Temporal trends of severity and outcomes of critically ill patients with COVID-19 after the emergence of variants of concern: A comparison of two waves.

Introduction

Background: The emergence of SARS-CoV-2 variants led to subsequent waves of COVID-19 worldwide. In many countries, the second wave of COVID-19 was marked by record deaths, raising the concern that variants associated with that wave might be more deadly. Our aim was to compare outcomes of critically-ill patients of the first two waves of COVID-19. Methods: This retrospective cohort included critically-ill patients admitted between March-June 2020 and April-July 2021 in the largest academic hospital in Brazil, which has free-access universal health care system. We compared admission characteristics and hospital outcomes. The main outcome was 60‐day survival and we built multivariable Cox model based on a conceptual causal diagram in the format of directed acyclic graph (DAG). Results: We included 1583 patients (1315 in the first and 268 in the second wave). Patients in the second wave were younger, had lower severity scores, used prone and non-invasive ventilatory support more often, and fewer patients required mechanical ventilation (70% vs 80%, p<0.001), vasopressors (60 vs 74%, p<0.001), and dialysis (22% vs 37%, p<0.001). Survival was higher in the second wave (HR 0.61, 95%CI 0.50–0.76). In the multivariable model, admission during the second wave, adjusted for age, SAPS3 and vaccination, was not associated with survival (aHR 0.85, 95%CI 0.65–1.12). Conclusions: In this cohort study, patients with COVID-19 admitted to the ICU in the second wave were younger and had better prognostic scores. Adjusted survival was similar in the two waves, contrasting with record number of hospitalizations, daily deaths and health system collapse seen across the country in the second wave. Our findings suggest that the combination of the burden of severe cases and factors such as resource allocation and health disparities may have had an impact in the excess mortality found in many countries in the second wave.

The Coronavirus Disease 2019 (COVID-19) pandemic was the most severe public health crisis of the century. By July 2023, Brazil was the sixth country in number of confirmed cases in the world ranking, with more than 37 million cases [[1]], and second country in cumulative deaths, with more than 700,000 deaths registered [[1]]. Among critically ill patients with severe acute respiratory infection (SARI) and COVID-19-related acute respiratory distress syndrome (ARDS), mortality in the first wave was high [[2]], especially among patients who required mechanical ventilation [[3]].

As countries were recovering from the toll taken by the first COVID-19 wave, the rapid worldwide spread of SARS-CoV-2 contributed to the emergence of new genetic lineages, called "variants of concern" (VOCs) [[4]] which showed increased transmissibility [[4]–[6]], escaped immunity generated by vaccine or natural infection [[7]–[9]], and raised concerns about increased disease severity and mortality [[10]–[12]]. The VOCs led to the worsening of the world health scenario with subsequent waves of cases and deaths. In Brazil, the gamma variant, which originated in the Amazon region, was the driver of the second wave, which was evident by January 2021 after a period of 3 months of decreasing number of cases [[1], [13]]. In March 2021, the Brazilian government issued a warning about the presence of Gamma (P.1), Alpha (B.1.1.7) and Beta (B.1.351) variants in Brazil [[14]], with a proportion of the Gamma variant reaching approximately 90% of the SARS-CoV-2 lineages circulating in Brazil during this period [[5]].

During the second wave, the Brazilian health system collapsed with rapidly rising number of cases, record deaths [[11], [15]], ICU occupancy above 90% in several large cities [[11], [16]], similarly to what was seen in other parts of the world [[18]–[21]]. The causes for excessive deaths, especially among the younger patients, were not clear. One initial hypothesis was that the Gamma variant could be more virulent, causing more serious illness, including among younger patients [[22]–[25]]. Other possible explanations were lack of access to vaccination, which was restricted to older and the most vulnerable populations [[26]], strain on the health system due to increased transmissibility, limited access to intensive care beds and failure to implement public health interventions to contain community transmission [[11], [28]].

