each characteristic with regard to diagnosis. Because previous criteria have demonstrated significant sensitivity, but low specificity, the authors then developed a model with greater positive predictive value based on the results of regression analysis.This study is a thorough, prospective analysis of MRI characteristics with regard to their diagnostic utility, using prospective regression analysis to assess the predictive value of each parameter. On the basis of the findings, a model was developed using the four most predictive parameters. This model became the basis for the MRI criteria used in the McDonald criteria. This study thus does not directly assess the performance of the McDonald criteria, but serves as the basis for the MRI portion of the McDonald criteria. The only significant criticism is that the criteria are based on T2 lesions and gadolinium enhancement without analysis of FLAIR images, sagittal images, or images obtained from higher-strength magnets. These issues were appropriately addressed by the authors.
Duration of follow up: Minimum of 2 yr; median follow up among patients not diagnosed with MS at end of study was 39 mo (range, 23–96 mo)Dropouts: 17 (7 lost to follow up; 10 given definitive diagnosis other than MS and excluded from analysis)Clinically definite MS was diagnosed when clinical signs or symptoms developed in other areas of the central nervous system after a period of at least 1 month, and when other diagnoses had been excluded by appropriate clinical tests1) By regression analysis, the four dichotomized MRI parameters that demonstrated the greatest diagnostic utility were presence of 1 or more gadolinium-enhancing lesions, 1 or more infratentorial lesions, 1 or more juxtacortical lesions, and 3 or more periventricular lesions. The final regression model based on the presence of 3 or more of these 4 parameters demonstrated the following characteristics:QUALITY ASSESSMENT:
Location: 3 sites in Europe (1 each in The Netherlands, Italy, and UK)Completed: 741) MRI - not used in the diagnosis of clinically definite MSSensitivity - 82%Patients evaluated using Poser criteria regardless of results on initial tests?: Yes
Age: NR2) CSF- not used in the diagnosis of clinically definite MS Specificity - 78%Follow up > 80%?: Yes
3) VEP - not used in the diagnosis of clinically definite MS Accuracy - 80%
MRIs were analyzed during a single session by consensus of two observers who were unaware of the clinical findings PPV - 75%
 NPV - 84%
Brex, Miszkiel, O'Riordan, et al., 2001 Prospective cohort studyTotal no. at start: 81Clinically isolated syndrome (defined as the occurrence of a presumed inflammatory demyelinating event of acute onset in the CNS in a patient with no history suggestive of an earlier demyelinating episode); presenting symptom was optic neuritis in 45 patients, brain stem syndrome in 16, spinal cord syndrome in 6, and optic tract lesion in 1; age 16–50 at presentation; appropriate investigations ruled out alternative diagnosesBaseline MRIs performed at a median of 5 wk (range, 1–12 wk) after onset of symptoms1) Contrast enhancing lesion at baseline was the most predictive initial MRI characteristic with positive predictive value of 52%, specificity of 80%, and sensitivity of 61%.This study does not directly assess the utility of MRI as specifically used in the McDonald criteria, but it contributes to the idea that MRI scans performed serially augment the clinical criteria of Poser.
Duration of follow up: Median, 12 mo; range, 11–19 moDropouts: 13MRI - performed as part of the initial baseline evaluation and again after 3 mo, with and without contrast enhancement2) A single T2 lesion on baseline scan had highest sensitivity (89%) but poor specificity (36%).QUALITY ASSESSMENT:
Location: London, UKCompleted: 68 attended all 3 study visits and were included in analysisClinical assessment at 1 yr3) The combination of T2 lesions on baseline scan and new T2 lesions on follow-up scan yielded positive predictive value of 55%, sensitivity of 83%, and specificity of 76%.Patients evaluated using Poser criteria regardless of results on initial tests?: Yes
Age at presentation: Mean, 31; range, 17–504) The combination of enhancing lesions on T1 images of both examinations had the highest positive predictive value (70%) and specificity (94%), but had a very low sensitivity (39%).Follow up > 80%?: Yes -- 84%
CHAMPS Study Group, 2002 Prospective cohort studyTotal no. at start: 190First occurrence of an isolated, well-defined neurological event consistent with demyelination and involving the optic nerve (unilateral optic neuritis; n = 97), spinal cord (incomplete transverse myelitis; n = 42), or brain stem or cerebellum (n = 51); ≥ 2 clinically silent T2-hyperintense brain MRI lesions (≥ 3 mm in size, at least one characteristic of MS [periventricular or ovoid]); onset of symptoms 14 days or less before start of IV corticosteroid and 27 days or less before randomization (see under “Patients”); age 18–50Baseline MRI performed ≥ 4 days after patient completed initial IV corticosteroid therapy (commenced within 14 days of symptom onset and lasted 3 days), but while patient still receiving oral prednisone (lasted 15 days after IV therapy stopped); median time from onset of symptoms = 18 days, range = 8–39 days1) Overall, 27% of patients (51/190) developed clinically definite MS by 18 mo.This study does examine the impact of MRI data in the diagnosis of clinically definite MS - including various MRI criteria. It serves as background information regarding the utility of the addition of MRI criteria in the McDonald criteria.
