Investigating the cognitive capacity constraints of an ICU care team using a systems engineering approach

BMC Anesthesiology

Table 3 Comparisons of medication orders per patient per hour before and after the cutoff

Characteristic		Sample size	Cutoff	No. of medication orders per patient per hour (average; SD)		P
Characteristic		Sample size	Cutoff	Patient census≤Cutoff	Patient census>Cutoff	P
Overall		18,630	18	0.74(0.56)	0.65(0.48)	< 0.001
Time periods	Daytime	9294	18	0.79(0.53)	0.67(0.41)	< 0.001
	Nighttime	7784	18	0.61(0.47)	0.50(0.36)	< 0.001
	Rounding	1552	18	1.31(0.84)	1.15(0.71)	< 0.001
Severe patients	Low	16,252	18	0.74(0.56)	0.66(0.48)	< 0.001
Severe patients	High	2378	16	0.81(0.59)	0.63(0.47)	< 0.001
New patients	Low	10,170	18	0.63(0.52)	0.59(0.47)	< 0.001
New patients	High	8460	18	0.85(0.58)	0.74(0.48)	< 0.001
High mortality risk patients	Low	9623	18	0.72(0.56)	0.63(0.48)	< 0.001
High mortality risk patients	High	9007	18	0.77(0.56)	0.67(0.48)	< 0.001

The cutoff is the patient census where the rate of medication orders generated per patient started to change. The difference was tested by the two-sample t-test and the p-values are shown, for the overall (ungrouped) data, and subgroups, stratified by time and system workload factors, respectively. For workload factors, being “Low” indicates a low presence of the featured patients in the entire census. If a p-value is less than 0.05, the two quantities are considered significantly different

ISSN: 1471-2253