Predictors for prescription of noninvasive ventilation in the postoperative period of cardiac surgery: a systematic review

Abstract

Introduction

The postoperative (PO) period after cardiac surgery is associated with the occurrence of respiratory complications. Noninvasive positive pressure ventilation (NIPPV) is largely used as a ventilatory support strategy after the interruption of invasive mechanical ventilation. However, the variables associated with NIPPV prescription are unclear.

Objective

To describe the literature on predictors of NIPPV prescription in patients during the PO period of cardiac surgery.

Materials and methods

This systematic review was registered on the International Prospective Register of Systematic Reviews (PROSPERO) platform in December 2021 (CRD42021291973). Bibliographic searches were performed in February 2022 using the PubMed, Lilacs, Embase and PEDro databases, with no year or language restrictions. The Predictors for the prescription of NIPPV were considered among patients who achieved curative NIPPV.

Results

A total of 349 articles were identified, of which four were deemed eligible and were included in this review. Three studies were retrospective studies, and one was a prospective safety pilot study. The total sample size in each study ranged from 109 to 1657 subjects, with a total of 3456 participants, of whom 283 realized NIPPV. Curative NIPPV was the only form of NIPPV in 75% of the studies, which presented this form of prescription in 5–9% of the total sample size, with men around 65 years old being the majority of the participants receiving curative NIPPV. The main indication for curative NIPPV was acute respiratory failure. Only one study realized prophylactic NIPPV (28% of 32 participants). The main predictors for the prescription of curative NIPPV in the PO period of cardiac surgery observed in this study were elevated body mass index (BMI), hypercapnia, PO lung injury, cardiogenic oedema and pneumonia.

Conclusions

BMI and lung alterations related to gas exchange disturbances are major predictors for NIPPV prescription in patients during the PO period of cardiac surgery. The identification of these predictors can benefit clinical decision-making regarding the prescription of NIPPV and help conserve human and material resources, thereby preventing the indiscriminate use of NIPPV.

Introduction

The postoperative (PO) period of cardiac surgery [1,2] is associated with the occurrence of respiratory complications – as a result of median sternotomy, surgical manipulation, general anaesthesia or cardiopulmonary bypass (CPB) – with pleural effusion, atelectasis, pneumonia, pulmonary oedema, phrenic nerve injury, pulmonary embolism, cardiogenic pulmonary oedema and pneumothorax, being the most common after interruption of invasive ventilatory support [3–6].

Thus, respiratory complications can impair lung function (by the reduction in forced vital capacity (FVC) and forced expiratory volume at the first minute of expiration (FEV1), and by the reduction of alveolar ventilation associated with pathological conditions or pain in the immediate PO period), favouring the progress of hypoxemia and acute respiratory failure (ARF) and the necessity of return for invasive ventilatory support [5,7–11].

Currently, noninvasive positive pressure ventilation (NIPPV) is widely used as a ventilatory support strategy after interruption of invasive mechanical ventilation (IMV) to maintain or improve alveolar expansion, targeting the optimization of gas exchanges, reduction of atelectasis in poorly ventilated areas, reduction of respiratory work, and improvement of haemodynamics and hypoxemia [10,12,13]. The use of NIPPV during the PO period of cardiac surgery supposedly improves heart rate (HR), myocardial performance, peripheral oxygenation, vital capacity (VC) and respiratory rate, decreasing the risk of ARF, reducing the need for reintubation, and increasing survival rates, respectively [13–16]. Thus, ARF is a significant contributor to morbidity, mortality, increased length of hospital stay, and consequently to health costs after cardiac surgery, and respiratory intervention strategies to prevent ARF, such as NIPPV, could improve patients’ outcomes [17].

