Performance of oxygenation indices and risk scores to predict invasive mechanical ventilation and mortality in COVID-19

doi:10.21203/rs.3.rs-3246647/v1

Download PDF

Research Article

Performance of oxygenation indices and risk scores to predict invasive mechanical ventilation and mortality in COVID-19

https://doi.org/10.21203/rs.3.rs-3246647/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 02 Feb, 2024

Read the published version in BMC Pulmonary Medicine →

You are reading this latest preprint version

Background

Information on the performance of oxygenation indices (IO) and risk scores in patients requiring invasive mechanical ventilation (IMV) is limited. We determine the performance of the IOs and risk scores in hospitalized patients with COVID-19 to predict the requirement of IMV and death at 7, 14 and 28 days after admission.

Methods

Retrospective cohort study in patients admitted to the emergency department, hospitalization, and intensive care unit diagnosed with COVID-19 and in whom oxygenation indices were measured within the first 6 hours of admission, between 6 and 12 hours, 12 to 24 hours and greater than 24 hours. The receiver operating characteristic curve (ROC-curve) were built with the OIs and risk scores to predict IMV and mortality.

Results

A total of 1402 subjects entered the final analysis, of whom 19.5% (274/1402) received IMV and 23.0% (323/1402) died at 28 days. In the general population, the mean age was 59.9 years (SD 16.19), the male sex represented 63.3% (888/1402) and the duration from the onset of symptoms was 7.3 days (SD: 13.26). Analysis with the De Long test showed that at 7, 14 and 28 days the ROC-curves with the best performance for IMV were PaO2/FiO2, 4C score, SOFA score and SaO2/FiO2, the ROC curves of the rest of the OI evaluated were lower (p < 0.001). At 28 days, the ROC-curve with the best behavior for death was 4C compared to the ROC-curves of the rest of the OI and risk scores evaluated (p < 0.001).

Conclusion

COVID-19

Performance

Mechanical ventilation

Mortality

Infection caused by severe acute respiratory syndrome coronavirus type 2 (SARSCoV-2) is responsible for the rapid global spread and current pandemic of coronavirus disease 2019 (COVID-19), which continues to be a threat to public health due to the persistence of reported cases, 6.5 million deaths and the physical sequelae associated with the disease.(1, 2) 80% or more of COVID-19 cases are asymptomatic or mild, however, less than 15% patients with genetic predisposition, comorbidities, or advanced age may develop severe or critical illness with multiple organ failure and acute respiratory distress syndrome, complications causing a high expenditure on medical care resources and a high mortality rate.(2, 3)

The main clinical finding in patients with moderate to severe COVID-19 is decreased oxygen saturation values due to diffuse alveolar damage present in acute respiratory distress syndrome and an uncontrolled inflammatory state,(4, 5) generating an imbalance of gas exchange due to destruction of the lung parenchyma, hypercatabolic state and a greater requirement of oxygen pressures for its adequate diffusion through the alveolar capillary membrane.(5, 6) Therefore, patients with severe COVID-19 may require early invasive mechanical ventilation, continuous positive airway pressure, or high-flow nasal cannula, together with strict follow-up consisting of clinical and laboratory tests, including oxygenation indices (OIs).(4, 7, 8)

The use of OIs as predictors of clinical evolution in patients with COVID-19 has already been studied by several authors; Sinatti et al,(9) conducted a cohort study with 150 patients and found that arterial pressure of oxygen/fraction of inspired oxygen (PaO2/FiO2 ratio) can be considered a reliable prognostic biomarker to differentiate severe from mild disease with yield based on the receiver operating characteristic curve (ROC-curve) of 0.838. Xie J et al,(10) showed in 140 patients with pneumonia associated with COVID-19 that supplemental oxygen with a peripheral oxygen saturation (SpO2) lower than 90% was related to sample in more than 60% of the patients, and that an increase in SpO2 improved survival by 8% (HR: 0.92; 95% CI: 0.91 to 0.94; p < 0.001). These results suggest that the evaluation of baseline OI values and their change values in the first hours of clinical follow-up can guide the evolution and prognosis of these patients. The other hand, ROX index (‘Respiratory rate-OXygenation’) is the ratio between peripheral blood oxygen saturation (SpO2) to fraction of inspired oxygen (FiO2) and respiratory rate, which has shown promising performance in successful prognosis of invasive and non-invasive oxygen therapy. (11–13)

OIs are useful markers for the prediction of complications and mortality in patients with lung involvement, including subjects with COVID-19.(4, 9, 10) Currently, information on the performance and discriminatory capacity of OIs in patients requiring invasive ventilatory support and mortality are limited, generating significant variability in clinical practice.(8, 14) Based on this, this study aims to determine the performance of OIs, delta of partial pressure of arterial oxygen to fraction of inspired oxygen (PaO2/FiO2) ratio, delta SpO2/FiO2 ratio, and risk scores in hospitalized patients with COVID-19 to predict invasive mechanical ventilation (IMV) and mortality at 7, 14 and 28 days after admission.

