Oxygen Delivery Methods in a Model of a Spontaneously Breathing Tracheostomized Child

Abstract

Background:

Heated tracheostomy collar (HTC) with different interfaces (direct connection or tracheostomy mask) is used to provide humidification of inspired air to spontaneously breathing tracheostomized children. At our hospital we use a HTC with a blender system (FIO₂ 0.21). If supplemental O₂ is needed, low flow O₂ (FIO₂ 1.0) is given through a tracheostomy ring (t-ring). This is done in the event the HTC circuit becomes disconnected. Sometimes, despite using a maximum flow, the desired O₂ saturation is not achieved. We hypothesized increasing the FIO₂from the blender, using a direct connection interface, and increasing t-ring flow would result in higher tracheal FIO₂s.

Methods:

We used a model of a spontaneously breathing tracheostomized child consisting of a 4.0 pediatric tracheostomy tube with cuff inflated (Bivona TTS) inserted into a tracheal model connected in series to a t-piece with O₂ analyzer (tracheal FIO₂), a filter and a breathing simulator (RR 30, V_T 140 mL, I-time 0.7 seconds). Oxygen was delivered through a t-ring attached to the tracheostomy (0-4 L/min), the HTC (15 L/min) at different FIO₂s, or with both. We tested 2 different interfaces: a direct connection adapter and a tracheostomy mask. Tracheal FIO₂was recorded after 5 minutes and each experiment was repeated 4 times. Interfaces were compared with T-test, and effect of flows and FIO₂s were compared with ANOVA. A P < .05 was considered statistically significant.

Results:

The direct connect provided higher tracheal FIO₂s than trach mask in most occasions. Increasing t-ring O₂ flow and increasing HTC FIO₂ increased tracheal FIO₂.

Conclusions:

The HTC used with blender FIO₂alone achieved higher tracheal Fi02s than the t-ring in a model of spontaneously breathing tracheostomized child. Increasing t-ring flow increased tracheal FIO₂ with HTC FIO₂ of 1. The addition of the t-ring O₂flow to blender FIO₂provided only minimal increases in FIO₂. When the blender was set to deliver 1.0 FIO₂, the addition of t-ring flow had no impact. In general, there were no significant differences in FIO₂between interfaces.

Results (X ± SD)

T-ring L/min	Blender Fi02	FiO2 (direct connection)	FiO2 (tracheostomy mask)
0	.21	0.211 ± 0.1	0.21 ± 0.3
0	.50	0.46 ± 0.1*	0.418 ± 0.7
0	1.0	0.881 ± 0.9*#	0.789 ± 0.9#
1	.21	0.286 ± 0.4	0.29 ± 0.3
1	.50	0.524 ± 0.1*	0.508 ± 0.3
1	1.0	0.882 ± 0.9*#	0.83 ± 1.7#
2	.21	0.347 ± 0.4	0.372 ± 0.2*
2	.50	0.549 ± 0.5	0.539 ± 1.3
2	1.0	0.877 ± 1.4*#	0.846 ± 1.2#
3	.21	0.41 ± 0.4	0.442 ± 0.4*
3	.50	0.615 ± 0.1*&	0.598 ± 0.3
3	1.0	0.877 ± 0.3#	0.89 ± 2.8#
4	.21	0.458 ± 0.8	0.513 ± 0.5*
4	.50	0.615 ± 0.8	0.613 ± 0.7
4	1.0	0.869 ± 1.5*#	0.832 ± 2.3#

*> than other interface (p < .05). #No difference with addition of t-ring flow (p < .0001). &No difference with increasing t-ring flow (p < .0001)

Get full access to this article

View all access options for this article.