However, the indication for prophylactic use of NIPPV is still controversial [17,18], and its application should be limited to patients at greater risk of developing PO ARF, since there is evidence that the indiscriminate use of NIPPV does not reduce the reintubation rates and length of hospital stay, probably due to lack of definition of appropriate predictors [13,16,19], leading to increased financial and professional resources [18]. Moreover, the inverse association between the number of physiotherapists in cardiac intensive care units and length of hospital stay is related to routine physiotherapeutic interventions other than NIPPV prescription, such as pain management assistance, positioning, cardiorespiratory rehabilitation, inspiratory muscle training, and application of expansive and clearance techniques [20], which also prevent PO respiratory complications, reducing the length of hospital stay, ICU readmission rates, and, consequently, hospital costs [21–23]. Thus, the indiscriminate prescription of prophylactic NIPPV can increase the physiotherapist’s assistance time, negatively impacting the wide performance in rehabilitation settings, burdening the health system, and slowing the recovery of patients.

Therefore, this study aimed to review the literature on the potential predictors for the prescription of NIPPV in patients during the PO of cardiac surgery.

Materials and methods

This study represents a systematic review conducted in accordance with the guidelines outlined in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [24], as well as in compliance with the Tutorial for Writing Systematic Reviews [25].

Search strategy

The research question was designed using the PICO strategy as follows: (P) Population: adult patients in the PO period of cardiac surgery; (I) Intervention: NIPPV; (C) Comparator: patients not submitted to NIPPV; (O) Outcomes: evaluation of clinical, functional and laboratory variables associated with the indication of NIPPV, and as a secondary outcome, to describe the mode of prescription for NIPPV, success, length of hospital stay in the ICU, and mortality rate of patients submitted to NIPPV.

The systematic review was registered on the International Prospective Register of Systematic Reviews (PROSPERO) platform in December 2021, and bibliographic searches were performed in February 2022 using the PubMed, Lilacs, Embase and PEDro databases, without year and language restrictions. The Medical Subject Headings descriptors were ‘Thoracic surgery’, ‘Noninvasive Ventilation’, ‘Cardiac surgical procedures’ and ‘Postoperative Period’. The complete search strategy can be found at https://osf.io/fb9xw.

Eligibility, inclusion and exclusion criteria

Studies published in any language and year describing the use of NIPPV as an intervention in adult patients in the PO of cardiac surgery in intensive care units were eligible for the review. Case Reports, Observational Studies, Reviews and Systematic Reviews were included in this search. Review studies were included in the searches because systematic reviews are often indexed in the databases only by the term ‘review’; therefore, this term was used to include the largest possible number of systematic review studies. Abstracts of congress, proceedings of events, and narrative reviews were excluded from the study.

First, the principal investigator searched all databases using Covidence systematic review software (Veritas Health Innovation, Melbourne, Australia, available at www.covidence.org) to identify and eliminate duplicated articles. Two different researchers read the titles and abstracts of all remaining articles, excluding those unrelated to the review topic. Finally, two other researchers read the full texts of the selected articles, extracted the data, and assessed the methodological quality.

Data extraction

Two independent reviewers extracted the data, ensuring blindness to each other’s research decisions. Discrepancies were resolved by the principal investigator. The extracted data from the studies encompassed the following variables: authors, year of publication, country of origin, sample size, sex and age of the study population, as well as the motivation for prescribing NIPPV (prophylactic or curative); intervention protocols (initial ventilatory support, adjustments and duration of NIPPV); NIPPV outcome (describing the occurrence and predictors of failure (as described in the original paper) or success (as described in the original paper) of NIPPV therapy); and possible predictors for NIPPV prescription (Table 1). The variables occurring before the acute need for NIPPV were considered predictors of NIPPV prescription, such as morbidities or physiological alterations conditions occurring previously of surgery or in the PO period of cardiac surgery.

Table 1.

Main results of studies that evaluated patients in the PO of cardiac surgery.