Duration of follow up: 18 moDropouts: NRMRI - performed ≥ 4 days after initial corticosteroid therapy (see above) and at 6, 12, and 18 months for those patients not meeting the primary study endpoint of clinically definite MS due to recurrence2) The best predictive model for clinically definite MS by 18 mo consisted only of whether patients had ≥ 2 enhancing lesions. None of the other MRI characteristics at their optimized cut-points improved the model fit.QUALITY ASSESSMENT:
Location: 50 sites in the US and CanadaCompleted: NR3) A higher percentage of those patients meeting the Barkhof criteria (≥ 9 T2 lesions) developed clinically definite MS (31%) by 18 mo than did patients who did not meet the criteria (16%) (RR = 1.94, 95% CI = 1.02 to 3.72).Patients evaluated using Poser criteria regardless of results on initial tests?: Yes
Age (mean ± SD): 33 ± 74) The highest risk of clinically definite MS was seen among those with ≥ 2 enhancing lesions, with 52% of these patients reaching clinically definite MS by 18 mo compared with 24% of those with < 2 enhancing lesions (RR = 2.16, 95% CI = 1.35 to 3.46).Follow up > 80%?: Uncertain (dropouts not clearly reported
Patients were enrolled in an RCT comparing interferon beta-1a (30 μg weekly by IM injection; n = 193) vs. placebo (n = 190); all were treated with a course of corticosteroids before the start of the trial. Only placebo patients are considered in this publication.
Comi, Filippi, Barkhof, et al., 2001 Prospective cohort studyTotal no. at start: 309Clinical syndrome indicating unifocal or multifocal involvement of the CNS; first neurological episode suggesting MS in the previous 3 mo; 1 or more abnormalities on neurological exam; positive brain MRI (presence of ≥ 4 white-matter lesions on T2-weighted scans or presence of ≥ 3 white-matter lesions if at least one of these was intratentorial or contrast enhancing); age 18–40Baseline MRI performed as part of pre-study screening, within 3 mo of first neurological episode suggesting MS1) 34% of patients treated with interferon β-1a (52/154) and 45% of patients treated with placebo (69/154) converted to clinically definite MS during the 2-yr study.This was a placebo-controlled treatment trial in patients with clinically isolated syndromes. The study does include a small amount of data regarding the predictive value of initial evaluations in the diagnosis of MS. Although MRI was used prospectively, the report does not contain data regarding the diagnostic performance of serial MRI studies. This study therefore does not answer question 1a directly but provides some background information regarding the utility of MRI in the diagnosis of MS.
