Table E1. Univariate and Multivariate Logistic Analysis to Identify Factors Associated

Table E1. Univariate and multivariate logistic analysis to identify factors associated with endotracheal intubation and 28-day mortality from HFNC application in patients with HFNC treatment

Variables / Endotracheal intubation / 28-day mortality from HFNC application
Univariate analysis / Multivariate analysis / Univariate analysis / Multivariate analysis
Odds ratio (95% CI) / P valuea / Odds ratio (95% CI) / P valuea / Odds ratio (95% CI) / P valuea / Odds ratio (95% CI) / P valuea
Age, years / 1.009 (0.986–1.031) / 0.453 / 1.022 (0.998–1.046) / 0.067
Sex, female / 0.650 (0.315–1.340) / 0.243 / 1.174 (0.571–2.410) / 0.663
Body mass index, kg/m2 / 1.007 (0.933–1.086) / 0.861 / 0.933 (0.863–1.010) / 0.086
Smoker / 2.500 (1.271–4.918) / 0.008* / 1.869 (0.536–6.518) / 0.326 / 1.410 (0.727–2.735) / 0.310
Smoking, pack years / 1.020 (1.004–1.037) / 0.014* / 1.006 (0.977–1.036) / 0.679 / 1.005 (0.989–1.020) / 0.554
Underlying disease
Diabetes mellitus, yes / 1.520 (0.715–3.230) / 0.276 / 1.175 (0.559–2.472) / 0.671
Hypertension, yes / 1.711 (0.890–3.288) / 0.107 / 1.404 (0.734–2.686) / 0.306
Solid malignancies except lung, yes / 1.016 (0.430–2.400) / 0.972 / 1.448 (0.610–3.439) / 0.401
Hematologic diseases, yes / 1.833 (0.641–5.247) / 0.259 / 9.868 (2.169–44.899) / 0.003* / 9.233 (1.491–57.164) / 0.017*
Chronic kidney disease, yes / 2.000 (0.649–5.164) / 0.227 / 3.291 (0.997–10.859) / 0.051
Liver cirrhosis, yes / 0.524 (0.180–1.524) / 0.236 / 1.429 (0.503–4.061) / 0.503
Use of immunosuppressive agents, yes / 7.100 (0.851–59.218) / 0.070 / 3.409 (0.665–17.472) / 0.141
Coronary artery disease, yes / 1.833 (0.641–5.247) / 0.259 / 2.880 (0.961–8.635) / 0.059
Chronic heart failure, yes / 0.700 (0.246–1.990) / 0.503 / 1.909 (0.656–5.553) / 0.236
APACHE II score / 1.067 (1.018–1.118) / 0.006* / 1.032 (0.978–1.089) / 0.249 / 1.191 (1.115–1.273) / <0.001* / 1.157 (1.074–1.247) / <0.001*
HFNC initial FiO2, % / 1.031 (1.008–1.055) / 0.009* / 1.011 (0.983–1.039) / 0.459 / 1.042 (1.017–1.068) / 0.001* / 1.031 (0.995–1.068) / 0.090
HFNC initial flow, L/min / 1.034 (0.996–1.075) / 0.084 / 1.043 (1.004–1.085) / 0.032* / 0.979 (0.923–1.038) / 0.475
Duration of HFNC application time, hours / 0.988 (0.982–0.995) / <0.001* / 0.991 (0.984–0.997) / 0.007* / 0.992 (0.987–0.998) / 0.006* / 0.996 (0.990–1.002) / 0.188
Type of ICU, MICU / 1.819 (0.855–3.868) / 0.120 / 4.000 (1.729–9.253) / 0.001* / 1.962 (0.691–5.571) / 0.206
Systolic blood pressure before HFNC, mmHg / 0.992 (0.980–1.005) / 0.243 / 0.985 (0.972–0.998) / 0.026* / 0.985 (0.968–1.003) / 0.103
Diastolic blood pressure before HFNC, mmHg / 1.003 (0.983–1.023) / 0.804 / 0.991 (0.971–1.011) / 0.383
Pulse rate before HFNC, /min / 0.999 (0.985–1.014) / 0.922 / 1.003 (0.989–1.018) / 0.639
Respiration rate before HFNC, /min / 1.012 (0.967–1.060) / 0.598 / 1.051 (1.002–1.103) / 0.041* / 1.023 (0.963–1.086) / 0.467
ABGA pH before HFNC / 0.553 (0.015–20.075) / 0.746 / 1.565 (0.043–56.680) / 0.807
ABGA PO2 before HFNC, mmHg / 0.997 (0.988–1.006) / 0.538 / 0.993 (0.981–1.004) / 0.194
ABGA PCO2 before HFNC, mmHg / 0.958 (0.926–0.991) / 0.015* / 0.959 (0.911–1.008) / 0.101 / 0.971 (0.939–1.003) / 0.077
ABGA HCO3 before HFNC, mEq/L / 0.927 (0.876–0.982) / 0.009* / 1.014 (0.928–1.107) / 0.760 / 0.957 (0.908–1.010) / 0.109
ABGA SpO2 before HFNC, % / 1.000 (0.959–1.043) / 0.995 / 1.005 (0.964–1.049) / 0.805
ABGA FiO2 before HFNC / 11.808 (1.964–70.980) / 0.007* / 7.160 (0.694–73.904) / 0.098 / 8.722 (1.471–51.712) / 0.017* / 2.209 (0.116–42.077) / 0.598
ABGA PaO2/FiO2 before HFNC, mmHg / 0.996 (0.991–1.000) / 0.073 / 0.991 (0.985–0.998) / 0.006* / 0.997 (0.988–1.005) / 0.483
ABGA lactic acid before HFNC, mmol/L / 1.144 (0.967–1.354) / 0.117 / 1.062 (0.921–1.223) / 0.408

