Skip to main content

Table 2 Method of analysis

From: Longitudinal randomised controlled trials in rehabilitation post-stroke: a systematic review on the quality of reporting and use of baseline outcome values

Measurement of outcome

Baseline- 1. After intervention- 2. Follow-up

46 % (39/84)

 

Othera

54 % (45/84)

aMore follow-ups after intervention or assessment during intervention

Repeated measure data was analysed

Cross sectional at each time-point

38 % (32/84)

 

Longitudinal

39 % (33/84)

 

Both longitudinal and cross sectional

8 % (7/84)

 

Repeated data not fully analysed

14 % (12/84)

Method of analysis

  

Cross sectional

t-test

22 % (10/46)

 

ANOVA/ANCOVA

35 % (16/46)

 

Regression

7 % (3/46)

 

Non parametric test/dichotomised data

37 % (17/46)

Correction for multiple testing due to repeated measures

76 % (31/41)

Longitudinal

Mixed Model

21 % (7/33)

 

Repeated measure ANOVA

72 % (24/33)

 

Generalised Estimating Equations

6 % (2/33)

Results presented

  
 

Mean (SD)b at each time-point per group

66 % (47/71)

 

F values (for ANOVA/ANCOVA)

54 % (19/35)

 

Regression coefficients

71 % (6/12)

 

None

8 % (6/71a)

  1. a13 of the 84 studies were protocols. bStandard deviation