Duration of follow up: 2 yrDropouts: 311) MRI - performed as part of the initial baseline evaluation and again at 12 and 24 mo2) The only baseline clinical and MRI variables that were significantly predictive of outcome were multifocal onset (odds ratio 1.99 [95% CI, 1.14 to 3.46]; p = 0.015) and T2 lesion number > 8 (3.64 [1.30 to 10.2]; p = 0.014).QUALITY ASSESSMENT:
Location: 57 sites in 14 European countriesCompleted: 2782) CSF - performed only in those patients with initial manifestations suggestive of spinal cord lesionPatients evaluated using Poser criteria regardless of results on initial tests?: Yes
Age: Mean, 28.5Follow up > 80%?: Yes - 90%
Patients were enrolled in an RCT comparing interferon β-1a (22 μg weekly by SC injection; n = 154) vs. placebo (n = 155); patients were offered open-label interferon after conversion to clinically definite MS
Dalton, Brex, Jenkins, et al., 2002 Prospective cohort studyTotal no. at start: 55Clinically isolated syndrome suggestive of MS, defined as a single event of acute onset in the CNS suggestive of demyelination. In study population, 38 had unilateral optic neuritis, 11 brain stem syndrome, 5 spinal cord syndrome, and 1 a hemianopia due to an MRI lesion in the optic tract.Baseline MRIs conducted within 3 mo of onset of symptoms14/55 patients (25%) developed clinically definite MS and 4 (7%) probable MS according to Poser criteria during the 1-yr follow up. 27 of 55 patients met McDonald criteria for MS at 1 yr.This study provides minimal data on the relative sensitivity of the Poser and McDonald criteria. This was not the primary purpose of the study, but it does demonstrate increased sensitivity of the McDonald criteria.
Duration of follow up: Median, 12 mo (range, 11–16 mo)Dropouts: 0Exclusion criteria: History of neurological symptoms suggestive of demyelination; age < 17 or > 50MRI - performed at baseline, 3 mo later, and approximately 1 yr after presentationMRI data focused on ventricular volume changes.
Location: London, UKCompleted: 55QUALITY ASSESSMENT:
Age: Mean, 29.6; range, 21–41Patients evaluated using Poser criteria regardless of results on initial tests?: Yes
Follow up > 80%?: Yes - 100%
Dalton, Brex, Miszkiel, et al., 2002 Prospective cohort studyTotal no. at start: 119Clinically isolated syndrome, defined as an acute isolated event affecting one region of the CNS and presumed to be demyelinating, with no previous history of possible demyelinating events. In study population, 87 had acute unilateral optic neuritis, 2 bilateral consecutive optic neuritis, 19 brain stem syndrome, 10 spinal cord syndrome, and 1 demyelinating optic tract lesion. Maximal symptoms and signs evident within 14 days of symptom onset. Alternative diagnoses excluded. Age 16–50.Baseline MRIs conducted within 3 mo of onset of symptoms1) Clinically definite MS was present in 7% of patients (7/95) at 3 mo, 20% (16/79) at 1 yr, and 38% (19/50) at 3 yr follow up.This study specifically evaluates the performance of the McDonald criteria in comparison with the Poser criteria. This is a preliminary report of a 3-yr study in which fewer than 80% of patients had completed the 1-yr evaluation. The study demonstrates a significant increase in sensitivity of the McDonald criteria.
Duration of follow up: 3 mo to 3 yr (follow up ongoing - see under “Patients,” at right)Follow up ongoing at time of publication: 95 patients studied at 3 mo, 79 at 1 yr, and 50 at 3 yrMRI of the brain was performed at baseline, 3 mo, 1 yr, and 3 yr. MRI of the spinal cord was performed at baseline, 1 yr, and 3 yr.2) Performance of the McDonald criteria at 3-mo evaluation for predicting the development of clinically definite MS at 1 yr:QUALITY ASSESSMENT:
Location: London, UKDropouts: 1 (died of asthmatic attack) Sensitivity = 73%Patients evaluated using Poser criteria regardless of results on initial tests?: Yes
Completed: Follow up ongoing; see above Specificity = 87%Follow up > 80%?: No - at the time of this report the study was ongoing with fewer than 80% of patients having had 1-yr evaluations
Age: Median at onset, 31; range, 16–50 PPV = 58%
 NPV = 93%
 Accuracy = 84%
3) Performance of the McDonald criteria at 1-yr evaluation for predicting the development of clinically definite MS at 3 yr:
 Sensitivity = 94%
 Specificity = 83%
 PPV = 77%
 NPV = 96%
 Accuracy = 87%
Filippi, Horsfield, Morrissey, et al., 1994 Prospective cohort studyTotal no. at start: 129Clinically isolated syndrome of the optic nerves (n = 40), brainstem (n = 16), or spinal cord (n = 28) suggestive of MS; syndrome fully developed within 14 days of symptom onset; age 10–50 at presentation; appropriate studies (including initial MRI) ruled out alternative diagnosesBaseline MRIs were conducted within 60 days of onset of symptoms in 69/84 patients (82%), later in remaining 15 patients1) During 5-yr follow up, 34 patients (40%) developed clinically definite MS: 18 of 40 (45%) with initial optic neuritis, 10 of 28 (36%) with initial spinal cord syndrome, and 6 of 16 (38%) with initial brainstem syndrome. 4 patients (5%) developed clinically probable MS - 2 with initial optic neuritis and one each with spinal cord or brainstem syndrome.The MRI criteria used here are similar to those used in the McDonald criteria but not precisely the same. This study supports the use of MRI findings in the diagnosis, but does not directly compare with the MRI criteria adopted in the McDonald criteria.