Currently, there are limited data comparing outcomes of critically ill patients across different waves of COVID-19 focusing on the emergence of VOCs [[9], [12], [18]]. More data are needed to clarify the interplay between different VOCs, disease severity and availability of health care resources and its impact on mortality. Therefore, we designed a cohort study to compare characteristics, clinical management, and outcomes of critically ill patients hospitalized in the first and second waves of COVID-19 in a large academic hospital in Brazil. We hypothesized that in-hospital mortality would be greater in the second wave.

Methods

Study design and location

This was a retrospective cohort conducted at Hospital das Clínicas, a public hospital affiliated with the University of Sao Paulo Medical School, and the largest academic hospital in Brazil. The hospital is public, and patients are treated at no cost in accordance with the Brazilian universal health system. We compared admission characteristics and hospital outcomes of patients admitted to ICUs dedicated exclusively to the care of COVID-19 in the first and second waves of the pandemic. During the first wave, in 2020, there were 20 COVID-19 ICUs with a total of 300 beds, of which 206 beds were operating rooms or ward beds converted into ICU beds. At that time, our hospital complex was the most important referral center for COVID-19 patients from the metropolitan region of São Paulo comprising a population comparable to countries the size of Portugal. In 2021, during the second wave, referral of severe COVID-19 cases was more organized and less concentrated in our hospital, in which five ICUs remained dedicated to COVID-19, with 58 beds.

The Research Ethics Committee approved the study (number CAAE: 50340521.6.0000.0068). Requirement for informed consent form was waived because of the observational nature of the study.

Study population

From March 30 until June 30, 2020 (first wave) and April 1st until July 31, 2021 (second wave), all consecutive patients admitted to the ICU with COVID-19 were screened. ICU admission was regulated by a team of physicians and nurses who received ICU admission requests from the emergency rooms and wards, and outside hospitals. Inclusion criteria were age ≥18 years and confirmed COVID-19 with Reverse Transcription Polymerase Chain Reaction (RT-PCR) or the rapid-antigen test or antibody (serology) test for IgM and IgG (more details in S1 Text). Exclusion criteria were palliative care in the first 24 hours after ICU admission, less than 24 hours of ICU stay, admission to the ICU after more than 7 days of invasive ventilatory support in another health institution or more than 14 days of hospitalization. We also excluded patients who were admitted to ICU due to other diagnosis and without symptoms of COVID-19, but who had SARS-CoV-2 RT-PCR positive on admission screening, defined as asymptomatic COVID-19.

Patient care

Patient care was at the ICU teams' discretion. The hospital developed institutional protocols specifically for COVID-19 patients, since the beginning of the pandemic, including the best evidence that emerged related to the care of critically ill patients with COVID-19. The main treatment protocol changes between 2020 and 2021 were use of corticosteroid in hypoxemic patients, use of prophylactic doses of anticoagulants to patients without evidence of thromboembolism, use of antibiotics restricted to patients with suspected secondary bacterial infection [[29]] and reduced restrictions for the use of noninvasive ventilation (NIV) and high-flow nasal cannula (HFNC).

Outcomes

The main outcome was 60-day hospital survival. The secondary outcomes were use of protective ventilation, rescue therapies for refractory hypoxemia, need for renal replacement therapy (RRT) and/ or vasopressors.

Data collection

Patient information was collected from electronic medical records. Data were accessed from August 19, 2021, to November 12, 2021. We used an online case report form, managed on REDCap-Research Electronic Data Capture, an online platform, [[30]]. Confidentiality of the information was ensured by restricted access to the redcap database to the two principal investigators and use of deanonymized databases on export. We followed patients until hospital discharge or transfer to other health services.

We collected data related to ICU admission, management in the first 24 hours, and outcomes. These data included demographic information, comorbidities, duration of symptoms, laboratory tests, prognostic scores such as Simplified Acute Physiology Score III (SAPS 3) [[31]] and Sequential (Sepsis-related) Organ Failure Assessment (SOFA) [[32]], specific COVID-19 treatment, need for advanced life support therapies, and ventilator parameters. We defined protective ventilation as use of tidal volume < 8 mL/Kg and plateau pressure < 30 cmH2O. We also collected ICU and hospital outcomes and clinical complications reported during the course of illness.

