Prone ventilation in H1N1 virus-associated severe acute respiratory distress syndrome: A case series.

BACKGROUND: Management of H1N1 viral infection-associated acute respiratory distress syndrome (ARDS) has primarily been focused on lung protective ventilation strategies, despite that mortality remains high (up to 45%). Other measures to improve survival are prone position ventilation (PPV) and extracorporeal membrane oxygenation. There is scarcity of literature on the use of prone ventilation in H1N1-associated ARDS patients.
METHODS: In this retrospective study, all adult patients admitted to medical intensive care unit (ICU) with H1N1 viral pneumonia having severe ARDS and requiring prone ventilation as a rescue therapy for severe hypoxemia were reviewed. The patients were considered to turn prone if PaO2/FiO2 ratio was <100 cmH2O and PaCO2 was >45 cmH2O; if no progressive improvement was seen in PaO2/FiO2 over a period of 4 h, then patients were considered to turn back to supine. Measurements were obtained in supine (baseline) and PPV, after 30-60 min and then 4-6 hourly.
RESULTS: Eleven adult patients with severe ARDS were ventilated in prone position. Their age range was 26-59 years. The worst PaO2/FiO2 ratio range on the day of invasive ventilation was 48-100 (median 79). A total of 39 PPV sessions were done, with a range of 1-8 prone sessions per patient (median three sessions). Out of the 39 PPV sessions, PaO2/FiO2 ratio and PaCO2 responder were 38 (97.4%) and 27 (69.2%) sessions, respectively. The median ICU stay and mechanical ventilation days were 15 (range: 3-26) and 12 (range: 2-22) days, respectively. The common complication observed due to PPV was pressure ulcer. At ICU discharge, all except two patients survived.
CONCLUSION: PPV improves oxygenation when started early with adequate duration and should be considered in all severe ARDS cases secondary to H1N1 viral infection. Copyright:

INTRODUCTION

Post pandemic 2009 H1N1 viral infection, multiple small epidemic outbreaks have been reported worldwide.[1] Mortality up to 45% among patients admitted to intensive care units (ICUs) has been reported in a recent epidemic.[23] The main cause of death in patients with pneumonia caused by H1N1 viral infection is multiorgan failure secondary to hypoxemia because of acute respiratory distress syndrome (ARDS) and secondary bacterial infection.[45] As with other causes of ARDS, management of H1N1 viral infection ARDS is focused on lung protective ventilation (LPV) strategies, but despite that mortality remains high.[6] Other measures to improve survival include prone position ventilation (PPV) and extracorporeal membrane oxygenation (ECMO); although enough evidences showed that PPV reduces mortality (16%) compared to supine position ventilation (32.8%),[278] there are underutilization of this strategy, which is found in recent trials.[59] In addition, there is scarcity of literature on the use of prone ventilation in H1N1 ARDS patients.[101112] We studied the effect of prone ventilation in severe ARDS patients admitted with the diagnosis of H1N1 during an epidemic in the winter months of 2014–2015 in a southern state of India.

METHODS

This study was conducted at a tertiary health-care center in southern part of India. All patients admitted to medical ICU with H1N1 viral pneumonia having severe ARDS requiring prone ventilation as a rescue therapy for severe hypoxemia were reviewed prospectively. Polymerase chain reaction-confirmed H1N1 was included in this series. ARDS was defined as per the Berlin definition.[13] Patients with cardiogenic pulmonary edema and spinal instability were excluded from the study. All the included patients were studied while sedated with fentanyl and midazolam and paralyzed with atracurium. The ventilatory settings were set as per the treating physician. Positive end-expiratory pressure (PEEP) and FiO2 were adjusted according to the PEEP-FiO2 table of ARDSnet trial.[6] Measurements were obtained in supine (baseline) and PPV, after 30–60 min and then 4–6 hourly. The patients were considered to turn prone if PaO2/FiO2 ratio was <100 cmH2O and PaCO2 was >45 cmH2O. Patients were classified as nonresponders if the PaO2/FiO2 ratio does not increase by >20% and PaCO2 decrease by 6 cmH2O. In PPV, if no progressive improvement was seen in PaO2/FiO2 and PaCO2 over a period of 4 h, then the patients were considered to turn supine. In this descriptive study, parametric and nonparametric data were presented as means (standard deviation) and medians, respectively.