Study design

This was a retrospective cohort study in patients admitted to the emergency department, hospitalized, and admitted to intensive care unit (ICU) of the Clínica Universidad of La Sabana between March 2020 and March 2022.

Eligibility criteria

Accordingly to its eligibility criteria, the patients included in this study were: at least 18 years of age, hospitalized for COVID-19 for more than 72 hours, with an infection by SARS-CoV-2 confirmed with a positive reverse transcription polymerase chain reaction, obtained with a nasopharyngeal swab/tracheal aspirate or sputum sample, without any history of congenital heart disease determining chronic hypoxemia, with no need of mechanical ventilation within 6 hours of admission to the emergency room, and with at least 2 arterial blood gas measurements. Patients without clinical follow-up and patients referred to other hospitals were excluded.

Analyzed variables

Demographic data, presence of comorbidities (Charlson index), signs and symptoms, severity test on admission, complete blood count, and arterial blood gas values were analyzed. Ios (SpO2/FiO2 ratio, PaO2/FiO2 ratio) and ROX index were calculated from the first 6 hours of

admission, between 6 to 12 hours, 12 to 24 hours, and over 24 hours. (15–17) Each of the indices were calculated from oximetry records and arterial blood gas measurements. The duration of hospitalization, vasopressor support in the ICU, IMV requirement (7, 14 and 28 days), and death (7, 14 and 28 days) were evaluated. The ROX index was interpreted with the cut-off point ≥ 4.88 at 12 hours as an indicator of failure.(18) SOFA score and 4C were calculated as additional risk scores based on the original validation studies. (19, 20)

Sample size

To estimate the sample size, data from the study carried out by Alberdí et al,(21) that evaluated SpO2/FiO2 ratio and ROX index, where a sensitivity of 26.8%, specificity of 89.5% for these indices and an expected mortality of 35.3% are reported. For a confidence level of 95% and a precision of 10%, a minimum of 856 patients is required.(22)

Statistical analysis

Data was fully collected and compiled using a secure server (Research Electronic Data Capture, REDCap software) and later analyzed in the SPSS (Statistical Package for the Social Sciences, Chicago, IL) for Windows version 25.(23, 24) Quantitative variables were summarized in means and standard deviations if their distribution was normal, or median and interquartile range if their distribution was not normal. The qualitative variables were summarized in frequencies and percentages. A bivariate analysis was carried out comparing the OIs and the study variables with the outcomes of IMV and mortality. For missing data, a weighted median was performed for quantitative variables and logistic regression for qualitative variables. (25)

Differences (delta) were calculated between the SpO2/FiO2 ratio, the PaO2/FiO2 ratio on admission, with those obtained in the first 6 hours, between 6 and 12 hours, between 12 and 24 hours, and over 24 hours. In addition, risk scores (ROX index, 4C score, and SOFA score) were calculated to construct the ROC-curve with the results of IMV requirement and mortality. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (LR+), negative likelihood ratio (LR-) with their respective 95% confidence intervals were calculated. The cut-off points for the calculated deltas were obtained through the Youden index. A comparison was made between the different ROC-curves obtained, using the DeLong test.(26) A p value adjusted by Bonferroni is considered for the comparison of the different OI and risk scores less than < 0.006.

Ethical considerations

This study was approved by the Ethics Committee of the Clínica Universidad de La Sabana (approval number: 20220602), considering it as risk-free research according to resolution 8430 of 1993, and respecting the protection of personal data according to the habeas data law 1266 of 2008.

Patient and public involvement

Patients were not involved in the development of the research question, design, recruitment, or intervention burden assessed; no patient advisors were required, and data were analyzed anonymously. The results will be disseminated to the scientific community in academic writing.

A total of 1402 subjects entered the final analysis, of whom 19.5% (274/1402) received IMV and 23.0% (323/1402) died at 28 days. In Fig. 1 the entry flow of subjects to the study is shown. In the general population, the mean age was 59.9 years (SD 16.19), the male sex represented 63.3% (888/1402) and the duration from the onset of symptoms was 7.3 days (SD: 13.26). In the deceased population, 46.4% (150/323) had systemic arterial hypertension compared to 34.5% (372/1079) of the surviving patients (p < 0.001). Cough and crackles occurred in 47.7% (154/323) and 38.4% (124/323) of the patients who died, respectively. The general characteristics are described in Table 1.

