Fig. 1.
Fig. 2.
Fig. 3.
Fig. 4.
Comparison of Demographic, Severity of illness, Inflammatory markers between NIV success vs Failure patients
| Total | NIV Success (n=179) | NIV failure (n=262) | p value | |
|---|---|---|---|---|
| Age | 60 (19–90) | 54(53–58) | 62(61–64) | <0.001 |
| M:F | 311:130 | 126:53 | 185:77 | 0.960 |
| DM | 241(54.6%) | 106(59.2%) | 135(56%) | 0.111 |
| HTN | 268(60.8%) | 103(57.5%) | 165(63%) | 0.251 |
| CAD | 64 (14.5%) | 22(12.3%) | 42(16%) | 0.274 |
| CKD | 24(5.4%) | 7(3.9%) | 17(6.5%) | 0.241 |
| COAD | 22(5%) | 8(4.5%) | 14(5.3%) | 0.629 |
| Symptom-admission duration (days) | 7 (0–14) | 3(0–8) | 4(1–14) | 0.783 |
| Baseline P/F ratio | 125(56–280) | 145(140–156) | 117(110–121) | <0.001 |
| ARDS category | <0.001 | |||
| Mild | 32 (7.3%) | 18(56.3%) | 14(43.8%) | |
| Moderate | 282 (63.4%) | 136(48.2%) | 146(51.8%) | |
| Severe | 127(28.8%) | 25(19.7%) | 80.3% | |
| Remedesivir | 286 (64.9%) | 117(65.4%) | 169(64.5%) | 0.853 |
| TCZ | 23 (5.2%) | 11(6.1%) | 12(4.6%) | 0.468 |
| Vasopressors | 172(39%) | 2(1.1%) | 170(64.9%) | <0.001 |
| Steroid | <0.001 | |||
| Early High | 283 (64.2%) | 126(70.4%) | 157(59.9%) | |
| Early low | 113 (25.6%) | 50(27.91%) | 63(14.3%) | |
| Indeterminate | 45 (10.2%) | 3 (1.7%) | 42(9.5%) | |
| BSI | ||||
| NG | 351(79.6%) | 164(91.6%) | 187(71.4%) | <0.001 |
| Gram neg | 56(12.7%) | 9(5%) | 47(17.9%) | |
| Positive | 14 (3.2%) | 3(1.7%) | 11(4.2%) | |
| Fungal | 20(4.5%) | 3(1.7%) | 17(6.5%) | |
| Shock | 134 (30.4%) | 1(0.6%) | 133(50.8%) | <0.001 |
| AKI | 109(24.7%0 | 8(4.5%) | 101(38.5%) | <0.001 |
| SOFA | 4(2–14) | 3(3–4) | 6(6–7) | 0.0001 |
| Baseline CRP | 101(1.98–870) | 75(68–98.70) | 116.20(98–135) | 0.006 |
| Baseline ALC | 760(120–3640) | 860(760–920) | 720(650–780) | <0.001 |
| ICU LOS | 9(1–90) | 7(7–8) | 11(10–13) | <0.001 |
| Hospital LOS | 14(1–95) | 14(12–16) | 14(13–15) | 0.349 |
Ventilatory parameter comparison- HACOR/ROX
| Total | Success (n=175) | Failure (n=262) | Univariate analysis OR (CI) | |
|---|---|---|---|---|
| HACOR D1 | 5(0–9) | 4(4–5) | 6(6–7) | 1.963(1.696–2.273) |
| ROX D1 | 4.13(1.9–9.27) | 5.14(4.9–5.3) | 3.53(3.33–3.68) | 0.309(0.245–0.389) |
| HACOR D2 | 5(0–10) | 3(3–4) | 6(6–7) | 2.735(2.250–3.325) |
| ROX D2 | 4.1(1.7–12.15) | 5.62(5.36–5.85) | 3.32(3.16–3.50) | 0.194(0.143–0.262) |
| HACOR D3 | 5(0–9) | 3(3–4) | 7(7–8) | 3.064(2.435–3.857) |
| ROX D3 | 4.5(1.68–13.22) | 6.06(5.85–6.27) | 3.3(3.13–3.50) | 0.186(0.131–0.263) |
| HACOR D4 | 5(0–10) | 3(3–4) | 7(7–8) | 3.119(2.416–4.027) |
| ROX D4 | 2.7(2–14.1) | 6.08(5.83–6.50) | 3.29(3.10–3.57) | 0.182(0.125–0.266) |
| HACOR D5 | 4.0(0–10) | 2(2–3) | 7(7–8) | 3.745(2.647–5.300) |
| ROX D5 | 5.2(1.95–10.47) | 6.3(6.15–6.6) | 2.95(2.8–3.20) | 0.182(0.119–0.279) |
| HACOR D6 | 4 (0–10) | 2.0(2–3) | 7.5(7–8) | 3.536(2.442–5.119) |
| ROX D6 | 5.4(1.5–12.65) | 6.84(6.35–7.14) | 2.89(2.66–3.20) | 0.159(0.093–0.272) |
| No of days with HACOR>5 | 2(0–18) | 1(1–2) | 3(3–4) | 1.87(1.58–2.09) |
Summary of studies depicting NIV failure
| Reference, Year (Country) | Setting | Modes | NIRS failure definition | HACOR score in Failure | ROX Score in Failure | NIRS Failure | Mortality | Predictors of Failure |
|---|---|---|---|---|---|---|---|---|
| Duan et al. 2021 (China)19 | ICU (non-COPD) | NIV | Need of IMV as rescue therapy | 7 (at initiation) 5 (at 1–2 hours) | 14.7 % | 9.4 % | HACOR score, Duration of NIV, Heart rate, pH, GCS, Mean Arterial BP, P/F ratio, PaCO2 | |
