Table 3

Regressions with the proportion of Caesarean sections per hospital as the dependent variable.

Variable

Main analysis

Imputation for

missing values

Test for whether

the effect of

hospital revenue

per bed varies

according to

number of births

I

II

III

Intercept

-1.262

**

-1.207

**

-3.210

**

(2.38)

(2.45)

(3.12)

rev_hosp (lagged)

0.137

**

0.153

**

0.462

**

(2.38)

(2.63)

(2.88)

age_le20¹

-0.012

0.014

-0.016

(0.36)

(0.42)

(0.48)

age_hi35¹

0.111

**

0.123

**

0.110

**

(2.78)

(3.12)

(2.77)

edu_univ²

-0.064

0.015

-0.094

(0.78)

(0.20)

(1.13)

edu_uss²

-0.372

**

-0.361

**

-0.424

***

(2.97)

(2.88)

(3.34)

weight_le2500³

-0.013

-0.009

-0.002

(0.53)

(0.39)

(0.11)

weight_hi4500³

0.052

0.036

0.053

(1.59)

(1.12)

(1.61)

ab_present

0.316

****

0.316

****

0.300

****

(10.80)

(10.81)

(9.95)

preeclam

0.126

****

0.122

****

0.130

****

(6.57)

(6.34)

(6.77)

mult_birth

1.709

*

1.715

*

1.283

(1.89)

(1.90)

(1.39)

no_births

-0.122

**

-0.106

**

0.153

(3.06)

(2.65)

(1.15)

weekend

-0.345

***

-0.261

**

-0.333

**

(3.34)

(2.92)

(3.23)

rev_hosp (lagged) × no_births

-0.046

**

(2.17)

Number of observations

1260

1334

1260

Number of hospitals

46

46

46

R²

0.74

0.74

0.74

All variables ln transformed. Regression coefficients with t-values in brackets.

¹ Reference category: the proportion of mothers aged 20-35 years

² Reference category: the proportion of mothers with compulsory school education

³ Reference category: the proportion of babies with a birthweight 2500-4500 g

**** p < 0.0001

*** p < 0.001

** p < 0.05

* p < 0.10

Grytten et al. BMC Health Services Research 2011 11:267   doi:10.1186/1472-6963-11-267