Table 1

Demographic characteristics, medical history, and risk scores.
	Total population n = 1402	Death n = 323	Alive n = 1079	p value
Age years, mean(sd)	59,9 (16,19)	69.0 (14,22)	57,2 (15,77)	< 0,001
Male gender, n (%)	888 (63,3)	212 (65,6)	676 (62,7)	0,329
Days since symptoms start, mean(sd)	7,3 (13,26)	6,2 (5,44)	7,6 (14,8)	0,009
Pathological history, n(%)
Systemic arterial hypertension	522 (37,2)	150 (46.4)	372 (34.5)	< 0.001
Smoking	542 (38,7)	157 (48,6)	385 (35,7)	< 0.001
Myocardial infarction	42 (3)	15 (4,6)	27 (2,5)	0.048
Heart failure	50 (3,6)	22 (6,8)	28 (2,6)	< 0.001
Peripheral vascular disease	10 (0,7)	2 (0,6)	8 (0,7)	< 0.001
Cerebrovascular disease	33 (2,4)	18 (5,6)	15 (1,4)	< 0.001
Chronic lung disease	102 (7,3)	41 (12,7)	61 (5,7)	< 0,001
Diabetes	218 (15,5)	61 (18,9)	157 (14,6)	0,059
Chronic kidney disease	75 (5,3)	29 (9)	46 (4,3)	< 0,001
Charlson index, mean (sd)	2,3 (2,18)	3,6 (2,45)	2 (1,94)	< 0.001
4C, mean (sd)	8,7 (3,75)	10,6 (3,63)	8 (3,58)	< 0.001
SOFA, mean (sd)	3,6 (2,28)	4,4 (2,5)	3,4 (2,14)	< 0.001
Notes: sd: standard deviation, ROX: Respiratory rate-OXygenation index, SOFA: Sequential Organ Failure Assessment.

The mean pH was 7.43 (SD: 0.07) in the patients who had a fatal outcome versus 7.45 (SD: 0.06) in the survivors (p < 0.001). In the population that died, the average C-reactive protein was 146.4 (SD: 115.74) compared to 121.5 (SD: 107.42) in living patients (p < 0.001). Laboratory tests and diagnostic images are described in Table 2. The OIs changes in the different hourly periods (less than 6 hours, from 6 to 12 hours, from 12 to 24 hours, and greater than 24 hours) versus the outcome of IMV and mortality are described in Supplementary table 1 and 2.

Table 2

Laboratory tests. and diagnostic images.
	Total population n = 1402	Death n = 323	Alive n = 1079	p value
Ph, mean(sd)	7.45 (0.06)	7.43 (0.07)	7.45 (0.06)	< 0.001
PaO2 (mmHg), mean(sd)	69 (24.02)	68.3 (24.25)	69.2 (23.96)	0.580
PaCO2(mmHg), mean(sd)	31.7 (6.74)	31.9 (8.35)	31.6 (6.18)	0.859
HCO3 (mE q/L), mean(sd)	22.2 (4.22)	21.7 (4.89)	22.3 (3.99)	< 0.001
BE (mE q/L), mean(sd)	-1.2 (4.05)	-1.9 (4.75)	-1 (3.8)	0.002
Lactate (mmol/L), mean(sd)	1.9 (7.56)	3 (15.72)	1.5 (0.92)	0.089
SaO2 (%), mean(sd)	90.6 (9.76)	89.9 (11.26)	90.8 (9.27)	0.196
Leukocytes, mean(sd)	9537.6 (6981.9)	10410.1 (12053.09)	9275.4 (4420.95)	0.097
Lymphocytes, mean(sd)	1145.7 (950.82)	985.9 (904.55)	1193.6 (959.49)	< 0.001
Neutrophils, mean(sd)	7794.1 (5876.31)	8202.8 (4584.64)	7671.4 (6208.4)	0.094
Hemoglobine, mean(sd)	14.2 (2.44)	13.5 (2.83)	14.4 (2.27)	< 0.001
Hematocrit, mean(sd)	42 (7.68)	40.7 (8.45)	42.5 (7.38)	< 0.001
Plateles, mean(sd)	246220.1 (100052.78)	231270 (104610.61)	250710.8 (98249.81)	< 0.001
Albumin, mean(sd)	3 (0.62)	2.9 (0.6)	3.1 (0.64)	< 0.001
Creatinine, mean(sd)	1.5 (3.57)	1.8 (2.64)	1.4 (3.8)	0.032
Ureic nitrogen, mean(sd)	22.1 (15.79)	29.1 (22.25)	20 (12.38)	< 0.001
D dimer, mean(sd)	1827.2 (4861.18)	2811.6 (7119.7)	1536.7 (3914.08)	0.002
LDH, mean(sd)	430.8 (252.17)	506.5 (309.53)	407.9 (227.34)	< 0.001
GOT, mean(sd)	53.7 (55.81)	57 (45.96)	52.6 (58.52)	0.158
GGT, mean(sd)	54.1 (63.57)	48.7 (49.79)	55.7 (67.24)	0.042
Troponin, mean(sd)	37 (101.94)	71.2 (154.89)	26.7 (76.66)	< 0.001
Procalcitonin, mean(sd)	2.1 (8.03)	2.9 (10.1)	1.7 (6.76)	0.047
CRP, mean(sd)	127.3 (109.87)	146.4 (115.74)	121.5 (107.42)	< 0.001
Bilirubin, mean(sd)	0.8 (2.74)	1.2 (5.51)	0.7 (0.46)	0.093
Ferritin, mean(sd)	1322.4 (2107.67)	1552.8 (2139.67)	1246.3 (2093.1)	0.022
CPK, mean(sd)	453.1 (677.2)	574.1 (835.19)	358.7 (508.47)	< 0.001
Normal limits, mean(sd)	93 (6.63)	13 (4)	80 (7.4)	0.032
Interstitial opacity on CT, mean(sd)	637 (45.44)	148 (45.8)	489 (45.3)	0.874
Alveolar opacity on CT, mean(sd)	446 (31.81)	115 (35.6)	331 (30.7)	0.095
Ground-glass opacity on CT, mean(sd)	556 (39.66)	143 (44.3)	413 (38.3)	0.053
Notes: sd: standard deviation, PaO2: arterial oxygen pressure, PaCO2: arterial carbon dioxide pressure, HCO3: bicarbonate, BE: base excess, SaO2: arterial oxygen saturation, LDH: lactate dehydrogenase, GOT: glutamic-oxaloacetic transaminase, GGT: gamma-glutamyl transferase, CRP: C Reactive protein, CPK: Creatine phosphokinase, CT: computer tomography.

