Table 2.

GRADE assessment of the included trials, observational studies and case series

	No. of patients	Population	Dates	Pathogen	Design	Outcome	Death OR (95% CI)	Quality of evidence
Rayamajhi et al. [11]	22 (11 IVIG, 11 placebo)	Pediatric	05–09/2009	JE	Prospective trial	Death/survival; neurological outcome; adverse effects; duration of hospital stay; Glasgow Coma Score	At discharge: 3⋅29 (0⋅12–89⋅82) at 3–6 months: 0⋅45 (0⋅01–8⋅4)	Very lowa, b, j
Zhang et al. [13]	80	Pediatric	03—09/2014	EV	Prospective case series	Clinical outcome; adverse effect	nck	Very lowc
Odessky et al. [15]	41 (27 IVIG, 14 controls)	Pediatric	1952	Measles	Observational study	Death/survival; neurological sequelae; adverse effects	0⋅75 (0.11–5⋅11)	Very lowb, c, g
Neu [14]	72	Adults	1970-1979	Mixed	Observational study	Death/survival; neurological sequelae	0⋅032 (0⋅0033–0⋅3024)	Very lowb, d, e, f, g
Wang et al. [80]	97 (Treatment/outcome data on 34 patients: 14 IVIG, 20 no IVIG)	Pediatric	04–12/1998	EV	Observational study	Survival/death	1⋅2 (0⋅26–5⋅59)	Very lowb, c, g, h, i, j

Legend see Table 1. nc not calculable

^aUnclear risk of detection bias

^bSmall sample size

^cNo control group and/or no randomized design

^dHeterogeneous cohort

^eHigh risk of selection or allocation bias

^fHigh risk of detection bias

^gRetrospective design

^hLack of information on part of cohort

ⁱLack of information in treatment protocol; possibly heterogenous treatment

^jLoss to follow-up

^kStudy designed as a randomized controlled trial for the use of methylprednisolone in enteroviral encephalitis; no control intervention for IVIG—> study classified as case series