Study	Study design	Sample size Gender Age	Intervention (preventive/curative)	Intervention protocols	Fail or success of NIPPV	Possible predictors for NIPPV prescription
De Santo et al. [26]	Pilot prospective study	Total: 465; Male: 32; Female: 11; Age: 65.7 ± 9.4 years.	Performed NIPPV: 43 (9.2%); Curative: 100%; (Reason: ARF after extubation).	Initial ventilatory support: CPAP by flow generator wearing helmet; EPAP: 7.5 cmH2O; Adjustments: EPAP can be reduced to 5 or increased to 10 cmH2O. Goal: PaO2 >70 mmHg. Duration: 33.8 ± 24.0 h.	NIPPV failure: 25.6% Predictors of failure: Pneumonia (9 patients): 55% reintubated NIPPV success: 74.4% Predictors of success: Lung injury after CPB: 21 (90.5%) Cardiogenic dysfunction: 13 (69.2%)	Post-CPB lung injury: 21 (48.8%) Cardiogenic edema: 13 (30.2%); Pneumonia: 9 (21%).
García-Delgado et al. [27]	Retrospective observational study	Total: 1225; Male: 31; Female: 32; Age: 66 ± 11 years.	Performed NIPPV: 63 (5.1%) Curative: 100% (Reason: ARF after extubation).	Initial ventilatory support: PEEP 4–6 cmH2O; IPAP 10–14 cmH2O; Adjustments: In patients with persistent hypoxemic ARF (SaO2 <90%), PEEP was increased in steps of 2 cmH2O. Minimum FIO2 required for SaO2 >92%. Patients with hypercapnic ARF (PaCO2 >45 mmHg) required IPAP increments of 2 cmH2O to adjust VT and RR. Duration: Not reported.	NIPPV failure: 30 (48%); Predictors of failure: Duration of spontaneous breathing (from extubation to NIPPV application) <24 h increased 4.6 times the rate of failure OR:4.6 (IQ 1.2–17.9) NIPPV success: 33 (55%); Predictors of success: Obesity – BMI >30 kg.m−2 OR: 0.22 (IQ 0.05–0.91).	Lobar atelectasis; Acute cardiogenic edema; Acute lung injury Hypercapnia (PaCO2 >45 mmHg); Pneumonia; Others (pneumothorax, shock); BMI > 30 kg.m−2
Daniel et al. [28]	Retrospective observational study	Total: 109; Male: 66; Female: 43. Age: Not reported.	Performed NIPPV: 32 (29%); Curative: 72%; (Main reason: hypercapnia). Prophylactic: 28%.	Initial ventilatory support: Not reported Adjustments: After 2 h of NIPPV installation, the GSA analysis was performed. Patients with improved oxygenation and/or hypercapnia remained on the treatment. Duration: 3×/day for a period of 2 h, until resolution of the clinical condition.	NIPPV failure: 2 (6%). One hypoxemic and 1 hypercapnic patient; Predictors of failure: MIP <80 cmH2O, 2 was not associated with failure of NIPPV; NIPPV success: 30 (94%); Predictors of success: Hypercapnic, n (%): 13 (93); bronchospasm, n (%): 2 (100); Hypoxemia: 1 (50%); ARF: 5 (100%)	Curative NIPPV: Hypercapnia; Bronchospasm; Hypoxemia. Preventive NIPPV: Radiographic alterations (atelectasis and pleural effusion); Post-extubation.
Ampatzidou et al. [29]	Retrospective cohort study.	Total: 1657 Male: 106 Female: 52 Age: 68 ± 3 years.	Performed NIPPV: 145 (8.8%) Curative: 100%; (Reason: ARF).	Initial ventilatory support: PEEP: 4–5 cmH2O; IPAP: 10–12 cmH2O. Adjustments: Gradual increase in IPAP and/or PEEP during the first 2 h of application, aiming for a PaO2 of 60 mmHg, without acidosis. Duration: 16 h, in a period of 24 h.	NIPPV failure: 16 (11%) Predictors of failure: Older age: 75.5 (9) OR: 1.09 (IQ 1.01–1.17) Higher EuroSCORE II, SD: 2.6 (5.8) OR: 1.20 (IQ 1.03–1.39) Chronic atrial fibrillation, n (%): 6 (37.5) OR: 3.95 (IQ 1.27 − 12.28) COPD, n (%): 8 (50) OR: 4.38 (IQ 1.49–12.83) Heart failure (NYHA ≥2), n (%): 13 (81.3) OR: 3.54 (IQ 1.01–13.02) PO stroke, n (%): 3 (18.8) OR: 3.69 (IQ: 1.77–13.01) Renal replacement therapy, n (%): 8 (50) OR: 8.80 (IQ: 1.03–23.06) NIPPV success: 129 (89%) Predictors of success: Younger age, y: 69 (13) Higher BMI (m.kg−2): 31.2 (7.5) A lower burden of comorbidities: Chronic atrial fibrillation, n (%) 17 (13.2) COPD, n (%): 24 (18.6) Previous cardiac surgery, n (%): 1 (0.8) Heart failure (NYHA ≥2), n (%): 71 (55) Myocardial infarction: 0 PO stroke, n (%): 3 (2.3) Renal replacement therapy, n (%): 1 (0.8)	BMI (kg.m−2): NIPPV: 31.2 ± 5.2; No NIPPV: 28.7 ± 8.3; OR: 1.02 (IQ 1.01–1.04). EuroSCORE II (points): NIPPV: 3.1 ± 2.8; No NIPPV: 1.9 ± 2.7; OR: 1.11 (IQ 1.02–1.32). COPD (%): NIPPV: 22.8; No NIPPV: 5.5; OR: 4.0 (IQ 2.53–8.93). Negative predictor: Estimated glomerular filtration rate (cutoff: 60 mL.min−1.1.73 m2(−1)) [22]: NIPPV: 65.4 ± 20.5; No NIPPV: 76.4 ± 23.5; OR: 0.99 (IQ 0.98–0.99).

