This was a hospital-based observational and analytical study that was undertaken in the Respiratory Medicine Department at SRM Medical Hospital and Research Centre, Jaipur, Rajasthan, India. The study duration was one and a half years (February 2019- August 2020) from the date of obtaining ethical clearance (1636A/IEC/2019). A total of 42 patients with respiratory acidosis secondary to COPD (pH <7.35 and PaCO₂ >45 mmHg) attending respiratory medicine Outpatient Department (OPD).

Sample size: Mean PaCO₂ after treatment in two groups was 53.2±9, 45.4±7.9 based on a previous study [7] with 95% confidence and 80% power. The sample size was 42; 21 members in each group.

Inclusion and Exclusion criteria: COPD patients with respiratory acidosis (pH <7.35 mm Hg) and type 2 respiratory failure (PaCO₂ >45 mmHg) were included. Low Glasgow Coma Scale (GCS), Haemodynamic instability, facial deformity, obstruction of the upper airway from recent trauma/surgery, pneumothorax, cardiogenic pulmonary oedema, cardiopulmonary arrest, patients with Arterial Blood Gas (ABG) showing pH <7.10, haemoptysis or haematemesis, other causes of type 2 respiratory failure except for COPD were excluded.

Demographic data, history including co-morbidities, relevant investigations, modes of NIV setting AVAPS or S/T, sputum organism growth and clinical outcomes of all the patients were recorded in a structured pro forma. Data of patient’s ABG parameter, heart rate, blood pressure, and respiratory rate baseline values were noted and at an interval of one hour, six hours, and every 24 hours. The baseline parameters between two modes of NIV ventilation (AVAPS and S/T mode) among the NIV failure category were also compared. The duration and severity of COPD illness were not considered. There were two study endpoints:

Parameters were tabulated and analysed by SPSS software version 22.0. The quantitative variables were analysed using Student’s t-test and qualitative variables were analysed using the Chi-square test. A p-value of less than 0.05 was considered statistically significant.

Of 42 patients, there were 27 male and 15 female with a mean age of 56.83 years. Among them, 30 (71.42%) patients improved with NIV and 12 (28.58%) patients were considered NIV failure as per the criteria specified in the methodology [Table/Fig-1].

Among 42 patients, the mean age of patients in the success category was 55.13±5.823, while in the failure category, it was 61.08±5.452. It was observed that patients with age >60 years had a poor outcome in NIV management of acute respiratory failure secondary to COPD. On comparing the mean BMI between success (19.62±3.22) and failure category (26.65±5.649), it was noted that patients who were overweight (BMI >25) were predominantly present in failure category than in success category.

In the success category, the mean baseline pH was 7.26±0.034 while in failure category it was 7.2±0.049 which was statistically significant. It is evident that patients who present with initial pH >7.25 have a better outcome without any need for invasive ventilation. Out of 42 patients, 35 patients (83.33%) showed improvement in pH after one hour of initiating NIV treatment. At the same time, 7 (16.66%) had deterioration within one hour itself. Within the 35 patients who showed an initial improvement, five patients had decline in pH when ABG was repeated after six hours interval and clinically these patients deteriorated. So, these patients were also considered failure and were intubated.

Overall, 30 patients had successful outcome with respect to PaCO₂. The mean baseline PaCO₂ in success category was 69.98±7.102 while in failure category it was 78±5.461 [Table/Fig-1]. Out of 42 patients, 34 patients (80.95%) showed improvement in PaCO₂ after one hour of initiating NIV treatment. In contrast, 8 (19.04%) had deterioration within one hour itself. Within the 34 patients who showed an initial improvement, four patients had a decline in PaCO₂ when ABG was repeated after six hours interval. So, these patients were also declared as NIV failure and were intubated. Therefore, a patient who presents with initial PaCO₂ <70 and patients who show improvement within one hour of initiating NIV appears to a good predictor was successful NIV outcome.

At the time of admission, all patients presented with tachycardia and tachypnoea. The mean baseline HR was 116.92 while the RR was 36.04. The success category patients had significantly lesser HR (115.33±5.033) and RR (34.33±3.089) than the failure category (HR=120.92±8.393, RR=40.33±3.367), which was statistically significant with p-value <0.01 [Table/Fig-1]. So patients who present with tachycardia and tachypnoea on the lower limit (HR <120 and RR <35) have a better outcome.

Co-existing co-morbidities and infective exacerbation (in sputum) also play a significant role in predicting the outcome of NIV success. Out of 42 patients, 18 (42.85%) had co-morbidities, among which 9 (21.42%) were in failure category. With respect to sputum examination, 10 patients (23.8%) had growth, among them 6 (14.2%) were in failure category [Table/Fig-1].

