Table 2.

Summary of sensitivity analyses on overall incidence correction factor

	N	K	95% CI
Base case	352	3.5	(3.1 to 3.9)
Sensitivity analyses
Exclude anthropometric outliers by WHO method	352	3.7	(3.3 to 4.1)
Exclude anthropometric outliers by SMART method	352	4.7	(4.2 to 5.3)
Exclude coverage surveys without exact match by time and place	63	4.5	(3.7 to 5.5)
Exclude non-representative prevalence surveys	208	4.7	(4.1 to 5.3)
Exclude admissions data containing possible readmissions	114	4.3	(3.7 to 5.0)
Exclude implausible values of the incidence correction (K<1 by expert consensus or K>15.2 by Tukey’s criteria)	284	3.6	(3.4 to 3.9)
Exclude low quality prevalence surveys by SMART criteria	258	3.8	(3.4 to 4.4)

SMART, Standardised Monitoring and Assessment of Relief and Transitions.