Hidden in plain sight: bias towards sick patients when sampling patients with sufficient electronic health record data for research

Table 3 Negative binomial regression

	Laboratory results				Medication orders
Variable	Overall variable effect p-value	Ratio of expected number of days(SE)^*	95% Confidence interval	Regression p-value	Overall variable effect p-value	Ratio of expected number of days(SE)^*	95% Confidence interval	Regression p-value
ASA class	<.001^†				<.001^†
1		1.00					1.00
2		1.47(1.03)	1.38 – 1.57	<.001^†		1.74(1.05)	1.56 – 1.94	<.001^†
3		3.38(1.04)	3.11 – 3.67	<.001^†		4.78(1.07)	4.18 – 5.48	<.001^†
4		5.05(1.07)	4.41 – 5.77	<.001^†		6.85(1.12)	4.49 – 8.54	<.001^†
Sex	<.001^†				<.001^†
Age	<.001^†				<.001^†
Emergency status	<.001^†				0.84
Admission status	<.001^†				<.001^†
ICD-9 category	<.001^†				<.001^†
CPT category	<.001^†				<.001^†

ASA Class = American Society of Anesthesiologists Physical Status Classification; ICD-9 = International Classification of Diseases, Ninth Revision; CPT = Current Procedural Terminology; SE = Standard Error.
*The effects are ratios, obtained by exponentiating the model regression coefficients, of the expected number of days (with either laboratory results or medication orders) for each ASA Class compared to ASA 1, that is, an effect of 2.0 for ASA 3 indicates that ASA 3 is estimated to have 2 times the number of days as ASA 1.
^†Statistically significant at the 0.05 significance level.

ISSN: 1472-6947