HFNC high flow nasal cannula, CI confidence interval, APACHE Acute Physiology and Chronic Health Evaluation, FiO2 fraction of inspired oxygen, L/min liter per minute, ICU intensive care unit, MICU medical intensive care unit, mmHg millimeter of mercury, min minute, ABGA arterial blood gas analysis, PO2partial pressure of oxygen, PCO2partial pressure of carbon dioxide, HCO3bicarbonate, mEq/L milliequivalents per liter, SpO2 oxygen saturation, PaO2/FiO2, ratio of partial pressure arterial oxygen and fraction of inspired oxygen, mmol/L millimoles per liter

a Statistical comparisons of the data were performed using logistic regression analysis (enter method).

Statistically significant p values are indicated by asterisks
Table E2. Baseline characteristics of de novo acute hypoxemic respiratory failure patients group and acute hypoxemic respiratory failure patients with underlying chronic lung disease group before and after propensity score matching

Characteristics / Before matching / P valuea / After matching / P valueb
De novo AHRF patients group(n =92) / AHRF patients with underlying CLD group (n = 57) / De novo AHRF patients(n =38) / AHRF patients with underlying CLD group (n = 38)
Age, yearsc / 61.62 ± 15.40 / 68.96 ± 12.11 / 0.003* / 71.00 (55.00–77.00) / 67.50 (60.00–73.25) / 0.839
Male sex, n (%) / 60 (65.2) / 48 (84.2) / 0.012* / 31 (81.6) / 30 (78.9) / >0.999
Body mass index, kg/m2c / 23.39 ± 4.33 / 21.51 ± 3.91 / 0.008* / 22.31 (19.74–24.71) / 22.03 (18.82–24.86) / 0.925
Smoker, n (%) / 49 (53.3) / 42 (73.7) / 0.013* / 29 (76.3) / 30 (78.9) / >0.999
Smoking, pack yearsc / 23.11 (20.86–26.04) / 21.65 (18.37–24.11) / 0.004* / 28.50 (11.25–42.50) / 30.00 (10.00–50.00) / 0.887
Underlying disease
Diabetes mellitus, n (%) / 26 (28.3) / 11 (19.3) / 0.218 / 10 (26.3) / 7 (18.4) / 0.581
Hypertension, n (%) / 43 (46.7) / 23 (40.4) / 0.445 / 21 (55.3) / 15 (39.5) / 0.286
Solid malignancies except lung, n (%) / 15 (16.3) / 10 (17.5) / 0.844 / 8 (21.1) / 7 (18.4) / >0.999
Hematologic diseases, n (%) / 9 (9.8) / 8 (14.0) / 0.427 / 3 (7.9) / 4 (10.5) / >0.999
Chronic kidney disease, n (%) / 7 (7.6) / 8 (14.0) / 0.205 / 2 (5.3) / 8 (21.1) / 0.070
Liver cirrhosis, n (%) / 9 (9.8) / 7 (12.3) / 0.632 / 5 (13.2) / 4 (10.5) / >0.999
Use of immunosuppressive agents, n (%) / 3 (3.3) / 5 (8.8) / 0.260 / 0 (0.0) / 5 (13.2) / 0.063
Coronary artery disease, n (%) / 8 (8.7) / 9 (15.8) / 0.186 / 3 (7.9) / 8 (21.1) / 0.125
Chronic heart failure, n (%) / 7 (7.6) / 9 (15.8) / 0.117 / 2 (5.3) / 6 (15.8) / 0.289
APACHE II scorec / 21.23 ± 8.32 / 22.04 ± 6.77 / 0.539 / 21.50 (17.00–27.25) / 23.00 (17.75–26.00) / 0.814
HFNC initial FiO2, %c / 50.00 (50.00–60.00) / 60.00 (50.00–60.00) / 0.736 / 60.00 (50.00–70.00) / 60.00 (50.00–62.50) / 0.548