Analysis of the methodological quality of the studies

Clinical trials were evaluated using the Physical Therapy Evidence Database of Residents in the PEDro database (PEDro Scale), which comprises the following criteria: (1) specification of eligibility criteria; (2) random allocation of individuals into groups; (3) concealment of allocation; (4) similarity of groups at baseline concerning crucial prognostic indicators; (5) blinding of all subjects; (6) blinding of all therapists administering the therapy; (7) blinding of all evaluators assessing at least one key result; (8) transience of measures for at least one key outcome in over 85% of individuals initially allocated to groups; and (9) administration of treatment or control condition as allocated to all subjects with available outcome measures, or, if not, analysis of data for at least one key outcome by ‘intention to treat’; (10) reporting of results of statistical comparisons between groups for at least one key outcome; (11) provision of point measures and measures of variability for at least one key outcome [30]. Retrospective observational studies were assessed using the Newcastle-Ottawa Scale (NOS) (Table 2), a specific instrument for case-control and cohort studies comprising three categories and eight items: selection (four items), comparability (one item) and achievement (three items). Each numbered item within the categories could receive a maximum of one star (*), whereas a maximum of two stars could be attributed to each item in the comparability category. The NOS ranged from 0 to 9, with 9 indicating the highest quality on this semi-quantitative scale [31]. If case report studies were included, the CARE checklist, developed by an international group of experts to enhance transparency, accuracy and precision in the use of case reports, would be employed [32].

Table 2.

Methodological quality assessment by the Newcastle-Ottawa Scale (NOS).

Assessment instrument scoring
Newcastle-Ottawa Scale (NOS)
Study	Design	Selection	Comparability	Outcome	Total
De Santo et al. [26]	Pilot prospective research	4	2	2	8/9
Ampatzidou et al. [29]	Retrospective cohort	4	2	2	8/9
García-Delgado et al. [27]	Retrospective observational study	4	2	2	8/9
Daniel et al. [28]	Retrospective cohort	4	2	2	8/9

The NOS ranges from 0 to 9, with 9 being the number that expresses the highest quality on this semi-quantitative scale [31].

