Intrathoracic airway closure: a progressively increasing barrier to ventilation during cardiopulmonary resuscitation?

Intrathoracic airway closure has been shown to occur during cardiopulmonary resuscitation (CPR).1, 2 This phenomenon is associated with lung volume reduction caused by chest compressions, local inflammatory processes, and loss of chest wall muscle tone; potentially impairing oxygenation and ventilation.2, 3 Previous studies have quantified airway closure using the airway opening index, analyzing oscillations in the capnogram generated by chest compressions.¹

Last year, our group introduced a novel method to detect intrathoracic airway closure based on the interpretation of low flow pressure–volume curves obtained during CPR. Following intubation, manual low flow inflations were performed during rhythm checks. The presence of a lower inflection point (LIP) on the resulting pressure–volume curves indicated airway closure.⁴

The LIP is defined as the pressure corresponding to the intersection of two lines representing minimal and maximal compliance. It is influenced by the amount of potentially recruitable lung, regional mechanical differences in pulmonary tissue (e.g. heterogeneous pulmonary injury) and chest wall characteristics.⁵

In our study, a LIP was detected in all included patients. The median pressure at which the LIP occurred was 5.56 cmH20 (IQR 4.80, 8.23 cmH₂0). If more time had elapsed between the start of CPR and the low flow maneuver, the LIP was higher, indicating more extensive airway closure.⁴ It remains unclear whether airway closure varies during ongoing CPR. One animal study has suggested a slight increase in airway closure during resuscitation.²

To explore this further, we analyzed two consecutive low flow maneuvers in a convenience sample of 10 adult, non-traumatic out-of-hospital cardiac arrest patients treated by a prehospital medical emergency team. Six additional patients were recruited for this analysis, data from four patients who had received consecutive low flow maneuvers during the initial study were also included (patients 1, 4, 8 and 10 in Table 1). This study was approved by the Ghent University Hospital Ethics Committee (reference EC/2008/025/AM02). Deferred informed consent was obtained.

Table 1.

Detailed results per patient.

Patient	Age (years)	Sex	LIP (first maneuver) (cmH2O/min)	Difference between second and first measured LIP (cmH2O/min)	Time between cardiac arrest and first low flow inflation (min)	Type of chest compressions performed	Bystander CPR	Initial cardiac rhythm	Presumed cause of cardiac arrest	Return of spontaneous circulation
1	83	M	6.42	0.18	20	Manual	Yes	PEA	Pulmonary embolism	No
2	80	M	6.88	0.12	17	Manual	No	Asystole	Cardiac ischemia	No
3	75	F	8.66	0.20	24	Mechanical	Yes	Asystole	Aortic stenosis	No
4	56	M	1.53	0.39	10	Manual	Yes	VF	Cardiac ischemia	Yes
5	93	M	7.44	−0.01	20	Manual	Yes	Asystole	Cardiac ischemia	No
6	45	M	5.06	0.56	15	Manual	No	Asystole	Cardiac ischemia	No
7	90	M	10.5	0.55	29	Mechanical	Yes	Asystole	Arrhythmia	No
8	53	M	5.29	0.11	31	Mechanical	No	VF	Cardiac ischemia	Yes
9	83	M	9.18	0.25	15	Manual	No	Asystole	Ruptured thoracic aneursym	No
10	73	M	2.83	0.20	4	Mechanical	Yes	VT	Cardiac ischemia	Yes

In this sample, nine patients were male, median age was 78 years. Bystander CPR was performed in six patients; return of spontaneous circulation was achieved in three patients. Median time between cardiac arrest and the first low flow inflation was 19 min (IQR 15, 24).

A LIP was detected in all patients, median calculated first LIP was 6.65 cmH20 (IQR 5.12, 8.35 cmH₂0). An increase in LIP was observed in nine patients. In one patient, a decrease was noted. Observed median difference between the second and first measurement was 0.20 cmH₂O/min (IQR 0.14, 0.36 cmH₂O/min). Detailed results per patient are shown in Table 1.

Several limitations exist when using manual low flow inflations during CPR.⁴ The use of a convenience sample possibly introduced sampling bias. Adding sensors to the ventilatory circuit causes additional resistance. Manually obtaining pressure–volume curves reduces accuracy. During initial development of this method, repeated low flow maneuvers were performed in 10 post-cardiac arrest patients (first using the manual method, then using a ventilator). Comparing the LIP calculated from these 2 maneuvers, observed coefficient of variation was 8%. The small changes in LIP seen in the current study may result solely from measurement error when performing manual low flow inflations. To enhance accuracy, automated and more precise methods of performing low flow inflations should be developed (e.g. using a ventilator to perform a low flow maneuver during rhythm checks).

In conclusion, airway closure was present in all included cardiac arrest patients. A small increase in LIP during ongoing CPR was observed in most patients. However, this change may be attributable to measurement error associated with the method used, rather than reflect a true change in airway closure. Further research should seek more accurate methods of detecting airway closure and explore which factors influence the occurrence of airway closure. It should be explored whether the use of ventilatory strategies aimed at preventing or reversing airway closure (e.g. applying PEEP) could improve outcomes.⁶

CRediT authorship contribution statement

Maxim Vanwulpen: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Data curation, Conceptualization. Davinia Coopman: Writing – review & editing, Investigation, Formal analysis, Data curation. Thomas Tackaert: Writing – review & editing, Investigation. Christophe Duchatelet: Writing – review & editing, Investigation. Abdo Khoury: Writing – review & editing, Supervision, Conceptualization. Saïd Hachimi-Idrissi: Writing – review & editing, Supervision, Conceptualization.

Funding

This work was supported by a grant from the Zoll Foundation (internal grant number KW/2191/NKU/001/013).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.