Skip to main content

Table 7 Comparison of prevalence ratios and respective confidence interval estimates (obtained by unadjusted Poisson/Cox regression, Poisson regression with scale factor adjusted by χ2 and deviance, Poisson/Cox regression with robust variances, log-binomial regression and logistic regression) and odds ratio with confidence interval estimate (obtained by logistic regression) with the Mantel-Haenszel prevalence ratio in the analysis of the original and modified data using underweight (weight for age Z-score < -2) as the outcome, previous hospitalization as the risk factor and low birth weight as confounder (situation1).

From: Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio

Original data

Point estimate

95% Confidence interval

 

value

% diff.

lower

upper

width

% diff.

PR Mantel-Haenszel

2.48

--

1.46

4.23

2.78

--

PR Poisson/Cox (unadj)

2.48

-0.2%

1.43

4.31

2.88

3.6%

PR Poisson (χ2)

2.48

-0.2%

1.44

4.26

2.82

1.5%

PR Poisson (deviance)

2.48

-0.2%

1.89

3.25

1.36

-51.1%

PR Poisson/Cox (robust)

2.48

-0.2%

1.46

4.22

2.76

-0.6%

PR log-binomial

2.48

-0.1%

1.46

4.22

2.77

-0.4%

OR logistic regression

2.64

6.3%

1.49

4.68

3.18

14.6%

Modified data

Point estimate

95% Confidence interval

 

value

% diff.

lower

upper

width

% diff.

PR Mantel-Haenszel

2.61

--

1.61

4.23

2.61

--

PR Poisson/Cox (unadj)

2.60

-0.4%

1.57

4.30

2.73

4.6%

PR Poisson (χ2)

2.60

-0.4%

1.59

4.26

2.67

2.4%

PR Poisson (deviance)

2.60

-0.4%

2.01

3.36

1.35

-48.3%

PR Poisson/Cox (robust)

2.60

-0.4%

1.61

4.19

2.58

-1.2%

PR log-binomial

2.61

-0.1%

1.61

4.21

2.60

-0.7%

OR logistic regression

2.85

9.3%

1.68

4.84

3.15

20.7%