Results

A total of 349 articles were identified in the searched databases (PubMed = 77, Embase = 142, Lilacs = 69, PEDro = 61), of which 65 duplicates were removed. The remaining 284 studies were screened, with 252 considered irrelevant after title and abstract reading. After reading the full text of the remaining 32 articles, four were deemed eligible and included in this review (Figure 1).

Figure 1.

Study flowchart.

Of the four studies included in this review, three were retrospective studies and one was a prospective safety pilot study. The total sample size in each study varied from 109 to 1657 subjects, with a total of 3456 participants, of whom 283 realized NIPPV. Curative NIPPV was the only form of NIPPV in 75% of the studies, which presented this form of prescription in 5–9% of the total sample size, with men around 65 years old being the majority of the participants receiving curative NIPPV. ARF is the main indication for curative NIPPV. Only one study realized prophylactic NIPPV (28% of 32 participants). Regarding the predictors for NIPPV prescription, the elevated body mass index (BMI), hypercapnia, PO lung injury, cardiogenic oedema and pneumonia were identified in half of the studies included in the present review (Box 1).

Box 1. Predictors for NIPPV prescription.

BMI: body mass index; COPD: chronic obstructive pulmonary disease; CPB: cardiopulmonary bypass.

Low glomerular filtration rate ≤60 mL.min⁻¹.1.73 m²⁽⁻¹⁾. Green marks indicate the study in which the predictor was identified.

Regarding the intervention protocols, one study did not report the initial ventilatory support, one study described the use of continuously positive airway pressure (CPAP) by a flow generator, and two studies used bi-level NIPPV. The initial positive end-expiratory pressure (PEEP) was set from 4 to 7.5 cmH₂O, and inspiratory positive airway pressure (IPAP) from 10 to 14 cmH₂O. Two articles described a 2 h deadline for the initial monitoring analysis and gradual adjustments in PEEP or IPAP with oxygenation targets estimated by arterial blood gas analysis (PaO₂ >60 mmHg) or peripheral oxygen saturation (>92%). The total duration of NIPPV prescription was not reported in one study, while others reported its use from 2 h, 3×/day, until resolution of the clinical condition (in the only study prescribing prophylactic NIPPV), to 33 h. The success rate of curative NIPPV ranged from 55% to 94%, with better percentiles obtained when the indication of NIPPV was post-extubation ARF.

In 2009, De Santo et al. [26] showed that of 465 patients, only 9.2% (43 of 465, male: 32; female: 11) had indications for curative NIPPV due to ARF after extubation. Possible predictors for indications were post-cardiopulmonary bypass (post-CPB) lung injury (n = 21 of 43; 48.8%), cardiogenic oedema (n = 13 of 43; 30.2%) and pneumonia (n = 9 of 43; 21%). Pneumonia was identified as a predictor of NIPPV failure in nine patients of 43, with 55% of nine patients being reintubated, while lung injury after CPB in 21 (90.5% of 21) and cardiogenic dysfunction in 13 (69.2% of 13) were predictors of NIPPV success. Adjustments of initial ventilatory support occurred in CPAP by flow generator using a helmet and the values were in EPAP: 7.5 cmH₂O, which could be reduced to 5 or increased to 10 cmH₂O, targeting a PaO₂ >70 mmHg. The duration of therapy was 33.8 ± 24.0 h (Table 1).

