Skip to main content

Table 5 Marginal effects from logistic regression models: factors associated with choosing a private sector provider vs. a public sector provider stratified by type of care

From: Health services utilization and out-of-pocket (OOP) expenditures in public and private facilities in Pakistan: an empirical analysis of the 2013–14 OOP health expenditure survey

Pr(Private) Outpatient Inpatient
  Marginal effect SE p-value Marginal effect SE p-value
Gender
 Male Ref.    Ref.   
 Female −0.025 0.01 0.02 0.034 0.04 0.41
Age
 0 to 5 Ref.    Ref.   
 6 to 20 −0.003 0.02 0.88 −0.042 0.07 0.52
 21 to 40 −0.005 0.02 0.78 −0.018 0.06 0.78
 41 to 60 −0.016 0.02 0.37 −0.096 0.07 0.14
 > 60 − 0.011 0.02 0.62 −0.024 0.07 0.73
Household wealth quintile
 Poorest Ref.    Ref.   
 Poorer 0.001 0.02 0.97 −0.010 0.07 0.87
 Middle 0.002 0.02 0.93 0.058 0.06 0.34
 Richer 0.058 0.02 0.001 0.004 0.06 0.95
 Richest 0.075 0.02 < 0.001 0.082 0.06 0.19
Region
 Rural 0.001 0.01 0.92 0.112 0.04 0.001
Province
 Punjab Ref.    Ref.   
 Sindh 0.062 0.01 < 0.001 0.053 0.04 0.14
 KP −0.133 0.02 < 0.001 −0.185 0.04 < 0.001
 Balochistan − 0.187 0.03 < 0.001 −0.065 0.06 0.30
Household size
 1–4 Ref.    Ref.   
 5–8 −0.050 0.01 < 0.001 0.028 0.04 0.53
 9–12 −0.051 0.02 0.001 −0.033 0.05 0.54
 13+ −0.060 0.03 0.03 −0.026 0.07 0.72
Type of illness
 Communicable Ref.    Ref.   
 Accident/Injury −0.107 0.05 0.04 0.005 0.08 0.95
 Chronic −0.014 0.02 0.40 −0.095 0.06 0.13
 Childbirth N/A    0.056 0.06 0.38
 Other female reproductive health concerns −0.053 0.03 0.08 0.159 0.08 0.05
 Other illness 0.033 0.01 0.02 −0.008 0.06 0.89
N 6770 1199
Pseudo R-squared 0.06 0.06
\