HFNC initial flow, L/minc / 50.00 (40.00–50.00) / 45.00 (40.00–50.00) / 0.252 / 50.00 (40.00–52.50) / 47.50 (40.00–50.00) / 0.117
Duration of HFNC application time, hoursc / 34.25 (11.63–68.50) / 56.58 (12.00–112.75) / 0.212 / 25.25 (5.75–53.50) / 28.53 (6.75–93.54) / 0.500
Type of ICU, MICU, n (%) / 63 (68.5) / 49 (86.0) / 0.016* / 32 (84.2) / 31 (81.6) / >0.999
Systolic blood pressure before HFNC, mmHgc / 126.2 ± 23.16 / 125.3 ± 29.00 / 0.844 / 124.50 (109.75–139.75) / 118.00 (105.50–138.50) / 0.572
Diastolic blood pressure before HFNC, mmHgc / 74.03 ± 15.92 / 74.32 ± 16.35 / 0.917 / 77.00 (64.50–84.25) / 75.50 (60.00–86.50) / 0.845
Pulse rate before HFNC, /minc / 105.40 ± 22.37 / 106.98 ± 23.17 / 0.680 / 107.00 (87.25–119.25) / 103.00 (85.75–117.50) / 0.786
Respiration rate before HFNC, /minc / 25.00 (22.00–32.00) / 24.00 (20.00–28.00) / 0.119 / 26.50 (22.00–32.25) / 25.00 (21.00–28.50) / 0.147
ABGA pH before HFNCc / 7.44 (7.38–7.50) / 7.42 (7.37–7.47) / 0.250 / 7.44 (7.38–7.48) / 7.43 (7.38–7.48) / 0.805
ABGA PO2 before HFNC, mmHgc / 61.00 (52.00–81.75) / 63.00 (51.00–79.00) / 0.850 / 54.00 (51.80–76.00) / 61.20 (50.75–71.55) / 0.809
ABGA PCO2 before HFNC, mmHgc / 37.71 ± 7.84 / 42.55 ± 12.86 / 0.012* / 38.00 (31.75–44.25) / 36.10 (33.00–43.25) / 0.789
ABGA HCO3 before HFNC, mEq/Lc / 25.07 ± 5.73 / 26.69 ± 7.11 / 0.128 / 26.05 (21.48–29.30) / 24.85 (21.03–26.53) / 0.572
ABGA SpO2 before HFNC, %c / 90.50 (86.05–96.00) / 91.00 (87.95–95.85) / 0.837 / 89.50 (85.75–96.25) / 91.00 (87.83–95.00) / 0.533
ABGA FiO2 before HFNCc / 0.50 (0.44–0.80) / 0.50 (0.44–0.80) / 0.589 / 0.80 (0.49–0.80) / 0.50 (0.49–0.80) / 0.200
ABGA PaO2/FiO2 before HFNC, mmHgc / 114.32 (78.19–151.70) / 116.00 (80.88–156.63) / 0.930 / 90.38 (67.19–122.09) / 114.57 (77.25–143.75) / 0.202
ABGA lactic acid before HFNC, mmol/Lc / 1.50 (1.00–2.98) / 1.20 (0.80–2.08) / 0.087 / 1.50 (1.08–3.00) / 1.35 (0.88–2.45) / 0.586
Patients with hypercapnia (PCO2 ≥ 45 mmHg), n (%) / 17 (18.5) / 22 (38.6) / 0.007* / 9 (23.7) / 8 (21.1) / >0.999
Patients with hypercapnic acidosis (PCO2 ≥ 45 mmHg and pH < 7.35), n (%) / 5 (5.4) / 8 (14.0) / 0.081 / 3 (7.9) / 3 (7.9) / >0.999
CLD
Asthma, n (%) / 0 (0.0) / 4 (7.0) / 0.020* / 0 (0.0) / 1 (2.6) / >0.999
Emphysema, n (%) / 0 (0.0) / 26 (45.6) / <0.001* / 0 (0.0) / 16 (42.1) / <0.001*
Interstitial lung disease, n (%) / 0 (0.0) / 12 (21.1) / <0.001* / 0 (0.0) / 7 (18.4) / 0.016*
Bronchiectasis, n (%) / 0 (0.0) / 19 (33.3) / <0.001* / 0 (0.0) / 12 (31.6) / <0.001*
Pulmonary tuberculosis sequelae, n (%) / 0 (0.0) / 18 (31.6) / <0.001* / 0 (0.0) / 12 (31.6) / <0.001*
Lung cancer, n (%) / 0 (0.0) / 7 (12.3) / 0.001* / 0 (0.0) / 4 (10.5) / 0.125
Other chronic pulmonary disease, n (%) / 0 (0.0) / 3 (5.3) / 0.054 / 0 (0.0) / 1 (2.6) / >0.999

