Skip to main content

Table 5 Incremental linear robust models highlighting potential omitted variable bias

From: Is the timing of radiological intervention and treatment day associated with economic outcomes in DRG-financed health care systems: a case study

 

(1) Basic model (Omitted variable bias)

(2) Basic model

(3) Extended model

(4) Model incl. patient confounding

(5) Model incl. all confounders

Constant

0.095∗∗∗

0.084∗∗∗

0.563∗∗∗

0.647∗∗∗

0.675∗∗∗

 

(0.008)

(0.008)

(0.011)

(0.024)

(0.049)

Basic model weekends

     

 Admission WE

−0.053∗

−0.063∗

−0.066∗∗

−0.064∗∗

−0.041·

 

(0.025)

(0.025)

(0.021)

(0.021)

(0.021)

 Scanning WE

−0.147∗∗∗

0.011

0.061∗∗

0.061∗∗

0.051∗

 

(0.032)

(0.028)

(0.023)

(0.023)

(0.023)

 Discharge WE

0.160∗∗∗

0.167∗∗∗

0.035∗

0.033·

0.015

 

(0.020)

(0.021)

(0.017)

(0.017)

(0.017)

 Treatment WE

−0.104∗∗

0.149∗∗∗

0.086∗∗∗

0.087∗∗∗

0.079∗∗∗

 

(0.038)

(0.028)

(0.023)

(0.023)

(0.023)

 Interaction WE scan and treatment

0.536∗∗∗

    
 

(0.054)

    

 Number of exams

  

-0.005

−0.005·

0.001

   

(0.003)

(0.003)

(0.003)

 Diff. scan and admission (in days)

  

−0.048∗∗∗

−0.048∗∗∗

−0.049∗∗∗

   

(0.002)

(0.002)

(0.002)

 Diff. treatment and scan (in days)

  

0.004∗

0.004∗

0.003·

   

(0.002)

(0.002)

(0.002)

 Diff. dismissal and scan (in days)

  

−0.041∗∗∗

−0.041∗∗∗

−0.039∗∗∗

   

(0.001)

(0.001)

(0.001)

Patient data

     

 Female

   

−0.075∗∗∗

−0.080∗∗∗

    

(0.012)

(0.011)

 Age

   

−0.001∗

−0.001∗

    

(0.000)

(0.000)

 Private insurance

   

−0.012

−0.007

    

(0.013)

(0.013)

Clinical data

     

 Deceased

    

0.464∗∗∗

     

(0.032)

 Emergency

    

−0.132∗∗∗

     

(0.013)

 Internal medicine case

    

0.113∗∗∗

     

(0.034)

 Surgical case

    

−0.065·

     

(0.033)

 MRI

    

0.115∗∗∗

     

(0.032)

 CT

    

0.082∗∗

     

(0.030)

 X-ray

    

0.091∗∗

     

(0.030)

 Complicated radiology case

    

−0.104∗∗∗

     

(0.020)

Adj. R2

0.017

0.009

0.323

0.326

0.355

Num. obs.

11371

11371

11371

11371

11371

  1. ∗∗∗ p<0.001, ∗∗ p<0.01, ∗ p<0.05, · p<0.1 Table indicating heteroskedasticity-consistent robust standard errors and p-values with omittted variable bias by excluding the interaction term
Back to article page

BMC Health Services Research

ISSN: 1472-6963

Contact us