García-Delgado et al. [27] observed that from 1225 individuals, only 5.1% (63 patients, male: 31; female: 32) required NIPPV and that the predictors for this indication were lobar atelectasis, acute cardiogenic oedema, acute lung injury/acute respiratory distress syndrome, hypercapnia (PaCO₂ >45 mmHg), pneumonia, others (pneumothorax, shock), BMI >30 kg.m⁻², CPB time (NIPPV group: 138 ± 68 min). The study reported that the initial ventilatory adjustments were PEEP 4–6 cmH₂O and IPAP 10–14 cmH₂O, and that among patients with persistent hypoxemic ARF (SaO₂< 90%), PEEP increased in steps of 2 cmH₂O, whereas in hypercapnic ARF (PaCO₂ >45 mmHg), increments of 2 cmH₂O in IPAP were necessary to adjust TV and RR. For oxygen supplementation in NIPPV, the minimum FIO₂ required to maintain an SaO₂ >92% was considered. The duration of the therapy was not reported by the authors. Forty-eight percent of patients undergoing NIPPV therapy failed, and a possible predictor of failure was a spontaneous breathing trial with a duration (from extubation to NIPPV application) of <24 h, which increased the failure rate by 4.6 times – odds ratio, 4.6; 95% CI, 1.2–17.9; p = .02). Meanwhile, therapy was successful in 55% (n = 63) of patients who had obesity as a predictor of NIPPV success (BMI >30 kg.m⁻²: odds ratio, 0.22; 95% confidence interval, 0.05–0.91; p = .03) (Table 1).

Daniel et al. [28] showed that out of 109 individuals (male: 66; female: 43), only 32 (29% of 109) needed NIPPV in the PO of cardiac surgery, with 72% of 32 being curative and only 28% of 32 prophylactic NIPPV. This study did not report initial ventilatory support or any other ventilatory parameters, and the clinical adjustments of these parameters were completely dependent on the individual’s assessment and the expertise of the health professional. The duration of therapy was carried out three times a day for 2 h until the resolution of the condition. NIPPV failed in only two (6%) individuals, one with hypoxemia and the other with hypercapnia. In addition, hypercapnia is a predictor of successful NIPPV therapy. The study did not report any variables associated with the predictors of failure. Hypercapnia, ARF, bronchospasm and hypoxemia are the most common predictors of curative NIPPV. For preventive NIPPV, radiographic abnormalities (atelectasis and pleural effusion) and post-extubation findings were indicated (Table 1).

A study published by Ampatzidou et al. [29] observed that only 8.8% of 1657 patients required curative NIPPV related to ARF, and the predictors for NIPPV indication were higher BMI (kg/m²), EuroSCORE II (points), COPD and low estimated glomerular filtration rate (cutoff: 60 mL.min⁻¹.1.73 m⁻²). The reported initial ventilatory support was PEEP: 4–5 cmH₂O, IPAP: 10–12 cmH₂O, and ventilatory adjustments were gradually increased in both IPAP and PEEP during the first 2 h of application, aiming at a PaO₂ of 60 mmHg without acidosis. The duration of therapy was 16 h, in a 24 h period. NIPPV failed in 16 individuals (11% of 145) and the possible failure predictors were: age (>75.5 years), EuroSCORE II: (>2.6), BMI (>26.4 kg.m⁻²) (5.9), higher burden of comorbidities as chronic atrial fibrillation (37.5%), COPD (50%), heart failure (with NYHA ≥2; 81.3%) and more frequent PO complications like myocardial infarction (6.3%), PO stroke (18.8%) and renal replacement therapy (50%). One hundred and twenty-nine subjects obtained the following characteristics as predictors of success of NIPPV: age (69 years), BMI (31.2 kg.m⁻²) (7.5), less comorbidities occurrence, as chronic atrial fibrillation (13.2%), COPD (18.6%), previous cardiac surgery (0.8%), heart failure (NYHA ≥2; 55%), PO stroke (2.3%) and renal replacement therapy (0.8%) (Table 1).

Discussion

The main finding of the present study was that elevated BMI, hypercapnia, PO lung injury, cardiogenic oedema and pneumonia can be considered as the main predictors of NIPPV prescription in the PO period of cardiac surgery. In terms of the NIPPV prescription method, large heterogeneity across articles was observed, since one study [28] did not mention the mode of NIPPV used, two other studies [27,29] used bi-level pressures, and one study [26] used the CPAP mode. In addition, there was no standardization of the duration of NIPPV or ventilatory adjustments used, with those variables being totally individualized and totally dependent on the patient’s response to the proposed therapy.

