High sensitivity and specificity in proposed clinical diagnostic criteria for anti-N-methyl-D-aspartate receptor encephalitis

To determine the validity of the proposed clinical diagnostic criteria for anti-N-methyl-d-aspartate receptor (NMDAR) encephalitis in paediatric patients.

The diagnostic criteria for anti-NMDAR encephalitis proposed by Graus et al. (2016) use clinical features and conventional investigations to facilitate early immunotherapy before antibody status is available. The criteria are satisfied if patients develop four out of six symptom groups within 3 months, together with at least one abnormal investigation (electroencephalography/cerebrospinal fluid) and reasonable exclusion of other disorders. We evaluated the validity of the criteria using a retrospective cohort of paediatric patients with encephalitis. Twenty-nine patients with anti-NMDAR encephalitis and 74 comparison children with encephalitis were included.

As expected, the percentage of patients with anti-NMDAR encephalitis who fulfilled the clinical criteria increased over time. During the hospital inpatient admission, most patients (26/29, 90%) with anti-NMDAR encephalitis fulfilled the criteria, significantly more than the comparison group (3/74, 4%) (p<0.001). The median time of fulfilling the criteria in patients with anti-NMDAR encephalitis was 2 weeks from first symptom onset (range 1-6). The sensitivity of the criteria was 90% (95% confidence interval 73-98) and the specificity was 96% (95% confidence interval 89-99).

The proposed diagnostic criteria for anti-NMDAR encephalitis have good sensitivity and specificity. Incomplete criteria do not exclude the diagnosis.

The proposed clinical diagnostic criteria for anti-N-methyl-d-aspartate receptor (NMDAR) encephalitis by Graus et al. (2016) have high sensitivity and specificity in paediatric patients. The median time of fulfilling the criteria in patients with anti-NMDAR was 2 weeks from first symptom onset.