Duration of follow up: Mean ± SD, 63 ± 11 mo; range, 43–84 moDropouts: 40 of original cohort not included in this 5-yr follow up1) MRI - repeat MRI scans were performed after a mean of 63 mo. Quantitative semi-automated computer assessment of T2 lesion load was performed in a manner shown to have an intrarater reproducibility of 6%.2) 52 patients with abnormal MRI at presentation with median total brain lesion volume 0.83 cm3 (range, 0.09–52.41), with median infratentorial lesion volume of 0 cm3 (range, 0–1.82)Additional reports on this study population are provided in Morrissey, Miller, Kendall, et al., 1993; and O'Riordan, Thompson, Kingsley, et al., 1998, below.
Location: London, UKCompleted: 89 re-examined and re-scanned at 5-yr follow up; 84 had complete data available (initial MRI unavailable at follow up for 5)2) Clinical examination - patients were re-examined after a mean of 63 mo with assessment of EDSS. MS was diagnosed solely on clinical grounds using Poser criteria.3) Patients developing MS had significantly higher total and infratentorial lesion loads at presentation than those who did not: median total lesion volumes were 1.15 cm3 (range, 0–52.41) versus 0 cm3 (range, 0–25.6), p < 0.0001; the median infratentorial lesion volumes were 0 cm3 (range, 0–1.82) versus 0 cm3 (range, 0–0.59), p < 0.0001.QUALITY ASSESSMENT:
Age at baseline presentation: Mean, 31; range, 13–504) Lesion load of 1.23 cm3 at presentation afforded the highest probability of separating patients developing MS from those who did not. Patients then divided into three groups: Group A - patients with total lesion volume ≥ 1.23 cm3, Group B - patients with abnormal MRI but total lesion volume < 1.23 cm3, and Group C - patients with normal MRI at baseline. Results:Patients evaluated using Poser criteria regardless of results on initial tests?: Yes
Group A - 19 of 21 (90%) patients developed MS (18 clinically definite, 1 clinically probable)Follow up > 80%?: Yes - 84%
Group B - 17 of 31 (55%) developed MS (15 definite and 2 probable)
Group C - 2 of 32 (6%) developed MS (1 definite and 2 probable)
5) 18 of 20 (90%) patients with infratentorial lesions developed MS (all clinically definite), whereas 44 of 64 (69%) without such lesions did not.
6) A significant correlation was found between total and infratentorial lesion load on the initial MRI (Spearman rank correlation coefficient = 0.649; p < 0.0001).
Ghezzi, Martinelli, Torri, et al., 1999 Prospective cohort studyTotal no. at start: 112Acute isolated optic neuritisBaseline paraclinical tests performed within 6 mo of onset of optic neuritis; mean interval, 45 ± 24 days36% of patients (37/102) developed clinically definite MS in 2.3 ± 1.6 yr of follow up after initial attack of optic neuritis.This study evaluated the utility of paraclinical tests in predicting those patients with clinically isolated syndromes who would progress to develop clinically definite MS. The data presented provide background information regarding the utility of paraclinical tests, but do not directly evaluate the McDonald criteria in that the paraclinical tests were not applied in the same manner as used in the McDonald criteria.