Overall, 23 (54.76%) patients received AVAPS-S/T mode, and 19 (45.23%) patients received S/T mode of NIV treatment. Among the failure category with 12 patients, it was observed that only 4 (33.33%) were on AVAPS mode and 8 (66.67%) were in S/T mode [Table/Fig-2].

Also, in failure group in was observed that the mean baseline of pH, PaCO₂, HR and RR was comparable between AVAPS and S/T group. This implies that the AVAPS and S/T group has the same baseline value, but still, the incidence of failure in S/T group was higher.

In this study, 23 patients received AVAPS mode of ventilation in which 4 (9.52%) patients belonged to NIV failure category while remaining 19 patients received S/T mode in which 8 (19.04%) patients were NIV failure [Table/Fig-2]. Although the baseline values in the AVAPS group and S/T group were comparable and statistically insignificant, the incidence of NIV failure was significantly higher in the S/T mode when compared to AVAPS mode. Hence, it can be concluded that the selection of mode ventilation also plays an essential role in the outcome of NIV success rate among COPD patients.

In this study, the possible early predictors and associated factors influencing the outcome of NIV in acute respiratory failure due to COPD were assessed. Previous studies have provided some predictors for successful outcome of NIV which includes lower age group, lower APACHE II score, patient-ventilator synchrony, better GCS, hypercarbia (45<PaCO₂>92 mmHg), acidosis (7.35<pH>7.10), oxygenation and heart and respiratory rates within the first 1-2 hour after the initiation of NIV [7,8].

From the present study, it can be concluded that patients with age >60 years had a poor outcome in NIV management of acute respiratory failure secondary to COPD. This was similar to a study conducted by Van Gemert A et al., the authors stated that older age and low GCS are an essential factor for conversion of NIV to IMV in AHRF secondary COPD [9,10]. In another study by Kida Y et al., NIV failure rate was higher in elderly patients with acute hypercapnic respiratory failure [11].

In the success category, the mean baseline pH was 7.26±0.034, so it can be concluded that patients who present with initial pH >7.25 have a better outcome without any need for invasive ventilation. Fewer studies have shown that the improvement in pH within a short period after initiation of NIV predicts the outcome. A study conducted by Claude KH et al., proved that improvement in pH and reduction in PaCO₂ observed within 2 hours of initiation of NIV in COPD patients [12]. Lightowler JV et al., did a similar study showed that patients with baseline pH <7.22 went for NIV failure in COPD patients [13]. Another study was done by Liu J et al., also concluded that severe acidosis is a significant predictor for NIV failure in COPD patients [10]. Cavalleri M et al., in his study, has shown that patients with initial pH <7.23 are more prone to NIV failure [14].

The mean baseline PaCO₂ in success category was 69.98±7.102 while in failure category it was 78±5.461. Therefore, a patient who presents with initial PaCO₂ <70 appears to a good predictor of successful NIV outcome. Liu J et al., did a similar study showed that patients with baseline PaCO₂>88 went for NIV failure in COPD patients [10]. Cavalleri M et al., showed that patients with initial PaCO₂ >85 mmHg were liable for NIV failure [14].

The mean baseline HR was 116.92 while the RR was 36.04. The success category patients had significantly lesser HR and RR than the failure category, which was statistically significant. So, patients who present with tachycardia and tachypnoea (HR <120 and RR <35) have a better outcome. Liu J et al., also showed that patients with initial mean HR >121 and RR >30 have a poor outcome and went for NIV failure [10].

In this study, the percentage of co-morbidities was higher in NIV failure group. Moretti M et al., showed that incidence of comorbidities were more in NIV failure category than in success category [15]. But Talwar D and Dogra V concluded that the incidence of comorbidities was equal both in success and failure group [16]. Also, while weaning the COPD patients from mechanical ventilation both groups had equal outcomes.

It can be concluded that the selection of mode ventilation also plays an important role in the management of COPD patients in terms of reducing the incidence of NIV failure. Similar studies were done by Claudett KH et al., and Shaaban L et al., they concluded that Bilevel Positive Airway Pressure (BiPAP)-AVAPS mode is an effective and safer mode compared to BiPAP-S/T mode in non invasive ventilatory management of acute respiratory failure in COPD patients [12,17].

The fact that it was an observational study in which patients were not randomised. A randomised control trial can be conducted in future to assess the other outcomes. Also, in future the technology and modes of ventilation used in NIV may change and these predictors cannot be used to assess the outcome in COPD patients.

In this study, it was observed that patients with acute respiratory failure secondary to COPD who were treated with NIV had a better outcome with reduced incidence of NIV failure. Among the failure group, elder age (>60 years), obese patients, severe acidosis (pH <7.2), hypercarbia (PaCO₂ >78), HR >120/min and RR >35/min at the time of presentation predicts the NIV failure. And also we observed that NIV failure was higher in patients who received S/T mode than AVAPS-S/T mode of NIV.