Information about vaccination against SARS-CoV-2 was collected from electronic medical records or telephone call to the patient's next of kin. In early 2021, vaccines approved for use in Brazil were those produced by Sinovac/Butantan Institute, AstraZeneca/Fiocruz, Pfizer/Wyeth and Janssen [[27]]. In São Paulo, vaccination started in mid-January 2021 for healthcare workers and was gradually expanded with priority for the most vulnerable population, the elderly and immunosuppressed. By April 2021, residents over 63 years-old had been offered their first shot and by July 2021, the cutoff was at 28 years-old [[27]].

Nasal swab and tracheal aspirate samples from patients admitted in the second wave were tested for detection of SARS-CoV-2 variants, using the QuantStudio™ 5 Real-Time PCR System (Applied Biosystem, Foster City, California, USA). Variant testing could not be performed for patients when samples were unavailable, either because RT-PCR was done in another health service, or because COVID-19 diagnosis was made using antigen or serologic tests (more details in S1 Text). Variant testing was not performed on samples from patients admitted in the first wave because at the time, there were no VOCs circulating in Brazil.

We report the results according to the recommendations the Strengthening The Reporting of Observational Studies in Epidemiology (STROBE) guidelines [[33]].

Statistical analysis

Continuous variables were expressed as mean and standard deviation (SD) or median and interquartile range (IQR) as appropriate and compared using the independent samples Student's t-test or Mann‐Whitney U test. Categorical variables were presented as absolute and relative frequencies and compared using the Chi-square test.

We plotted Kaplan–Meier curves to estimate 60-day survival in each of the pandemic waves. 60-day survival was defined as the time interval between ICU admission and patient death from any cause or hospital discharge. Patients discharged home or transferred to other health services were considered alive at the end of follow-up. In the unadjusted model, a log rank test was used to compare the survival of patients in the two pandemic waves. In addition, Cox proportional hazard models were used to compare the survival of patients in the two pandemic waves, both without adjustments, and adjusting for potential confounders. We built the multivariable Cox model based on a conceptual causal diagram in the format of directed acyclic graph (DAG) [[34]], including the association between admission in the second wave and survival, adjusted by the most relevant variables. Variables in the DAG conceptual model were selected based on prior knowledge and are depicted in S1 Fig. A sensitivity analysis using an alternative DAG model not including vaccination was also performed (S2 Fig). We tested the proportional hazards assumption in Cox models with the Schoenfeld residuals method. The multivariable model did not include interaction terms or any other higher order terms.

Unadjusted and adjusted hazard ratios (HR) and 95% confidence intervals (95%CI) were used to measure the association between each variable and 60-day survival. All hypothesis tests are two-tailed and p-value < 0.05 was considered statistically significant.

The analyses were performed using the statistical software R (R Foundation for Statistical Computing Platform, version 4.2.1) [[35]–[38]].

Results

Study population

We screened 1,955 patients admitted to the ICU during the study periods, and included 1,583 patients, of whom 1,315 were admitted in the first and 268 were admitted in the second wave (Fig 1).

Graph: Flow of potentially eligible participants in the study, and final numbers included and analyzed. ICU: Intensive Care Unit; COVID-19: Coronavirus Disease 2019.

Follow-up was complete for all patients until hospital discharge, death in the hospital or transfer to other health services. In 2020, 104 (8%) patients were transferred to a long-term care facility, and in 2021 only 4 (1.5%) patients were transferred.

Baseline characteristics

The baseline characteristics of the subjects, stratified by year of admission are shown in Table 1. In the first wave, patients were older, 742 (56%) were ≥60 years old, than in the second wave, when 114 (43%) were ≥60 years old. Male sex was predominant in both periods, 795 (61%) in the first wave and 165 (62%) in the second wave. At ICU admission, patients in the first wave were more likely to be receiving mechanical ventilation (61% vs 45%, p<0.001) and vasopressors (39% vs 30%, p = 0.005), had more comorbidities, and the mean SAPS3 (64 ± 16 vs 56 ± 13, p<0.001) and the median SOFA 7 [[3]–[10]] vs 4 [[2]–[7]], p<0.001) were higher. Laboratory findings are shown in S1 Table.

Graph

Table 1 Baseline characteristics at ICU admission.