| Guia et al. 2021 (Italy, Portugal, USA)15 | ICU (COVID) | NIV (CPAP) | Need of IMV as rescue therapy and Death | 5 | 27.3 % | 23 % | HACOR score, Age, P/F ratio | |
| Valencia et al. 2021 (Colombia)13 | Emergency Department | HFNC (COVID) | Need of IMV as rescue therapy and Death | 7.14 ±3.6 | 5.61 ±4.1 | 62.0 % | 29.3 % | HACOR score, GCS, ROX index, Heart Rate, Respiratory Rate, PaCO2, SaO2, pH, PaO2, P/F ratio, Chronic Kidney Disease, Co-Infections |
| Jog S et al. 2021 (India)20 | ICU (COVID) | NIV/HFNC | Need of IMV as rescue therapy and Death | 64.1 % | 56.9 % | Age, Medical Comorbidities, Admission SpO2, Non-Respiratory organ dysfunction | ||
| Prakash J et al. 2021 (Meta-analysis)16 | ICU (COVID) | HFNC | 5 (24 hr) | ROX score | ||||
| Prakash J et al. 2021 (Meta-analysis)16 | ICU (COVID) | HFNC | 5 (24 hr) | ROX score | ||||
| Khan MS et al., 2022 (India)21 | ICU (COVID) | HFNC | Need of NIV or IMV as rescue therapy | 3.4 ± 0.4 (1 hr) | 28.9 % | 27.1 % (28-day) | Age, APACHE ll, SOFA, Lag time to HFNC, Duration of HFNC, ICU LOS, PF ratio, ROX index, D-dimer, IL-6, RBS, Admission SpO2 | |
| Santus et al. 2022 (Italy)14 | HDU (COVID) | CPAP (Helmet) | Need of IMV as rescue therapy, Transfer to ICU and Death | 5 (1 hr) | 6.2 (4.7–7.7) (1hr) | 38.4 % | 12.4 % | Age, Ischaemic Heart Disease, COPD, CPAP duration, Hospital LOS, WBC count, D-dimer, CRP, PEEP, FiO2, PF ratio, Respiratory rate, A-a O2 gradient, GCS, ROX index, mROX index, HACOR score PaCO2, |
| Chacko et al. 2022 (India)18 | ICU (COVID) | NIV | Need of IMV as rescue therapy or Death | 36.4 % | 30.1 % | Age, Disease Severity, Admission PF ratio, Respiratory Rate, High CK-MB, Need for organ support, Duration of continuous NIV, ICU LOS and hospital LOS |
Multivariable models for NIV failure_ The models were created on 327 participants having complete information about all covariates_ The model diagnostics indicate the substantial reduction in error variances after theory driven sequential sections of variables
| Predictors | Ref. Category. NIV_Success Odds Ratios CI p | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | CI | p | OR | CI | p | OR | Cl | p | OR | CI | p | OR | CI | p | |
| R0X>4.47 | 0.19 | 0.13–26 | <0.001 | 0,18 | 0.12–0.25 | <0.001 | 0,18 | 0.12–0.25 | <0.001 | 0,17 | 0.11–0.25 | <0.001 | 0,15 | 0.09–0.23 | <0.001 |
| Age>60Yrs | 5, | 2.47–11.83 | <0.001 | 5, | 238–11.74 | <0.001 | 5, | 230–1158 | <0.001 | 4, | 1.66–10.18 | 0,003 | |||
| GenderF | 0,64 | 0.29–1.41 | 0,271 | 59 | 0.26–1.33 | 0,208 | 0,65 | 0.28–1.47 | 0,300 | 37 | 0.13–1.00 | 57 | |||
| Diabetes+ | 0,61 | 0.29-1.27 | 0,184 | 59 | 0.28–1.25 | 0,170 | 59 | 0.27–1.26 | 0,176 | 55 | 023–1.30 | 0,175 | |||
| Illness_ICU_IntervalT2(Int) | 0,72 | 0.29–1.72 | 0,456 | 0,76 | 031–1.86 | 355 | 0,77 | 0.27–2.15 | 0,617 | ||||||
| Illness_ICU_IntervalT3(High) | 2, | 0.87–5.25 | 0,100 | 222 | 0.90–5.62 | 0,086 | 3, | 1.16–9.26 | 0,027 | ||||||
| ALCDI_Tertile2(Int) | 0,45 | 0.17–1.11 | 87 | 0,87 | 0.29–258 | 0,796 | |||||||||
| ALCDI_Tertile3(High) | 0,43 | 0.16–1.07 | 0,071 | 0,75 | 0.26–2.16 | 585 | |||||||||
| SOFA Score | 2, | 1.60–2.71 | <0.001 | ||||||||||||
| Observations | 327 | 327 | 327 | 327 | 327 | ||||||||||
| R2 Tjur | 393 | 0,636 | 0,646 | 0,654 | 0,729 | ||||||||||
| AIC | 224, | 209, | 207, | 206582 | 169, | ||||||||||
| log-Likelihood | −110213 | −99334 | −96387 | −94291 | −74, | ||||||||||