Multi-infusion with integrated multiple pressure sensing allows earlier detection of line occlusions

Table 2 Validation set alarm delays in minutes for the regression and SD algorithms compared to calculated conventional alarm delays with an alarm threshold set at 400 mmHg

Syringe (product name, internal volume)	IV tube (product name, internal diameter, length)	N	Rate (ml/h)	Detection algorithm	Alarm delay mean ± SD (minutes)	Conventional alarm delay mean ± SD (minutes)¹	P-value²
BD Plastipak 20 ml	MPH Medical Devices 12040150E, Ø 1.0 mm, 1500 mm	2³	1	SD	0.3 ± 0.2	5.9 ± 1.2	0.08
		3		Regression	1.1 ± 0.3	6.3 ± 1.1	< 0.05
		2³	5	SD	0.3 ± 0.1	1.5 ± 0.5	0.20
		3		Regression	0.9 ± 0.1	1.4 ± 0.4	0.18
		3	25	SD	0.2 ± 0.0	0.3 ± 0.1	0.06
				Regression	N.A.⁴		N.A
Fresenius Injectomat 50 ml	MPH Medical Devices 12040150E, Ø 1.0 mm, 1500 mm	3	1	SD	0.7 ± 0.2	37.4 ± 9.9	< 0.05
		3		Regression	5.4 ± 1.9	37.4 ± 9.9	< 0.05
		3	5	SD	0.4 ± 0.0	5.0 ± 0.7	< 0.01
		3		Regression	1.5 ± 0.1	5.0 ± 0.7	< 0.01
		3	25	SD	0.3 ± 0.0	1.1 ± 0.2	< 0.05
		3		Regression	1.2 ± 0.2	1.1 ± 0.2	< 0.05
Fresenius Injectomat 50 ml	Codan 71.4021, Ø 3.0 mm, 1500 mm	3	25	SD	0.4 ± 0.0	1.9 ± 0.1	< 0.01
		3		Regression	1.4 ± 0.1	1.9 ± 0.1	< 0.05
		3	50	SD	0.3 ± 0.0	0.8 ± 0.0	< 0.001
		3		Regression	1.4 ± 0.1	0.8 ± 0.0	< 0.01
		3	75	SD	0.2 ± 0.0	0.5 ± 0.0	< 0.001
				Regression	N.A.⁴		N.A

¹Calculated alarm delay for our local threshold setting of 400 mmHg
²Paired Student t-test
³A case was excluded due to the occurrence of a false positive that overlapped with the onset of the occlusion
⁴The regression algorithm was not able detect an occlusion in this case (i.e. a false negative classification occurred)
In order to test the robustness of the algorithms, syringes and IV tubes were used that were different from those used to develop the algorithms

ISSN: 1472-6947