RESULTS

Between September 1, 2014, and March 15, 2015, 93 patients were treated for H1N1 viral pneumonia. Of these, 11 adult patients developed severe ARDS which did not responded to lung-protective ventilator strategies and were ventilated in prone position [Table 1]. Their age range was 26–59 years (median: 48), with four patients under the age of 40 years. Seven patients were female, including one postpartum. All presented with fever, cough, and dyspnea for 4–12 days. The number of days from symptom onset to invasive ventilation ranged from 7 to 12 days (median: 8). Five patients had comorbid illness, the common being morbid obesity and diabetes mellitus. On the day of ICU admission, the median Acute Physiological and Chronic Health Evaluation-II and Sequential Organ Failure Assessment scores were 20 (range: 16–27) and 7 (range: 5–12), respectively. Noninvasive ventilation trial was given in eight patients before invasive ventilation, out of which two required invasive ventilation within 12 h and the rest in 48 h. The worst PaO2/FiO2 ratio range on the day of invasive ventilation was 48–100 (median: 79). Out of 11 patients, PPV was started within 24 h in 8 patients and within 48 h in 3 patients. Preprone, ventilator, and blood gas parameters are summarized in Table 2; the response of PaO2/FiO2 ratio to prone ventilation is shown in Figure 1. Four patients had cardiac dysfunction, with one having viral myocarditis (global hypokinesia ejection fraction of 23%), and three patients had pulmonary arterial hypertension with acute right heart failure. A total of 39 PPV sessions were done, with a range of 1–8 prone sessions per patient (median: 3). A total of 580 h of prone ventilation was done with an average of 14 h 52 min/session (range: 8–24) of prone ventilation. Out of the 39 PPV sessions, PaO2/FiO2 ratio and PaCO2 responder were 38 (97.4%) and 27 (69.2%) sessions, respectively.

Table 1

Demographic and clinical characteristic of patients

Patient number	Age	Gender	Comorbid	At ICU admission		Days of illness before invasive ventilation	Severity of ARDS: PaO2/FiO2 ratio on the day of prone ventilation	Intubation to prone ventilation in days	Number of prone ventilation sessions	MV days	ICU days	Outcome at discharge
				APACHE-II score	SOFA score
1	48	Female	-	27	12	8	55	1	4	12	16	Survived
2	55	Female	-	22	7	8	100	2	5	15	19	Survived
3	40	Female	DM	20	8	8	57	0	3	14	15	Survived
4	59	Female	HTN, COPD	18	5	8	79	1	5	19	22	Survived
5	52	Male	-	19	10	7	75	2	8	16	21	Survived
6	52	Male	CKD, Ca tongue on chemotherapy	24	10	8	73	1	3	9	10	Died
7	35	Male	-	19	5	12	79	1	2	7	9	Survived
8	44	Female	DM, morbid obesity	22	7	9	90	2	4	22	26	Survived
9	26	Female	-	16	6	9	90	1	2	12	14	Survived
10	54	Male	Morbid obesity	23	10	11	48	0	1	3	3	Died
11	26	Female	Postpartum	16	6	8	82	1	2	9	12	Survived

APACHE-II score: Acute Physiological and Chronic Health Evaluation score, SOFA score: Sequential Organ Failure Assessment score, ARDS: Acute respiratory distress syndrome, PaO2/FiO2: Partial pressure of oxygen in arterial blood/fractional inspired oxygen concentration, MV: Minute ventilation, ICU: Intensive care unit, DM: Diabetes mellitus, HTN: Hypertension, COPD: Chronic obstructive pulmonary disease, CKD: Chronic kidney disease, Ca: Carcinoma

Table 2

Ventilator parameters and blood gases before the first prone session

Patient number	PBW	Preprone ventilator parameters
		FiO2	PaO2/FiO2 ratio	PCO2	PEEP	Plateau pressure	Compliance	Tv	LIS	MV
1	58	90	62	49	12	31	18	360	3.5	10.6
2	60	80	48	58	14	32	16	360	3.75	12
3	52	80	57	38	12	31	18	320	3.5	10
4	56	70	100	56	11	30	28	340	3	13
5	70	80	75	38	14	31	22	420	3	11.8
6	62	70	52	40	12	31	22	370	3	10.6
7	68	65	80	57	14	32	20	400	3	10.2
8	56	80	70	49	14	32	14	340	3	12.4
9	52	75	90	40	12	30	19	320	2.5	10.2
10	60	100	50	60	16	35	14	390	3.25	13
11	52	80	80	44	14	32	14	339	3	12

PBW: Predicted body weight (kg), PaO2/FiO2: Partial pressure of oxygen in arterial blood/fractional inspired oxygen concentration, PEEP: Positive end-expiratory pressure (cmH2O), Tv: Inspired tidal volume (ml), MV: Inspired minute ventilation (L/min), LIS: Lung injury score

Figure 1

Changes in PaO2/FiO2 ratio in response to prone position ventilation session. Supine PaO2/FiO2: Before prone position; prone PaO2/FiO2: End of prone position ventilation

