Machine Learning Models for Analysis of Vital Signs Dynamics: A Case for Sepsis Onset Prediction

Table 3

Separation of populations by vital signs’ features.

	Septic patients		No sepsis
					p value
(AP)
n	21.71	8.48	11.17	10.94	<0.00001
Median	10.15	9.27	21.14	15.16	<0.00001
Mean	12.48	8.6	22.44	14.58	<0.00001
Min	3.35	7.08	13.64	14.53	<0.00001
Max	35.15	17.6	37.62	18.49	<0.00001
(HR)
n	23.56	6.32	22.96	5.88	<0.00001
Median	6.9	5.32	6.117	3.98	<0.00001
Mean	9.17	5.79	8.591	4.89	<0.00001
Min	1.31	0.88	1.231	0.75	<0.00001
Max	30.18	21.19	28.87	19.4	<0.00001
(RR)
n	15.76	10.02	11.53	11.3	<0.00001
Median	4.933	4.252	7.033	5.57	<0.00001
Mean	5.39	3.69	7.176	4.86	<0.00001
Min	1.83	2.86	4.086	5.22	<0.00001
Max	11.61	6.46	12	6.29	<0.00001
(TEMP)
n	7.76	5.64	8.73	5.02	<0.00001
Median	0.56	0.93	0.42	0.67	0.804374
Mean	0.75	0.91	0.63	0.67	<0.00001
Min	0.33	0.91	0.11	0.41	<0.00001
Max	1.68	1.5	1.72	1.46	<0.00001

AP: arterial pressure; HR: heart rate; RR: respiratory rate; TEMP: temperature; n: number of trend changes; Median: median intensity of change; Mean: mean intensity of change; Max: maximal intensity of change; Min: minimal intensity of change.