The ROC-curve of the delta PaO2/FiO2 ratio for the requirement of IMV and mortality at 28-day was 0.589 (95% CI: 0.546–0.632) and 0.567 (95% CI: 0.526–0.608), respectively Table 3. PaO2/FiO2 ≤ 300 shows a ROC curve of 0.669 (95% CI: 0.628–0.711) to predict IMV. PaO2/FiO2 ≤ 300 and 4C score in mortality at 28 days showed an ROC-curve of 0.624 (95% CI: 0.582–0.667) and 0.706 (95% CI: 0.669–0.742), respectively.

Table 3

Performance of invasive mechanical ventilation and mortality at 28-day.
	S (IC 95%)	Sp(IC 95%)	PPV (CI 95%)	NPV (CI 95%)	LR+ (CI 95%)	LR- (CI 95%)	ROC(IC95%)	p value
Invasive Mechanical Ventilation
PaO2 mmHg < = 60	55.3 (52.7–57.9)	36.8 (34.2–39.3)	44.7 (17.9–15.8)	76.8 (74.6–79)	1.21 (1.082–1.364)	0.88 (0.779–0.983)	0.536 (0.496–0.576)	0.085
SaO2% <= 90	61.4 (58.8–64)	31 (28.5–33.4)	17.5 (15.5–19.6)	77 (74.8-0)	1.25 (1.123–1.384)	0.89 (0.801–0.987)	0.545 (0.505–0.585)	0.032
PaO2/FiO2 ≤ 300	13.9 (12.1–15.8)	71.5 (69.1–73.9)	10.8 (9.2–12.5)	77 (74.7–79.2)	1.2 (0.884–1.641)	0.49 (0.358–0.665)	0.669 (0.628–0.711)	< 0.001
SaO2/FiO2 ≤ 350	54.7 (52-57.3)	25.5 (23.2–27.9)	15.1 (13.2–17)	69.9 (67.5–72.3)	1.77 (1.582–1.989)	0.73 (0.655–0.824)	0.629 (0.588–0.67)	< 0.001
ROX index ≥ 4.8	89.4 (87.8–91.1)	2.5 (1.7–3.4)	18.1 (16.1–20.1)	50 (47.3–52.7)	4.15 (3.978–4.331)	0.92 (0.88–0.958)	0.608 (0.565–0.65)	< 0.001
ΔPaO2/FiO2 ratio	83 (81–85)	5.4 (4.2–6.6)	17.7 (15.7–19.7)	56.3 (53.7–59)	3.16 (2.992–3.348)	0.88 (0.877–0.928)	0.612 (0.573–0.651)	< 0.001
ΔSaO2/FiO2 ratio	90.5 (88.9–92.2)	4.8 (3.6-6)	20.8 (18.6–23.1)	64.8 (62.1–67.5)	1.96 (1.884–2.047)	0.95 (0.912–0.992)	0.589 (0.546–0.632)	< 0.001
ROX index ≥ 5.2	66.2 (63.6–68.7)	17.7 (15.7–19.8)	0.1 (-0.1-0.2)	99.9 (99.7-100.1)	1.91 (1.744–2.091)	0.8 (0.734–0.882)	0.59 (0.548–0.632)	< 0.001
SOFA score ≥ 4	55.1 (51.6–58.7)	76.6 (73.6–79.6)	42 (38.5–45.5)	84.7 (82.2–87.3)	2.36 (2.039–2.724)	0.59 (0.507–0.677)	0.633 (0.595–0.671)	< 0.001
4C score ≥ 8	72.1 (69.1–75.1)	48.7 (45.4–52)	29.8 (26.7–32.8)	85.2 (82.9–87.6)	1.4 (1.301–1.515)	0.57 (0.532–0.62)	0.64 (0.602–0.677)	< 0.001
Charlson index ≥ 3	48.5 (45.9–51.1)	61.5 (59-64.1)	23.5 (21.2–25.7)	83.1 (81.2–85.1)	1.26 (1.158–1.374)	0.84 (0.768–0.911)	0.563 (0.524–0.602)	0.003
Mortality
PaO2 mmHg < = 60	59.2 (56.6–61.8)	37.6 (35.1–40.2)	22.1 (19.9–24.3)	75.5 (73.2–77.8)	1.08 (0.979–1.157)	0.95 (0.856–1.052)	0.515 (0.474–0.557)		0.449
SaO2% <= 90	66 (63.5–68.5)	32 (29.5–34.5)	22.3 (20.1–24.5)	76.1 (73.8–78.4)	1.06 (0.986–1.25)	0.97 (0.886–1.063)	0.513 (0.472–0.553)		0.536
PaO2/FiO2 ≤ 300	70.6 (68.2–73)	9.3 (7.8–10.8)	18.9 (18.9–21)	51.3 (48.7–54)	3.17 (2.943–3.411)	0.78 (0.723–0.838)	0.624 (0.582–0.667)		< 0.001
SaO2/FiO2 ≤ 350	58.8 (56.2–61.4)	25.9 (23.6–28.2)	19.1 (17-21.1)	67.9 (65.4–70.4)	1.59 (1.441–1.757)	0.79 (0.718–0.876)	0.581 (0.539–0.623)		< 0.001
ROX index ≥ 4.8	91.6 (90.2–93.1)	2.8 (2-3.7)	21.8 (19.6–24)	53.6 (50.9–56.2)	2.94 (2.836–3.05)	0.94 (0.911–0.977)	0.588 (0.545–0.63)		< 0.001
ΔPaO2/FiO2 ratio	81.1 (79-83.2)	6.5 (5.2–7.8)	20.7 (18.6–22.9)	53.2 (50.5–55.8)	2.92 (2.76–3.086)	0.87 (0.82–0.917)	0.547 (0.507–0.587)		0.021
ΔSaO2/FiO2 ratio	87.9 (86-89.7)	7.3 (5.9–8.8)	20.8 (18.5–23.1)	68.6 (66-71.2)	1.65 (1.574–1.734)	0.95 (0.904–0.995)	0.567 (0.526–0.608)		0.001
ROX index ≥ 5.2	70.2 (67.8–72.7)	18.2 (16.2–20.3)	20.5 (18.3–22.6)	67.1 (64.6–69.7)	1.63 (1.512–1.767)	0.86 (0.794–0.928)	0.573 (0.532–0.614)		< 0.001
SOFA score ≥ 4	58.1 (54.6–61.7)	75.9 (72.8–78.9)	37.2 (33.8–40.7)	88 (85.7–90.4)	2.41 (1.978–2.933)	0.55 (0.453–0.672)	0.633 (0.595–0.67)		< 0.001
4C score ≥ 8	74.9 (72-77.8)	48.2 (44.9–51.5)	25.3 (22.4–28.2)	89.1 (87.1–91.2)	1.45 (1.292–1.619)	0.52 (0.466–0.584)	0.706 (0.669–0.742)		< 0.001
Charlson index ≥ 3	49.8 (47.2–52.4)	61.3 (58.7–63.8)	19.2 (17.2–21.3)	86.8 (85.1–88.6)	1.29 (1.105–1.495)	0.82 (0.705–0.953)	0.672 (0.636–0.709)		< 0.001
Notes: S: Sensibility, Sp: specificity, PPV: positive predictive value, NPV: negative predictive value, LR +: positive likelihood ratio, LR-: negative likelihood ratio, CI: confidence intervals, ROC: receiver operating characteristic curve, PaO2: arterial oxygen pressure, SaO2: arterial oxygen saturation, Δ: delta, PaO2/FiO2 ratio: arterial oxygen pressure/inspired fraction of oxygen, SaO2/FiO2 ratio: arterial oxygen saturation in relation to the inspired oxygen fraction, ROX: Respiratory rate-OXygenation index, SOFA: Sequential Organ Failure Assessment.

