Prediction Rule for Scrub Typhus Meningoencephalitis in Children: Emerging Disease in North India.J Child Neurol. 2020 10; 35(12):820-827.JC
OBJECTIVES
To evaluate the proportion of scrub typhus meningoencephalitis among children with acute encephalitis syndrome and to outline its differentiating features. To develop a prediction rule for scrub typhus meningoencephalitis.
METHODS
A prospective cohort study was conducted at a tertiary care public hospital in Northern India. Consecutive patients of acute encephalitis syndrome who met our inclusion criteria were enrolled over 2 years. Standardized workup including serum IgM against Orientia tsutsugamushi was performed. Clinical and laboratory features were compared between IgM-positive and IgM-negative patients. The area under the receiver operating characteristic curve of the score derived from "independent predictors" was measured. Sensitivity, specificity, positive and negative predictive values, and positive and negative likelihood ratios were calculated at different cut-offs of the score.
RESULTS
Scrub typhus IgM enzyme-linked immunosorbent assay was positive in 66/352 patients (18.8%). Longer duration of fever and prodromal stage along with eschar, hepatomegaly, lymphadenopathy, and pneumonia were significantly more prevalent in scrub typhus meningoencephalitis. However, petechiae were frequent in non-scrub typhus patients. Leucocytosis, lymphocytosis, thrombocytopenia, hypoalbuminemia, and elevated levels of serum bilirubin, serum transaminases, and cerebrospinal fluid protein were associated with scrub typhus meningoencephalitis. Logistic regression revealed fever for >8 days, pneumonia, absence of petechiae, cerebrospinal fluid protein >1000 mg/L, and serum glutamic oxaloacetic transaminase >100 IU/L as independent "predictors" of scrub typhus meningoencephalitis. The area under the receiver operating characteristic curve (95% confidence interval) of the prediction score was 0.832 (0.78-0.89). Score at cutoff ≥1 had 91% sensitivity, 96.1% negative predictive value, and at cutoff ≥4 had 99.7% specificity, 88.9% positive predictive value, 83.1% negative predictive value, 40.3 positive likelihood ratio, 0.88 negative likelihood ratio for identifying scrub typhus meningoencephalitis.
CONCLUSION
Prediction score may help physicians in peripheral areas to identify and treat scrub typhus meningoencephalitis, an emerging cause of acute encephalitis syndrome in India.