	First wave (n = 1315)	Second wave (n = 268)	p value
Characteristics
Age (y), mean (SD)	61 ± 15	56 ± 14	<0.001
Age groups (y), n (%)			<0.001
18–40	138 (11)	46 (17)
41–60	435 (33)	108 (40)
>60	742 (56)	114 (43)
Female sex, n (%)	520 (39)	103 (38)	0.787
BMI	28 ± 7	30 ± 6	<0.001
SAPS 3	64 ± 16	56 ± 13	<0.001
SOFA, median [IQR]	7 [3 – 10]	4 [2 – 7]	<0.001
Duration of symptoms (d), median [IQR]	9 [6 – 12]	12 [9 – 14]	<0.001
Vasopressors, n (%)	515 (39)	80 (30)	0.005
Invasive mechanical ventilation, n (%)	809 (61.5)	122 (45.5)	<0.001
Corticosteroids, n (%)	146 (11)	207 (77)	<0.001
Vaccination, n (%)	0 (0.0)	77 (29)	-
1 dose, n (%)	0 (0.0)	52 (20)
2 doses, n (%)	0 (0.0)	25 (9)
SARS-CoV-2 variant screening, n (%)	0 (0.0)	67 (25)	-
Gamma (P1), n (%)	0 (0.0)	63 (94)
Alpha (B.1.1.7), n (%)	0 (0.0)	1 (1.5)
Indeterminate, n (%)	0 (0.0)	3 (4.5)
Racea, n (%)			<0.001
White	795 (60)	201 (75)
Black	94 (7)	8 (3)
Mix‐ethnicity (Pardo)	364 (28)	47 (18)
Asian	15 (1)	3 (1)
Not informed	47 (4)	9 (3)
Comorbidities, n (%)
Asthma	36 (3)	13 (5)	0.104
Cancer	129 (10)	10 (4)	0.002
Cardiovascular disease	194 (15)	54 (20)	0.034
Cerebrovascular disease	54 (4)	5 (2)	0.112
Chronic kidney disease	129 (10)	7 (3)	<0.001
Chronic pulmonary disease	77 (6)	17 (6)	0.868
Diabetes	501 (38)	79 (29)	0.009
Hypertension	751 (57)	122 (46)	0.001
Obesity	423 (35)	136 (51)	<0.001
HIV/AIDS	15 (1)	1 (0.4)	0.418

1 BMI: body mass index, kg/m2; IQR: interquartile range; SAPS 3: Simplified acute Physiology Score 3; SOFA: Sepsis-related Organ Failure Assessment. Data are presented as mean and standard deviation, unless otherwise stated; comparisons were made with t test, Mann–Whitney U test or Chi-square test as appropriate.

• 2 Missing year 2020: BMI for 106 (8%) patients; SAPS3, missing for 1 patient. Missing year 2021: BMI for 1 patient, SARS-CoV-2 variant screening for 201 (75%) patients in 2021.
• 3 a The categories represent the Brazilian official race categories.

Vaccination and variants of concern

In the second wave, nearly a third of patients were vaccinated with at least one dose of the COVID-19 vaccine. 67 (25%) patients in the second wave had samples available for SARS-CoV-2 mutations testing, 94% of whom were found to be infected with the Gamma variant (Table 1).

Ventilatory management in the first 24 hours after ICU admission

More patients required mechanical ventilation in the first 24 hours of their ICU stay in the first wave compared with the second, 878 (67%) vs 148 (55%), p<0.001, as shown in Table 2. The use of protective ventilation was common and similar in both waves (82% in the first wave vs 88% in the second wave, p = 0.08).

Graph

Table 2 Ventilatory management on the first 24 h after ICU admission.