Analysis with the De Long test showed that at 7, 14 and 28 days the ROC-curves with the best performance for IMV were PaO2/FiO2, 4C score, SOFA score and SaO2/FiO2, the ROC curves of the rest of the OI evaluated were lower (p < 0.001). At 14 days, ROC-curves with the best behavior for death were the 4C score and the Charlson index, the ROC-curves of the rest of the OI and SOFA scores evaluated were lower (p < 0.001). At 28 days, the ROC-curve with the best behavior for death was 4C compared to the ROC-curves of the rest of the OI and risk scores evaluated (p < 0.001). At 7 days, no statistically significant differences were found for multiple comparisons between the OI and risk scores assessed for death (p = 0.043).

In Supplementary table 3 the different performance values of the OIs and risk scores for the 7 and 14-days outcomes are shown. The ROC-curve of the OIs and the risk scores to predict IMV and mortality are shown in the Fig. 2 and Fig. 3.

The present study determined the performance of the change in OIs and risk scores in a cohort of 1402 hospitalized patients with COVID-19, finding that the PaO2/FiO2, 4C score, SOFA score y SaO2/FiO2 were weak predictors of the IMV requirement from admission, and 4C score was weak predictors of the mortality from admission. Characteristics such as age, pathological history, and clinical manifestations occurred more frequently among patients who died from SARS-CoV-2. The delta PaO2/FiO2 and the delta SaO2/FiO2 showed inferior performance for IMV and death compared to the other tools evaluated.

