Abstract
In acute decompensated pulmonary hypertension, elevated nonrespiratory SOFA was associated with a higher use of invasive mechanical ventilation independently of the hypoxaemia on ICU admission and pulmonary hypertension severity https://bit.ly/3YdOHpn
To the Editor:
In acute decompensated pulmonary hypertension (ADPH) patients, invasive mechanical ventilation (IMV) use remains challenging, because it has deleterious haemodynamic effects on the right ventricle (RV), which can result in haemodynamic collapse and cardiac arrest [1, 2]. IMV may decrease RV function by reducing RV preload and raising the pulmonary vascular resistance and RV afterload, both decreasing cardiac output. When the compensatory mechanisms are exhausted, ventriculoarterial uncoupling occurs, with drops in cardiac index and pulmonary pressures, as well as increases in central venous pressure, late markers of RV failure and imminent cardiovascular demise [3]. Therefore, unless absolutely required, endotracheal intubation and IMV are strongly advised against [4, 5].
The aim of our study was to identify early predictors of the use IMV in patients with ADPH. Identification of these predictors can minimise complications and supports decision-making about the use of IMV or palliative care.
We conducted a secondary analysis of a larger multicentre retrospective cohort, which included adults with both pulmonary arterial hypertension (PAH) and chronic thromboembolic pulmonary hypertension who were admitted, unplanned, to the intensive care unit (ICU) due to ADPH [6]. ADPH was defined by low cardiac output and/or RV failure [7]. Patients who had previously made the decision to forgo life-sustaining therapies (do-not-resuscitate orders) were excluded. The data for this study were collected from electronic healthcare records spanning the period between January 2014 and December 2019.
To identify predictors for IMV use, we conducted univariate analysis and developed a multivariable logistic regression model. Variables with p<0.05 in univariate analysis were included in the multivariable logistic regression model employing the backward conditional method of logistic regression. There were no missing data for any of the variables of interest. Categorical and continuous data are presented as frequencies (percentages) and median (25–75% interquartile range (IQR)). Categorical variables were compared using the Chi-squared test or Fisher's exact test. Continuous variables were compared using the Mann–Whitney test. Odds ratios and 95% confidence intervals were used to measure the association between each variable and renal replacement therapy use. Data analysis was done using Statistical Package for SPSS software (version 23.0; IBM, Armonk, NY, USA).
74 patients with ADPH were included and 11 (14.9%) patients required IMV during the first 48 h of ICU stay. The median (IQR) age was 47 (35–65) years; 75.7% of the patients were female; and 64.9% had group 1 pulmonary hypertension. Baseline and ICU admission data are depicted in table 1. Most common reasons for ADPH were infection (48.6%), hypervolaemia (18.9%) and arrhythmia (8.1%), and in 20.2% of cases no clear cause for ADPH was identified. No cases of acute pulmonary embolism were documented.
Patient characteristics before and at admission to the intensive care unit
Patients in the IMV group were mainly classified as high-risk pulmonary hypertension based on European Respiratory Society/European Society of Cardiology pulmonary hypertension risk assessment (a scoring system that predicts 1-year mortality of patients with pulmonary hypertension) (63.6% versus 23.8%, p=0.03) and had a higher sequential organ failure assessment (SOFA) score upon ICU admission (median 9 (IQR 8–12) versus 6 (4–8), p<0.01). Hypoxaemia was common during ICU admission, median (IQR) arterial oxygen tension (PaO2)/inspiratory oxygen fraction (FIO2) was 144 (89–221). Nonrespiratory (NR)-SOFA represents SOFA without the respiratory components and was significantly higher in the IMV group compared to the non-IMV group (median 7 (IQR 6–10) versus 4 (3–6), p<0.01).
In our sample, the most common reasons for IMV were worsening hypoxia (54.5%), respiratory distress (18.2%), cardiac arrest (18.2%) and hypercapnia (9.0%). ICU mortality of patients who required IMV was 100%. The IMV group had a shorter hospital stay (2 days) compared to the non-IMV group (9 days); of note, 54.5% of patients in the IMV group had a cardiac arrest following intubation.