Duration of follow up: Mean ± SD, 6.3 ± 2.2 yr; median, 5 yrDropouts: 10 lost to follow up1) MRI - performed at baseline only - details not delineatedNumber of patients developing MS in relation to the results of paraclinical tests performed at baseline:QUALITY ASSESSMENT:
Location: Gallarate, ItalyCompleted: 1022) CSF IgG Index was the parameter utilized; definition of abnormal not statedMS+Ms-P-valuePatients evaluated using Poser criteria regardless of results on initial tests?: Yes
Age: Mean ± SD, 29.2 ± 9.03) VEP - Multiple Evoked Potential studies were performed at baseline. No details regarding technique were presented.1) MRI:0.0001Follow up > 80%?: Yes - 91%
Negative3734
Positive031
2) CSF:0.19
Negative2229
Positive1231
3) VEP:0.95
Negative1016
Positive2648
4) BAEP, median nerve SEP, and upper limb MEP:0.7
Negative27
Positive1731
5) BAEP, median and tibial nerve
SEP:0.02
Negative95
Positive621
Morrissey, Miller, Kendall, et al., 1993 Prospective cohort studyTotal no. at start: 132Clinically isolated syndrome of the optic nerves (n = 44), brainstem (n = 17), or spinal cord (n = 28) suggestive of MS; syndrome acute or subacute in onset; age 10–50 at presentation; appropriate studies (including initial MRI) ruled out alternative diagnosesBaseline MRIs were conducted within 60 days of onset of symptoms in 74/89 patients (83%), later in remaining 15 patientsAfter 5 yr, progression to clinically definite MS occurred in 41 of 57 (72%) of patients who had had abnormal initial scans, but in only 2 of 36 (6%) patients whose initial scan was normal (P < 0.0001).This study provides background information regarding the utility of MRI in the diagnosis of MS but does not utilize MRI in the same manner as the McDonald criteria and therefore does not answer Question 1a specifically.
Duration of follow up: Mean, 63.6 mo; range, 43–84 moDropouts: 43 of original cohort not included in 5-yr follow up1) MRI - performed at baseline and a mean of 1.3 yr later and again at 5.3 yr - all scans were unenhancedAdditional reports on this study population are provided in Filippi, Horsfield, Morrissey, et al., 1994, above; and O'Riordan, Thompson, Kingsley, et al., 1998, below.
Location: London, UKCompleted: 89 re-examined and re-scanned at 5-yr follow up2) CSF - not performed in patients with clinically isolated optic neuritis, but was performed in patients with isolated spinal cord or brainstem syndromesQUALITY ASSESSMENT:
Age at baseline presentation: Mean, 31.3; range, 13–50Patients evaluated using Poser criteria regardless of results on initial tests?: Yes
Follow up > 80%?: No - 67%
Optic Neuritis Study Group, 1997Prospective cohort studyTotal no. at start: 388Acute unilateral optic neuritis with visual symptoms of 8 days or less; no previous history of optic neuritis or ophthalmoscopic signs of optic atrophy in the affected eye; no evidence of a systemic disease other than MS that might be associated with the optic neuritis; no previous treatment with corticosteroids for MS or for optic neuritis in the opposite eye; age 18–46 yrBaseline MRIs performed “on study entry” (within 8 days of onset of acute symptoms)1) 27% of patients (106/388) developed clinically definite MS with 5 yr, and an additional 9% (35 patients) developed probable MS.This study provides background information regarding the utility of MRI in the diagnosis of MS, but the utilization of MRI did not include serial studies as is the case for the McDonald criteria, and therefore this report does not provide direct data on the performance of the McDonald criteria.