Findings regarding the length of hospital stay in the ICU and mortality were poorly described in the included articles. Only De Santo et al. [26] mentioned a shorter length of stay in the intensive care unit and in the hospital for those patients who underwent NIPPV. Regarding hospital mortality, De Santo et al. reported a lower rate for patients who realized NIPPV compared to the control group. The causes of death in the NIPPV group were as follows: three cases of low cardiac output, two cases of multiple organ failure and one case of septic shock. The controls presented with the following causes of death: low cardiac output in 16 cases, multiple organ failure in five cases, and septic shock in seven cases. In this sense, the study by García-Delgado et al. [27] mentioned that the hospital mortality rate was higher for patients who failed NIPPV than for those in whom NIPPV was successful.

In addition, one study [28] did not describe the initial parameters of NIPPV. However, because the main objective was to investigate the relationship between preoperative respiratory muscle strength and the indication and performance of NIV in elective cardiac surgery, the author only stated that the institutional protocol for NIPPV prescription and implementation was used, which we considered a relevant bias.

Despite the heterogeneity of our results, the literature [33] recommends that NIPPV must be used in three ways: curative, facilitative or preventive; the latter is used for prophylaxis of ARF. Thus, both the Brazilian Recommendations on Mechanical Ventilation [33] and the Brazilian Guidelines on Mechanical Ventilation [10] highlight some risk factors for the application of preventive NIPPV, namely hypercapnia after extubation (>45 mmHg), heart failure, ineffective cough, copious secretions, more than one consecutive weaning failure, more than one comorbidity, upper airway obstruction, age >65 years, heart failure as a cause of intubation, APACHE >12 on the day of extubation, and >72 h of IMV. However, at bedside, these risk factors can be found in almost, if not all, patients in the ICU after cardiac surgery, so NIPPV would be prescribed for those individuals, perhaps without the real need. Interestingly, some of the risk factors for indicating curative NIPPV mentioned in the above-mentioned recommendations are similar to those observed by Daniel et al. [28] , including hypercapnia, bronchospasm, ARF and hypoxemia, emphasizing the higher success rate of NIPPV in patients with hypercapnia. In fact, NIPPV is the preferred form of ventilatory support for patients with acute hypercapnic respiratory failure, as it improves dyspnoea and gas exchange, avoids the need for intubation, reduces morbidity and mortality rates, and may be beneficial in patients with obesity-related hypoventilation [34].

Liu et al. [35] observed that BMI ≥25 kg.m⁻² was a predictor of NIPPV success and attributed this to the fact that overweight patients exhibited hypercapnia (PaCO₂ > 50 mmHg) more often than eutrophic patients. Thus, obese patients are at a higher risk of atelectasis due to hypoventilation, making NIPPV a useful tool in this population. This similarity can also be observed in the study by García-Delgado et al. [27], in which a small number of individuals (5.1%) were subjected to NIPPV due to ARF, with BMI >30 kg.m⁻² being the main predictor responsible for 55% success in NIPPV. These findings are possibly attributable to what the literature calls obesity hypoventilation syndrome (OHS), which makes this a high-risk population for the development of post-extubation ARF [36,37].

In the study by Ampatzidou et al. [29], factors such as higher BMI, higher EuroSCORE, presence of COPD, and lower preoperative estimated glomerular filtration rate were independently associated with the application of NIPPV and were considered possible predictors for therapy prescription. In contrast, declining age, higher EuroSCORE, presence of COPD, presence of heart failure, presence of atrial fibrillation, PO renal replacement therapy and PO stroke were associated with NIPPV failure. These findings were similar to those reported by Liu et al. [35], who reported that the higher the severity of the disease (assessed by the Sequential Organ Failure Assessment (SOFA) score), the higher the chance of NIPPV failure, even without a difference in the risk of death after cardiac surgery (assessed by EuroSCORE). However, both studies reported elevated BMI as a predictor of NIPPV success, and furthermore, in the study by Hill et al. [34], the researchers reported that NIPPV may still be useful in patients with obesity-related hypoventilation, reinforcing the importance of a high BMI being considered as a possible predictor for the indication of NIPPV.