Management	First wave (n = 878)	Second wave (n = 148)	p value
Tidal volume (mL/Kg ibw), mean (SD)	6.55 ± 1.3	6.17 ± 1.2	0.001
Respiratory Rate, median [IQR]	30 [26 - 35]	30 [25 - 31]	0.015
Minute volume, mean (SD)	12 ± 3.7	11 ± 2.7	<0.001
FiO2 (%), median [IQR]	50 [40 - 60]	50 [40 - 70]	0.178
PEEP (cmH2O), median [IQR]	10 [8 - 12]	10 [8 - 14]	<0.001
Plateau pressure (cmH2O), mean (SD)	22.6 ± 4.7	23.9 ± 4.3	0.001
Driving pressure (cmH2O), mean (SD)	12.6 ± 4	12.8 ± 3	0.686
Compliance (mLcmH2O−1), median [IQR]	32 [24 - 41]	30 [24 - 37]	0.070
Compliance (mLcmH2O−1.Kg−1ibw), median [IQR]	0.52 [0.41–0.65]	0.48 [0.39–0.62]	0.022
PaO2/FIO2 (%), mean (SD)	168 ± 70	164 ± 72	0.521
Arterial pH, median [IQR]	7.36 [7.30 - 7.42]	7.36 [7.29 - 7.42]	0.986
Arterial PaCO2 (mmHg), median [IQR]	42.3 [37.9 - 48.3]	47.00 [39.0 - 56.1]	<0.001
Arterial O2 saturation (%), median [IQR]	93 [91 - 96]	94 [91 - 96]	0.687
Ventilation Mode, n (%)			<0.001
Volume-controlled ventilation	503 (57)	94 (64)
Pressure-controlled ventilation	167 (19)	28 (19)
Pressure support ventilation	199 (23)	15 (10)
Other	9 (1)	11 (7)
Rescue therapy for hypoxemia, n (%)
Prone position	145 (16)	61 (41)	<0.001
PEEP titration	103 (12)	57 (38)	<0.001
Recruitment maneuvers	14 (1.6)	12 (8.1)	<0.001
Extracorporeal membrane oxygenation	1 (0.1)	2 (1.4)	0.079
Inhaled nitric oxide	1 (0.1)	0 (0.0)	-
Protective ventilation, n (%)	634 (82)	130 (88)	0.081

• 4 SD: standard deviation; IQR: interquartile range; O2: oxygen; PaCO2: arterial partial pressure of carbon dioxide; FIO2: inspired fraction of oxygen; PEEP: positive end-expiratory pressure; ibw: ideal body weight; PaO2/ FIO2: partial pressure of arterial oxygen and fraction of inspired oxygen ratio. Data are n. (%), unless otherwise stated; comparisons were made with t-test, Mann–Whitney U test or Chi-square test as appropriate.
• 5 Missing year 2020: Tidal volume (mL/Kg ibw) for 15 (1.7%) patients; PaO2/FIO2 was missing for 4 (0.5%) patients; plateau pressure (cmH2O) and driving pressure (cmH2O) were missing for 95 (11%) patients; PaCO2 for 58 (7%) patients; arterial pH for 52 (6%) patients and arterial O2 saturation for 61 (7%) patients. Missing year 2021: none.

Clinical management of patients in the first 24 hours after ICU admission

Differences in the clinical management of patients in the first 24 hours after ICU admission between waves are shown in S2 Table. In the first wave, the percentage of patients under sedation was higher than in the second wave (63% vs 55%, p<0.001), but use of neuromuscular blockade was more common on the second wave (27% in the first wave vs 39% in the second wave, p<0.001). In the second wave, the use of antibiotics was less common (69% vs 42%, p<0.001), and therapeutic anticoagulation was more frequently used in the second wave (9% vs 19%, p<0.001). Systemic corticosteroids were used for 98% of patients in the second wave, and only for 25% of patients in the first wave.

ICU and hospital outcomes

Relevant clinical outcomes are shown in Table 3. The median ICU stay was 11 [[6]–[19]] days in the first wave and 10 (7–19) in the second wave(p = 0.445). Hospital stay was longer, 21 days [[14]–[32]], on the second wave compared to the first wave, 17 days [[11]–[27]], p<0.001. More patients in first wave required mechanical ventilation (80% vs 70%, p<0.001), while the use of NIV and HFNC were more frequent in the second wave, (22% vs 46%, p<0.001) and (10% vs 41%, p<0.001), respectively. The duration of mechanical ventilation was similar for the two waves, 10 [[6]–[17]] vs 10 [[6]–[19]], p = 0.458. Prone position was more commonly used in the second wave (38% vs 57%, p<0.001), while vasopressors and RRT were required more often in the first wave (74% vs 60%, p<0.001).