Duration of follow up: 5 yrDropouts: 47Patients with a diagnosis of clinically definite or probable MS were excludedMRI - brain MRI was performed at baseline according to standardized protocols2) The presence of MRI abnormalities at the time of optic neuritis was the single most important predictor of the development of clinically definite MS by 5 yr. The probability of clinically definite MS was 16% in the 202 patients with no MRI abnormalities, 37% in the 60 patients with 1–2 MRI abnormalities, and 51% in the 89 patients with ≥ 3 MRI abnormalities.QUALITY ASSESSMENT:
Location: 15 sites in the USCompleted: 341 followed up for 5 yrPatients evaluated using Poser criteria regardless of results on initial tests?: Yes
Age (mean ± SD): 32 ± 7Follow up > 80%?: Yes - 88%
Patients were participants in an RCT comparing IV methylprednisolone vs. oral prednisone vs. oral placebo
O'Riordan, Thompson, Kingsley, et al., 1998 Prospective cohort studyTotal no. at start: 129Clinically isolated syndrome (defined as an acute or subacute episode suggestive of demyelination affecting the optic nerves [n = 42], brainstem [n = 16], or spinal cord [n = 23]); age 10–50 at presentation; appropriate studies (including initial MRI) ruled out alternative diagnosesNot clear when baseline MRIs conducted1) Patients with a normal baseline MRI (n = 27): Only 3 patients (11%) progressed to clinically definite MS, all of whom had benign disease. 2 additional patients (7%) had clinically probable MS. Of these 5 patients, 4 had 10-yr follow-up MRIs and all had developed new lesions. 22 patients of these original 27 (81%) were still classified as having clinically isolated syndromes.The MRI criteria used here are similar to those used in the McDonald criteria but not precisely the same. This study supports the use of MRI findings in the diagnosis, but does not directly compare with the MRI criteria adopted in the McDonald criteria.
Duration of follow up: Mean, 9.7 yrDropouts: 48 of original cohort not included in this 10-yr follow up1) MRI - baseline and follow-up scans up to the 5-yr scans were performed on a 0.5 T scanner using SE2000/60 sequences. 10-yr scans were performed on a 1.5 T scanner and used conventional dual spin echo technique. All scans were evaluated only for the presence of hyperintense lesions. Scans were considered abnormal only if one or more asymptomatic lesions characteristic for demyelination were present. The number of lesions compatible with demyelination was recorded. All scans were read with the baseline and 5-yr scans side-by-side for comparison.2) Patients with abnormal MRI at baseline (n = 54): After 10 yr, only 7 patients (13%) still had a diagnosis of clinically isolated syndrome, 2 patients (4%) had clinically probable MS, and 45 patients (83%) had progressed to clinically definite MS. Of those with clinically definite MS, 21 patients (39%) had benign disease, 11 patients (20%) relapsing/remitting disease with an EDSS of > 3, and 13 patients (24%) developed secondary progressive MS.Additional reports on this study population are provided in Filippi, Horsfield, Morrissey, et al., 1994; and Morrissey, Miller, Kendall, et al., 1993, above.
Location: London, UKCompleted: 81 re-examined and re-scanned at 10-yr follow up2) Diagnosis of MS was made solely on the basis of Poser criteria after 10 yr of follow upFor those with an abnormal baseline MRI, the presence of infratentorial lesions did not confer any greater risk for the subsequent development of clinically definite MS.QUALITY ASSESSMENT:
Age at baseline presentation: Mean, 32.3; range, 17–49Patients evaluated using Poser criteria regardless of results on initial tests?: Yes
Follow up > 80%?: No - 81 patients at 10-yr follow up of 129 patients in original cohort = 63%
Sastre-Garriga, Tintoré, Rovira, et al., 2003 Prospective cohort studyTotal no. at start: 59Episode of clinical brainstem dysfunction suggestive of inflammatory demyelination; clinical assessment within 3 mo of onset of symptoms; age < 50; other possible diagnoses excludedMean time between onset of symptoms and initial MRI 29 days1) Paty MRI criteria:Clinical diagnosis of MS was made based on the occurrence of neurological symptoms lasting over 24 hr without the requirement of objective findings on neurological examination. This definition is more sensitive but less specific than most clinical criteria in use, including the Poser criteria. Additionally, this study evaluated the ability of baseline parameters to predict the subsequent development of MS. These parameters were not performed serially to assess their correlation with clinical diagnosis.