PaO2/FiO2 reflects the severity of hypoxemia and given its performance, has been correlated with prognosis and hospital mortality in patients with acute respiratory failure due to COVID-19.(27–29). Direct blood sample collection for gasometric analysis it is the main reference for evaluating the oxygenation status of patients with respiratory diseases,(9) and its baseline value is the one that offers the most information in the evaluation of the patient.(9, 28, 29) In our study, the PaO2/FiO2 ratio showed a weak discriminatory capacity for IMV or death. In addition, the continuous evaluation of oxygenation status through gas analysis constitutes an expensive invasive procedure and was not available at the different levels of care, requiring the continuous study of other measurements that use pulse oximetry to obtain indices such as the SpO2/FiO2 ratio and ROX index, which can reflect the state of hypoxemia in a non-invasive way.(30)

Knight et al.,(19) developed and validated a 4C score (Coronavirus Clinical Characterization Consortium) of risk for the prediction of mortality in a derivation cohort with 35.463 patients and validation with 22.361 patients hospitalized for COVID-19. The variables of age, gender, number of comorbidities, respiratory rate, SpO2, state of consciousness, urea nitrogen, and C Reactive protein were included in the score; obtaining a scale of 0 to 21 points and a performance of 0.79 (95% CI: 0.78–0.79) in the derivation cohort and 0.77 (95% CI: 0.76–0.77) in the validation cohort for mortality, similar findings in our study where greater comorbidity measured by Charlson and greater severity by 4C were related to greater mortality in patients with COVID-19.(19, 31, 32).

Cattazzo et al.,(33) analyzed the efficiency of the ROX index and the SaO2/FiO2 ratio compared to PaO2/FiO2 ratio for the prediction of death or IMV requirement in 456 patients hospitalized in areas other than the ICU due to COVID-19. The results showed an ROC-curve of 0.67 (95% CI: 0.62–0.73; p < 0.001) for PaO2/FiO2 ratio, 0.69 (95% CI: 0.63–0.74; p < 0.001) for the ROX index and 0.66 (95% CI: 0.60–0.72; p < 0.001) for SpO2/FiO2 ratio; similar situation to what we find where the basal values of each of these indices have a weak discriminatory capacity for IMV or death. Baek et al.,(34) in patients with COVID-19 and supplemental oxygen with a high-flow cannula found that the change in the ROX index and SaO2/FiO2 ratio was related to early IMV; findings that were corroborated with our results where a drop in SaO2/FiO2 ratio of 150 points was associated with a greater requirement of IMV and death, the change in this variable being the one with the best efficiency for the prediction of complications in patients with SARS-CoV2 infection.

Grasselli et al.,(35) described demographic characteristics, comorbidities, and treatments of a cohort of 3,988 critically ill patients with SARS-CoV-2 infection. Hypertension, hypercholesterolemia, heart disease, diabetes, chronic obstructive pulmonary disease, and chronic kidney disease were associated with increased mortality. Our results confirm that deaths from COVID-19 were higher in patients with multiple comorbidities, a consequence of a pre-existing inflammatory state in chronic diseases and greater susceptibility to a cellular immune response and humoral activation, predominantly of tumor necrosis factor and interleukin 6.(36, 37). This was possibly related to what was observed in most clinical scenarios as described during the pandemic period, in which the presence of other respiratory diseases directly influences the loss of alveolar reserve, promoting greater pulmonary compromise secondary to viral infection.(38, 39)

Limitations

As it was a retrospective study based on medical records, this study may give rise to information biases; however, the research group has adequate training for the analysis and interpretation of the results. Similarly, being a single center study may limit the extrapolation of the results, despite this, there was a sufficient sample size to support them. The altitude above sea level where the care center was located can be considered a limiting factor, since exceeding 2.500 meters above sea level can alter oxygenation values, as described in previous studies. (13, 40) However, there are large numbers of the world's population residing at altitudes to which these findings could be applied. The lack of a uniform take in the evaluation of arterial gasses during the evolution of patients with SARS-CoV2 can cause confusion throughout the follow-up, however, each of the analyzes at different periods of time carried out in the study consistently show differences in the OIs evaluated. We consider it pertinent to carry out future studies that corroborate our findings.

PaO2/FiO2 ≤ 300, 4C score ≥ 8, SOFA score ≥ 4 y SaO2/FiO2 ≤ 300 were weak predictors of the IMV requirement from admission, and 4C score ≥ 8 was weak predictors of the mortality from admission in patients with pulmonary involvement by COVID-19. Age, pathological history, and clinical manifestations occurred more frequently among patients who died from SARS-CoV-2.

Acknowledgements

The authors are most thankful for the Universidad de La Sabana.

Authors' contributions

ARB, ETQ, MFA, AVM, HCA, JMV, SG, PM, MM, SP, DP, and MMB contributed substantially to the study design, data analysis and interpretation, and manuscript writing. ARB and EAT had full access to all study data and takes responsibility for data integrity as well as for accuracy of the included data analysis and, especially, any adverse effects. All authors read and approved the final manuscript.

Funding

This work was supported by Universidad de la Sabana Grant MEDMSc-63-2023.