The median ICU stay and mechanical ventilation days were 15 (range: 3–26) and 12 (range: 2–22) days, respectively. In five patients, tracheal aspirate showed bacterial growth (3 – Acinetobacter baumannii, 1 – Pseudomonas aeruginosa, and 1 – Klebsiella pneumoniae) and all survived. Apart from respiratory failure, two patients required vasopressor therapy and two had cardiovascular and renal failure, of which both patients died. Seven patients were tracheotomized. The median duration of spontaneous breathing trial from the day of prone ventilation was 8 (range: 5-16) days. The complications were observed as pressure ulcer in four patients; while back pain and critical illness neuromyopathy in one each patient. Two patients died due to shock and multiorgan failure.

DISCUSSION

In this single-center, retrospective observational study, all H1N1 infection-associated severe ARDS patients showed improvement in PaO2/FiO2 ratio (>20%) after the first prone session, with many of them undergoing multiple prone sessions also. There are multiple factors which may account for the benefit in PPV such as reduction in dependent atelectasis and pleural pressure gradient,[14] alveolar recruitment,[15] homogeneous ventilation,[16] and relief of the compressive segment of the lung by the heart[1718] and abdomen.[19] All these factors cause reduction in ventilation and perfusion mismatch, intrapulmonary shunt, and ventilator-induced lung injury (VILI). These all lead to improved oxygenation and decreased PaCO2. The oxygenation and PaCO2 response to PPV are variable and depend on the type of lung injury, percentage of recruitable lung tissue,[20] and duration of PPV.[421] Patients with diffuse lung injury have less nonaerated lung tissue and respond to prone ventilation by increase in oxygenation without decrease in PaCO2, whereas lobar lung injury has significant percentage of nonaerated recruitable lung tissue and shows marked decrease in PaCO2 along with improvement in oxygenation.[15] Alveolar recruitment in PPV is time dependent, with different nonaerated lung units having different plateaus for complete alveolar recruitment.[22]

Till 2013, a large number of randomized controlled trials were published which did not show any mortality benefit of PPV in ARDS patients, although improvement of oxygenation was seen.[21232425262728] Many reasons have been put forth for nonimprovement in mortality in earlier studies such as lack of power in the study, too short PPV, heterogeneous patient population, or improved oxygenation does not translate to decrease mortality as multiorgan failure being the most common cause of mortality followed by hypoxemia in ALI or ARDS patients.[2930]

We studied the effect of prone ventilation in severe ARDS due to H1N1, with early initiation, prolonged duration, and multiple sessions. Our study replicated the findings of Guérin et al. regarding the beneficial effect of PPV in severe ARDS when started early with prolonged duration.[8] Worldwide, ARDS is associated with substantial mortality, with unadjusted ICU and hospital mortality of 35.3% (95% confidence interval [CI], 33.3%–37.2%) and 40.0% (95% CI, 38.1%–42.1%), respectively.[5] H1N1 viral infection ARDS causes 36%–38% mortality in accordance with other causes of ARDS.[331]

Overall, worldwide, PPV is used in 7.9% (95% CI, 6.8–9) in all patients and 16.3% (95% CI 13.7–19.2) in severe ARDS patients.[5] As with other causes of ARDS, management of H1N1 viral infection ARDS is focused on LPV strategies, and PPV is a very safe and effective way of achieving improved oxygenation and preventing VILI.[1232] In recent published trials worldwide on severe H1N1 viral ARDS, only 20%, 45%, 34%, 28%, and 47% of patients were subjected to prone position ventilated before starting ECMO in Australian, French, UK, Italian, and German studies, respectively.[2373133]

Our study differs from that of previous with regard to the optimal duration of PPV. Earlier studies with prone ventilation used short duration PPV (8 h) with no mortality benefit,[2324] but later studies which used longer duration PPV showed mortality benefit when used in severe ARDS patients with high percentage of nonaerated lung unit.[821] In our studies, we used variable PPV duration (average: 14 h 52 min, range: 8–24 h) and continued until we did not find any improvement in oxygenation, decrease in PaCO2, or increase in tidal volume over a period of 4 h. In the present study, according to clinician's decision, multiple prone sessions were considered till PaO2/FiO2 ratio sustained >150 in supine position. One patient had only one prone session, and later he died [Figure 1].

CONCLUSION

Our study demonstrates that PPV improves oxygenation when started early with adequate duration and should be considered in all severe ARDS cases secondary to H1N1 viral infection.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Ethical conduct of research

This study was approved by the Institutional Review Board / Ethics Committee. The authors followed applicable EQUATOR Network (http://www.equator-network.org/) guidelines during the conduct of this research project.