Duration of follow up: Mean, 37 moDropouts: 8 (excluded because follow-up shorter than 12 mo)1) MRI - 1.0 or 1.5 T scanners including transverse proton density and T2-weighted conventional spin echo or fast spin echo, and T1-weighted spin echo. T1 images were repeated after administration of gadolinium. Sagittal T2 or transverse T2 FLAIR were also performed on most patients. A blinded neuroradiologist recorded the number and sites of abnormalities. The MRI diagnostic criteria of Paty, Fazekas, and Barkhof were then studied. Sensitivity = 89%QUALITY ASSESSMENT:
Location: Barcelona, SpainCompleted: 512) CSF - presence of oligoclonal bands were assessed, but not used in the diagnosis of MS Specificity = 52%Patients evaluated using Poser criteria regardless of results on initial tests?: No - symptomatic recurrence did not require objective examination abnormalities
Age: Mean at assessment, 29; range, 14–493) VEP - values of VEP and SEP results were assessed but not used in the diagnosis of MS PPV = 50%Follow up > 80%?: Yes - 86%
 NPV = 89%
 Accuracy = 65%
2) Fazekas MRI criteria:
 Sensitivity = 89%
 Specificity = 48%
 PPV = 48%
 NPV = 89%
 Accuracy = 63%
3) Barkhof MRI criteria:
 Sensitivity = 78%
 Specificity = 61%
 PPV = 52%
 NPV = 83%
 Accuracy = 67%
4) CSF - presence of oligoclonal bands:
 Sensitivity = 100%
 Specificity = 42%
 PPV = 44%
 NPV = 100%
 Accuracy = 63%
5) Evoked potential studies - no statistically significant differences for baseline VEP or SSEP parameters were found between patients who did and those who did not convert to MS
Tintoré, Rovira, Río, et al., 2003 Cohort study; data collected prospectively, McDonald criteria applied retrospectivelyTotal no. at start: 139Clinically isolated syndrome suggestive of CNS demyelination involving the optic nerve (41.5%), brainstem (24.5%), spinal cord (28%), or combinations of the above (6%), and not attributable to other diseases; age < 50 yrBaseline MRIs completed within 3 mo of onset of symptoms1) At 1 yr, 15 patients (11%) had a second clinical attack and therefore fulfilled the requirement for dissemination in time and space necessary for clinically definite MS according to the Poser criteria. Of these 15 patients, 10 also fulfilled the radiologic conditions of dissemination in time and space.This article precisely and specifically evaluates Question 1a.
Duration of follow up: Mean, 39 ± 17.2 mo; range, 12–77 mo; all patients were followed up for at least 1 yr (inclusion criterion), 122 for at least 2 yr, and 86 for at least 3 yrDropouts: 17 by 2 yr; 53 by 3 yrAnalysis included only patients with clinical and MRI examinations within 3 mo of onset of symptoms and clinical follow up of at least 12 mo1) MRI - standard MRI techniques used after the first demyelinating event and 12 mo later2) Fifty-one patients (37%) fulfilled MRI requirements for dissemination in time and space and therefore were considered to have MS according to the McDonald criteria. Ten of these 51 patients (20%) had a second clinical event during the first year of follow up. In total, 56 of 139 patients (40%) fulfilled the McDonald criteria for MS either by MRI or clinically.QUALITY ASSESSMENT:
Location: Barcelona, SpainCompleted: 139 were followed up for at least 1 yr (inclusion criterion), 122 for at least 2 yr, and 86 for at least 3 yr2) CSF - the presence of oligoclonal bands was assessed after the first demyelinating event3) The McDonald criteria showed a sensitivity of 74%, specificity of 85%, PPV of 80%, NPV of 80%, and accuracy of 80% in predicting conversion to clinically definite MS:Patients evaluated using Poser criteria regardless of results on initial tests?: Yes
Age: Mean, 30; range, 13–49Clinically definite MSFollow up > 80%?: Yes - 100% (first yr)
at 3 yr+-
McDonald+287
criteria
at 1 yr-1041
4) In the first year the Poser criteria allowed the diagnosis of clinically definite MS in 11% compared with 37% with the McDonald criteria.

From: Appendix F. Evidence Tables

Cover of Criteria to Determine Disability Related to Multiple Sclerosis
Criteria to Determine Disability Related to Multiple Sclerosis.
Evidence Reports/Technology Assessments, No. 100.
McCrory DC, Pompeii LA, Skeen MB, et al.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.