Availability of data and materials

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate

The study was conducted in accordance with the principles of the current Helsinki Declaration, as well as local, regional, and international regulations pertaining to clinical research, including Colombian Law on Biomedical Research. Ethical approval was obtained from the Medical Ethics Committee of the Clínica Universidad de La Sabana (approval number 20220602). Prior to participating in the study, all participants provided written informed consent, and the confidentiality of their data was strictly maintained throughout the study.

Consent for publication

Not applicable

Competing interests

The authors declare that there are no competing interests associated with the manuscript.

Coronavirus disease 2019 dashboard: World Health Organization; [5–12-22]. Available from: https://covid19.who.int/.
Singh SP, Pritam M, Pandey B, et al. Microstructure, pathophysiology, and potential therapeutics of COVID-19: A comprehensive review. J Med Virol. 2021; 93(1):275-299.
Safiabadi Tali SH, LeBlanc JJ, Sadiq Z, et al. Tools and Techniques for Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2)/COVID-19 Detection. Clin Microbiol Rev. 2021; 12;34(3):e00228-20.
Beloncle F, Studer A, Seegers V, et al. Longitudinal changes in compliance, oxygenation and ventilatory ratio in COVID-19 versus non-COVID-19 pulmonary acute respiratory distress syndrome. Crit Care. 2021; 25(1):248.
Meyer NJ, Gattinoni L, Calfee CS. Acute respiratory distress syndrome. Lancet. 2021; 398(10300):622-637.
Swenson KE, Swenson ER. Pathophysiology of Acute Respiratory Distress Syndrome and COVID-19 Lung Injury. Crit Care Clin. 2021; 37(4):749-776.
Bonnesen B, Jensen JS, Jeschke KN, et al. Management of COVID-19-Associated Acute Respiratory Failure with Alternatives to Invasive Mechanical Ventilation: High-Flow Oxygen, Continuous Positive Airway Pressure, and Noninvasive Ventilation. Diagnostics (Basel). 2021; 11(12):2259.
Perkins GD, Ji C, Connolly BA, Couper K, et al. Effect of Noninvasive Respiratory Strategies on Intubation or Mortality Among Patients With Acute Hypoxemic Respiratory Failure and COVID-19: The RECOVERY-RS Randomized Clinical Trial. JAMA. 2022 ;327(6):546-558.
Sinatti G, Santini SJ, Tarantino G, et al. PaO2/FiO2 ratio forecasts COVID-19 patients' outcome regardless of age: a cross-sectional, monocentric study. Intern Emerg Med. 2022; 17(3):665-673.
Xie J, Covassin N, Fan Z, et al. Association Between Hypoxemia and Mortality in Patients With COVID-19. Mayo Clin Proc. 2020; 95(6):1138-1147.
Roca O, Messika J, Caralt B, et al Predicting success of high-flow nasal cannula in pneumonia patients with hypoxemic respiratory failure: the utility of the roX index. J Crit Care 2016;35:200–5.
Roca O, Caralt B, Messika J, et al. An index combining respiratory rate and oxygenation to predict outcome of nasal high-flow therapy. Am J Respir Crit Care Med 2019;199:1368–76.
Reyes LF, Bastidas Goyes A, Tuta Quintero EA, et al. Validity of the ROX index in predicting invasive mechanical ventilation requirement in pneumonia. BMJ Open Respir Res. 2022 ;9(1):e001320.
Azoulay E, de Waele J, Ferrer R, et al. International variation in the management of severe COVID-19 patients. Crit Care. 2020 ;24(1):486.
Ferguson ND, Fan E, Camporota L, et al. The Berlin definition of ARDS: An expanded rationale, justification, and supplementary material. Intensive Care Med. 2012;38(10):1573–82.
Martinez G, Casas D, Bastidas A, et al. Índices de oxigenación como predictores de ventilación mecánica en neumonía a 2600 metros de altitud. Acta Médica Colomb. 2017 Oct 30;41(3).
Lu X, Jiang L, Chen T, Wang Y, et al. Continuously available ratio of SpO2/FiO2serves as a noninvasive prognostic marker for intensive care patients with COVID-19. Respir Res. 2020;21(1):1–4.
Roca O, Messika J, Caralt B, et al. Predicting success of high-flow nasal cannula in pneumonia patients with hypoxemic respiratory failure: The utility of the ROX index. J Crit Care. 2016; 35:200-5.
Knight SR, Ho A, Pius R, et al; ISARIC4C investigators. Risk stratification of patients admitted to hospital with covid-19 using the ISARIC WHO Clinical Characterisation Protocol: development and validation of the 4C Mortality Score. BMJ. 2020 ;370:m3339.
Pandharipande PP, Shintani AK, Hagerman HE, et al. Derivation and validation of SpO2/FiO2 ratio to impute for PaO2/FiO2 ratio in the respiratory component of the Sequential Organ Failure Assessment (SOFA) score. Crit Care Med. 2009;37:1317-21.
Alberdi-Iglesias A, Martín-Rodríguez F, Ortega Rabbione G, et al. Role of SpO2/FiO2 Ratio and ROX Index in Predicting Early Invasive Mechanical Ventilation in COVID-19. A Pragmatic, Retrospective, Multi-Center Study. Biomedicines. 2021; 18;9(8):1036.