AHRF acute hypoxemic respiratory failure, APACHE Acute Physiology and Chronic Health Evaluation, CLD chronic lung disease, HFNC high flow nasal cannula, FiO2 fraction of inspired oxygen, L/min liter per minute, ICU intensive care unit, MICU medical intensive care unit, mmHg millimeter of mercury, min minute, ABGA arterial blood gas analysis, pH potential of hydrogen, PO2partial pressure of oxygen, PCO2partial pressure of carbon dioxide, HCO3bicarbonate, mEq/L milliequivalents per liter, SpO2 oxygen saturation, PaO2/FiO2, ratio of partial pressure arterial oxygen and fraction of inspired oxygen, mmol/L millimoles per liter

a Statistical comparisons of the data were performed by using the chi-square test for categorical variables and the Mann–Whitney U test or independent T-test (according to the results of Kolmogorov-Smirnov test) for continuous variables.

b Statistical comparisons of the data were performed by using the McNemar test for categorical variables and the Wilcoxon signed-rank test for continuous variables.

c Continuous variables are presented as median (interquartile range) or mean ± standard deviation (according to the results of Kolmogorov-Smirnov test).

Statistically significant p values are indicated by asterisks

Propensity model discrimination was assessed with an area under the receiver operating characteristic curve (AUROC = 0.817) and calibration was assessed with Hosmer-Lemeshow statistics (χ2 = 6.720; df = 8; P = 0.567).

Fig. E1 Schematic flow chart of the study population

1