Apostolakis et al. [38] stated that CPB performed in cardiac surgery with median sternotomy may lead to pulmonary complications, such as hypoxemia, atelectasis, deterioration of respiratory mechanics, increase in respiratory secretions, bronchospasm, increased airway and thoracic resistance, and infections due to the release of pro-inflammatory cytokines, leading to worsening gas exchange and temporary pulmonary dysfunction. Similarly, Gray et al. [39] observed that NIPPV was associated with more significant reductions in dyspnoea, HR, acidosis and hypercapnia in the treatment of acute cardiogenic pulmonary oedema. He et al. [40] found that prolonged CPB time could increase the risk of ventilator-associated pneumonia (VAP) after cardiac surgery. Moreover, severe functional class, pulmonary hypertension, COPD, peripheral vascular disease, renal disease, emergency surgery, need for intra-aortic balloon counterpulsation (IABC), aortic cross-clamp time, mechanical ventilation time, reoperation and reintubation were also considered proven risk factors for VAP. These findings corroborate those of De Santo et al. [26], who described those patients requiring NIPPV presented with a systemic inflammatory reaction followed by post-CPB lung injury, cardiogenic oedema and pneumonia as possible predictive causes; all clinical conditions that may predispose to hypercapnia, another predictor for NIPPV prescription found in our study. Furthermore, pneumonia is a determinant of NIPPV failure in individuals with COPD [41], and it is worth noting that pneumonia is the main cause of NIPPV failure and reintubation. As previously reported, pneumonia is an independent predictive factor for NIPPV failure, and the majority of individuals in whom pneumonia occurred before NIPPV needed to return to invasive ventilatory support [35].

Conclusions

There is little evidence specifically describing the predictors for the application of NIV in the PO of cardiac surgery. The different intervention protocols and diversity of professionals’ recommendations for prescribing NIPPV make comparisons between studies difficult. Nevertheless, most studies have presented the main conditions that can be observed as possible predictors for the indication of curative NIPPV: elevated BMI, hypercapnia, PO lung injury, cardiogenic oedema and pneumonia. Despite the need for more robust studies, mainly multicentre clinical trials, to understand the role of these conditions as real predictors of the need for NIPPV in the PO of cardiac surgery, it seems that there is no need to perform NIPPV for all patients in the PO period of cardiac surgery, but only for those presenting the aforementioned predictors.

Funding Statement

This work was supported by Fundação Pró-Coração (FUNDACOR).

Author contributions

Jéssica Gonçalves de Lima participated in the design and planning of the work, as well as in the interpretation of evidence, and drafting and/or revision of preliminary and definitive versions. Victoria Maria Garcia de Medeiros participated in the design and planning of the work, as well as in the interpretation of evidence and drafting and/or revision of preliminary and definitive versions. Fernando Gomes de Jesus participated in the design and planning of the study, as well as in the interpretation of evidence. Thaísa Sarmento dos Santos participated in the design and planning of the study, as well as in the interpretation of evidence. Juliana Rega de Oliveira participated in the design and planning of the study, as well as in the interpretation of evidence. Drafting and/or revision of preliminary and definitive versions. Claudia Rosa de Oliveira participated in the design and planning of the study, as well as in the interpretation of evidence. Mauro Felippe Felix Mediano: drafting and/or revision of preliminary and definitive versions. Luiz Fernando Rodrigues Junior participated in the design and planning of the study, as well as in the interpretation of evidence, review of preliminary versions, and approval of the final version.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are openly available in Open Science Framework at https://osf.io/fb9xw, reference number fb9xw.