Obuchowski NA. Sample size calculations in studies of test accuracy. Stat Methods Med Res. 1998; 7(4):371-92.
Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)-A metadata-driven methodology and workflow process for providing translational research informatics support. Journal of Biomedical Informatics. 2009;42(2):377-81.
IBM Corp. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp; 2017. Available from: https://www.ibm.com/support/pages/downloading-ibm-spss-statistics-25
Akl EA, Shawwa K, Kahale LA, et al. Reporting missing participant data in randomised trials: systematic survey of the methodological literature and a proposed guide. BMJ Open. 2015 ;5(12):e008431.
DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988 ;44(3):837-45.
Babu S, Abhilash KP, Kandasamy S, et al. Association between SpO2/FiO2 Ratio and PaO2/FiO2 Ratio in Different Modes of Oxygen Supplementation. Indian J Crit Care Med. 2021; 25(9):1001-1005.
Sartini S, Massobrio L, Cutuli O, et al. Role of SatO2, PaO2/FiO2 Ratio and PaO2 to Predict Adverse Outcome in COVID-19: A Retrospective, Cohort Study. Int J Environ Res Public Health. 2021; 2;18(21):11534.
Bracegirdle L, Jackson A, Beecham R, et al. Dynamic blood oxygen indices in mechanically ventilated COVID-19 patients with acute hypoxic respiratory failure: A cohort study. PLoS One. 2022; 10;17(6):e0269471.
Tsai PH, Lai WY, Lin YY, et al. Clinical manifestation and disease progression in COVID-19 infection. J Chin Med Assoc. 2021; 1;84(1):3-8.
Tuty Kuswardhani RA, Henrina J, Pranata R, et al. Charlson comorbidity index and a composite of poor outcomes in COVID-19 patients: A systematic review and meta-analysis. Diabetes Metab Syndr. 2020;14(6):2103-2109.
Yang X, Yu Y, Xu J, et al Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study. Lancet Respir Med. 2020 ;8(5):475-481.
Cattazzo F, Inglese F, Dalbeni A, et al. Performance of non-invasive respiratory function indices in predicting clinical outcomes in patients hospitalized for COVID-19 pneumonia in medical and sub-intensive wards: a retrospective cohort study. Intern Emerg Med. 2022; 17(4):1097-1106.
Baek AR, Seong GM, Lee SI, et al. Late failure of high-flow nasal cannula may be associated with high mortality in covid-19 patients: A multicenter retrospective study in the republic of korea. J Pers Med. 2021; 11(10):1–10.
Grasselli G, Greco M, Zanella A, et al. Risk Factors Associated With Mortality Among Patients With COVID-19 in Intensive Care Units in Lombardy, Italy. JAMA Intern Med. 2020;180(10):1345-1355.
Tuta E, Vega M, Perdomo L, Pimentel J. Colchicina, perspectivas de un viejo amigo para la reumatología en la COVID-19: una revisión exploratoria. Rev Col Reuma. 2022; 29:S35–47.
Anka AU, Tahir MI, Abubakar SD, et al. Coronavirus disease 2019 (COVID-19): An overview of the immunopathology, serological diagnosis and management. Scand J Immunol. 2021; 93(4):e12998.
Liu K, Chen Y, Lin R, et al. Clinical features of COVID-19 in elderly patients : A comparison with young and middle-aged patients. J Infect. 2020;80(6):e14–8.
Chen T, Dai Z, Mo P, et al. Clinical Characteristics and Outcomes of Older Patients with Coronavirus Disease 2019 (COVID-19) in Wuhan, China: A Single-Centered, Retrospective Study. J Gerontol A Biol Sci Med Sci. 2020;75(9):1788–95.
Rojas-Camayo J, Mejia CR, Callacondo D, et al. Reference values for oxygen saturation from sea level to the highest human habitation in the Andes in acclimatised persons. Thorax. 2018; 73(8):776-778.

No competing interests reported.

Download PDF

Journal Publication

published 02 Feb, 2024

Read the published version in BMC Pulmonary Medicine →

Editorial decision: Revision requested
18 Oct, 2023
Editorial decision: Major revision
18 Oct, 2023
Reviews received at journal
14 Oct, 2023
Reviews received at journal
11 Oct, 2023
Reviewers agreed at journal
05 Sep, 2023
Reviewers agreed at journal
05 Sep, 2023
Reviewers invited by journal
05 Sep, 2023
Editor assigned by journal
05 Sep, 2023
Editor invited by journal
05 Sep, 2023
Submission checks completed at journal
05 Sep, 2023
First submitted to journal
08 Aug, 2023

You are reading this latest preprint version

Performance of oxygenation indices and risk scores to predict invasive mechanical ventilation and mortality in COVID-19

Status:

Journal Publication

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

INTRODUCTION

METHODS

Study design

Eligibility criteria

Analyzed variables

Sample size

Statistical analysis

Ethical considerations

Patient and public involvement

RESULTS

DISCUSSION

Limitations

CONCLUSIONS

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1