Skip to main content

Table 3 Logistic regression for being a Cochrane Review (dependent/response variable) by various predictor variables

From: Applicable or non-applicable: investigations of clinical heterogeneity in systematic reviews

 

OR

95 % CI

p value

Number of studies

0.94

0.92–0.97

<0.0001

Number of participants (log)

0.14

0.077–0.24

<0.0001

Quantitative synthesis (1 = yes, 0 = no)

0.38

0.20–0.73

0.0035

Individual patient data (IPD; 1 = yes, 0 = no)*

n/a

 

n/a

Aggregate patient data (APD; 1 = yes, 0 = no)

0.28

0.15–0.53

0.0001

Combination of IPD & APD (1 = yes, 0 = no)

1.00

0.24–4.11

1.00

Investigated an aspect of clinical heterogeneity (1 = yes, 0 = no)

2.14

0.94–4.87

0.070

  1. Note: 1. Bolded p value indicates significance. 2. OR of greater than 1 favors being a Cochrane Review for each variable
  2. *No Cochrane studies conducted individual patient data analysis