All IMV patients in our unit were ventilated with protective ventilation to reduce lung stress (low tidal volumes up to 6 mL·kg−1, limitation of plateau pressure <30 cmH20 and driving pressure <15 cmH20), normocapnia and prone positioning when indicated (PaO2/FIO2 ratio <150) [8].
Of note, during the time of this study, high-flow nasal cannula (HFNC) was unavailable and extracorporeal membrane oxygenation (ECMO) was rarely accessible.
NR-SOFA score and pulmonary hypertension risk assessment (independent variables) and IMV (dependent variable) were included in a multivariable logistic regression model. Higher NR-SOFA score was associated with increased IMV use (OR 1.77, 95% CI 1.21−2.57), while pulmonary hypertension risk assessment was not significantly associated with IMV use. The area under the receiver operating characteristic curve of the NR-SOFA score to predict IMV was 0.83 (0.71–0.95, p<0.01), suggesting good discrimination. In a sensitivity analysis, neither pulmonary hypertension group nor subtype of PAH had a significant impact on the use of IMV in our cohort.
The respiratory component of SOFA was similar between IMV group and non-IMV group, which is explained by the fact patients were similarly hypoxic at admission. NR-SOFA was heavily influenced by the cardiovascular component, reflecting more use of inotropes and vasopressors in the IMV group before intubation, and also influenced by acute kidney failure, highly prevalent on ICU admission.
Use of IMV was noted to be associated with poor prognosis and elevated mortality in many cohorts [6, 9, 10, 11]. In-hospital mortality was 100% in the IMV group in our cohort, similar to the 100% mortality of IMV patients described by Campo et al. [9]. In the study by Sztrymf et al. [12], no patient underwent IMV in the study period, as the authors stipulated that no intubations would be performed due to the poor prognosis associated with IMV. In spite of poor prognosis related to IMV, carefully selected patients with potentially reversible acute RV dysfunction could be considered for IMV [13].
The high mortality observed in the IMV group was explained by peri-intubation cardiac arrest, but probably reflects that sicker patients tend to receive IMV. NR-SOFA offers an appropriate refinement for the prediction of IMV, since it reflects how sick those patients were, apart from the hypoxaemia, which was common in this population, and should not be taken as an isolated indication for IMV. Oxygen induces pulmonary vasodilation in PAH patients and it can improve RV cardiac output. Therefore, the ideal method of oxygenation is the one that causes the least negative impact in the pulmonary circulation.
There is scarce literature on the alternatives to IMV in this high-risk population. HFNC was associated with improved oxygenation in ADPH in a preliminary study [14]. Physiological effects of HFNC in ADPH are promising, but its role remains unclear and we need further investigation in this area.
We were unable to assess the impact of noninvasive ventilation (NIV) due to a lack of consistent data on NIV in patient charts. It is worth noting that NIV was generally avoided in ADPH patients in our units.
ECMO may be considered for patients with refractory ADPH due to severe hypoxaemic respiratory failure in whom conventional support is not resulting in improvement, as a bridge to lung transplantation or a bridge to recovery. If possible, ECMO should preferably be used in awake and spontaneously breathing patients, who are non-intubated, to reduce risks related to general anaesthesia vasodilation and prevent the negative consequences of positive pressure [15].
NR-SOFA was associated with a higher use IMV in ADPH, independent of hypoxaemia and the baseline severity of pulmonary hypertension. In ADPH, IMV should be carefully considered, and alternative modes of oxygenation such as HFNC and ECMO should be considered.
Footnotes
Provenance: Submitted article, peer reviewed.
Author contributions: M.V.F. Garcia made substantial contributions to the conception, design, acquisition, analysis and interpretation of the work, and drafted the work and approved the version to be published. R. Souza made substantial contributions to the conception of the work, analysis and interpretation of data, revised for critically important intellectual content, and approved the final version to be published. P. Caruso made substantial contributions to the conception, design, analysis, and interpretation of the work, revised it for critically important intellectual content, and approved the version to be published.
Conflict of interest: Nothing to disclose.
- Received November 7, 2022.
- Accepted January 29, 2023.
- Copyright ©The authors 2023
This version is distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0. For commercial reproduction rights and permissions contact permissions{at}ersnet.org