Table 3.

Methods used to deal with missing primary outcome data in included published randomised controlled trials

Method used to deal with missing data	Frequency of use in RCTs in:
	Alzheimer’s disease (n = 47)		Parkinson’s disease (n = 31)
	n	%	n	%
Complete-case analysis	14	30	11	35
Survival analysis	3	6	10	32
Last observation carried forward	16	34	6	19
Mixed model	18	38	4	13
Data imputation	7	15	4	13

Twelve AD trials did not report any information regarding how missing data was dealt with, four used two methods and two used three methods and one used four methods. Five Parkinson’s disease trials did not report any information regarding how missing data was dealt with and four used two methods