Identification of factors associated with early and late NIV success or failure are important in hospital settings and help in timely decisions and prompt shifting of patients to ventilation facilities. In our study severe acidosis at initiation increased the risk of delay in NIV success, however no increased risk of failure was seen at 24-hours. This is consistent with other studies (18, 19, 23–25), although as compared to these studies our patient population included more diverse patients including obstructive, restrictive, and infective disorders and based in ICU rather than emergency department. Official ERS/ATS guidelines 2017 show that there is no lower limit of pH appropriate for NIV trial (26). In our study if the pH levels after one-hour of NIV were ≥ 7.3, chances of success were increased 5 folds at 24-hours, making one-hour pH a stronger predictor of overall NIV success which is also seen in previous studies (27).
Higher initial PCO2did not show the same correlations with either timing of success or overall NIV failure, however, PCO2 > 75mmHg at one-hour was associated with NIV failure at 24-hours.Previous studies show conflicting data on the association of levels of pCO2 and NIV outcomes. One study showed that high PCO2 didn’t predict NIV failure(23), while another study, in agreement with our study, showed PCO2 > 77mmHg at 1–2 hours as a critical predictor of failure(25).In our study initial pCO2 levels were positively correlated with the bicarbonate levels, implying compensation of respiratory acidosis. Thus, high PCO2 with metabolic compensation is not a predictor of NIV outcome at one-hour. Patients with low bicarbonate levels or concomitant metabolic acidosis failed NIV trial at one-hour. However, persistently high PCO2 after 1-hour is a predictor of NIV failure.
Another important predictor of NIV success was lesser IPAP given to wash out PCO2. Relatively lower IPAP < 20 cm H2O was associated with increased risk of NIV failure at 24-hours. Lesser mortality is shown to be associated with initiating treatment with high pressure regimen(27). Our findings in line with BTS guidelines2016, support the use of high IPAP of 20cm of H2O(28).
Initial PO2 levels ≥ 65 mmHg had lesser chances of success at one and24-hours of NIV. Higher FiO2 was associated with failure at 24-hours. This is most likely due to loss of hypoxia drive because of oxygen therapy, causing worsening of respiratory failure, accumulation of PCO2 and worsening of acidosis. Patients enrolled in INSPIRED COPD Outreach Program also showed high prevalence of hypercapnia in over-oxygenated patients(29). Therefore controlled oxygen therapy is a well-known concept for quite a long time(29).
Our study showed the association of mortality with high initial PO2, higher FiO2 requirements. Higher FiO2 requirement represent higher degree of hypoxia and disease severity which is likely to affect mortality as shown in other studies as well (30–32).High WBC counts were associated with mortality in our study. In pneumonia high WBC counts is a well known predictor of mortality (33),.implying the comorbidity with infection in obstructive and restrictive disorders was the contributing towards mortality.