Graph

Table 3 Clinical outcomes.

Outcomes	First wave (n = 1315)	Second wave (n = 268)	p value
ICU length of stay, median [IQR], d	11 [6 - 19]	10 [7- 19]	0.445
Hospital length of stay, median [IQR], d	17 [11 - 27]	21 [14 - 32]	<0.001
Invasive mechanical ventilation, n (%)	1051 (80)	187 (70)	<0.001
Duration of mechanical ventilation, median [IQR], d	10 [6 - 17]	10 [6 - 19]	0.458
Reintubation, n (%)	206 (15.7)	17 (6.3)	<0.001
Prone positioning, n (%)	401 (38)	107 (57)	<0.001
Noninvasive ventilationa, n (%)	291 (22)	123 (46)	<0.001
High‐flow nasal cannulaa, n (%)	128 (10)	109 (41)	<0.001
Extracorporeal membrane oxygenation, n (%)	6 (0.5)	5 (1.9)	0.033
Vasopressors, n (%)	976 (74)	162 (60)	<0.001
Renal replacement therapy, n (%)	481 (37)	58 (22)	<0.001
Tracheostomy, n (%)	160 (12)	32 (12)	0.999
Delirium, n (%)	430 (33)	28 (10)	<0.001
Ventilator‐associated pneumonia, n (%)	358 (27)	79 (29)	0.479
Thromboembolic event, n (%)	246 (19)	81 (30)	<0.001
Cardiac arrhythmia, n (%)	223 (17)	43 (16)	0.769
Treatment withhold or withdraw during hospital stay, n (%)	249 (19)	22 (8)	<0.001
Mortality at 28 days, n (%)	580 (44)	87 (32)	<0.001
Mortality at 60 days, n (%)	645 (49)	97 (36)	<0.001
ICU outcome, n (%)			<0.001
Discharged home	22 (1.7)	10 (3.7)
Discharged to the ward	641 (49)	162 (60)
Transferred to another ICU	50 (3.8)	6 (2.2)
Transfer to long-term care facility	1 (0.1)	1 (0.4)
Death	601 (46)	89 (33)
Hospital outcome, n (%)			<0.001
Discharged home	553 (42)	162 (60)
Transfer to long-term care facility	104 (8)	4 (1.5)
Death	658 (50)	102 (38)

• 6 ICU: intensive care unit; IQR: interquartile range; Data are n. (%), unless otherwise stated; comparisons were made with t test, Mann–Whitney U tests or chi-square test as appropriate.
• 7 a To avoid intubation or prior to intubation.

At the end of 60-days follow-up, 645 (49%) patients died in the first wave and 97 (36%) patients died in the second wave. After 60 days, another 13 (1%) patients in the first wave and 5 (2%) in the second wave died in the hospital.

Association of baseline characteristics with survival

Admission in the second wave was associated with higher survival at 60 days in the unadjusted analysis (logrank, p<0.001 and HR 0.61, 95%CI 0.49–0.76), as shown in Fig 2A, and in S3 Table. After adjusting for age, SAPS 3 and vaccination, according to our conceptual causal diagram, admission in the second wave was no longer associated with survival (aHR 0.85, 95%CI 0.65–1.12). In this multivariable model, the only variables independently associated with 60-day hospital survival were age (aHR 1.02, 95%CI 1.02–1.03) and SAPS 3 (aHR 1.03, 95%CI 1.03–1.04), shown in in Fig 2B, and in S3 Table. In the sensitivity analysis excluding vaccination from the DAG model, we found similar results, with admission not associated with survival (aHR 0.89, 95%CI 0.72–1.12), as shown in S4 Table.

Graph: Solid red line represents survival of patients who were admitted in 2020 (first wave) and solid blue line represents survival of patients who were admitted in 2021 (second wave). (A) Unadjusted survival, logrank p<0.001 and HR 0.61, 95%CI 0.50–0.76. (B) Survival adjusted by age, SAPS 3 and vacination (aHR 0.85, 95%CI 0.65–1.12). The p values were obtained with Cox proportional hazards models.

Discussion

In this retrospective cohort, we compare characteristics and outcomes of 1,583 patients with COVID-19 admitted to the ICUs of a referral center for COVID‐19 in Brazil in the first two waves of the pandemic. We found that 60-day survival was higher in the second wave compared to the first wave. However, in the multivariable analysis, adjusting by SAPS3, age and vaccination, admission to the ICU in the second wave was no longer associated with mortality. Patients admitted to the ICU in the second wave were younger, had better prognostic scores, and less need for vasopressors and RRT. Invasive mechanical ventilation was needed by most patients in both waves, 80% in the first and 70% in the second wave, and noninvasive ventilatory support and prone increased substantially in the second wave. Duration of mechanical ventilation and ICU length of stay were similar in the two waves, while hospital length of stay was longer in the second wave.

Survival at 60 days in our hospital was higher in the second wave than in the first wave, contrary to our study hypothesis, but in line with what was observed in studies from other countries [[39]–[43]]. Possible reasons for this difference include the admission of younger and less severe patients, earlier hospitalization, more organized and efficient ICU care, with better trained health care professionals, and more knowledge about the disease with better treatment protocols in the second wave. The impact of younger age and less severe disease at ICU admission on survival is supported by the finding that survival was similar in both waves when adjusted for age and SAPS3.

Our results contrast with the catastrophic outcomes seen in Brazil during the second wave, one the most affected countries in the world. By February 2021, the rapid spread of the Gamma variant was associated with a steep increase in number of cases [[15]], accompanied by a rise in the number of cases of SARI [[44]]. This combination of factors led to record hospitalizations and deaths [[15]] with signs of health system collapse [[11], [15]], resulting in worse outcomes and potentially preventable deaths [[45]].

Access to ICU beds and availability of resources are very discrepant across Brazil, and such disparities increased during the pandemic [[45]]. Availability of ventilators was up to 5 times greater in southern and southeastern capitals compared to northern and northeastern capitals [[45]], and age-standardized mortality rates varied by up to 5 times between the northern and southern states in the second wave [[44], [47]]. Our hospital is university-affiliated, located in the richest state of the country, with capability to increase ICU capacity and resources according to demand, and a centralized referral system which limited the number of patients admitted each day, in a striking contrast with frontline hospitals in the Amazon region, for example.

In Brazil and other countries, the second wave had an increased proportion of younger and previously healthy patients who required ICU care [[15], [39]–[41]]. This reduction in the average age could be related to the fact that vaccination coverage started in January 2021 in Brazil and was associated with a reduction in hospitalizations and deaths in adults older than 70 years [[26]], but was not available to younger populations at that time. Another possible explanation to the high burden of cases among younger people in Brazil was a greater exposure to the virus during the second wave, when fewer restrictive measures were in effect, adherence to the use of masks decreased, and part of the economically active population had returned to work [[28]].

Only 25 (9%) of patients of the second wave had completed two doses of vaccination when they were admitted to the ICU. This low number reflects the vaccination calendar in Brazil, since vaccination started in January 2021 and was scaled by age, starting with very old citizens [[27]]. As a result, despite documented protection against COVID-19 [[26]], vaccination probably had very little impact for patients of the second wave included in our study.

Patients in the second wave were less likely to need vasopressors and RRT during their ICU stay than patients in the first wave, as noted in other studies [[39], [48]]. These results support the finding that better clinical outcomes in the second wave could be associated with younger age and less severe disease at ICU admission.

During the second wave, we found that 94% of the samples analyzed detected the Gamma variant, in accordance to epidemiological findings in Brazil during this period [[5]]. The Gamma variant spread throughout Brazil and other countries and was found to be more transmissible [[4]–[6]], having a higher risk of hospitalization [[44], [49]]. While some studies found that it caused more severe disease and increased the risk of death [[44], [46], [50]], others did not see an increased mortality [[48], [51]]. It is possible that the elevated daily death counts observed during the second wave in Brazil was due to a combination of increased risk of development of SARI [[44]] and lack or delayed access to healthcare and strain on the healthcare system. Future comparisons with other cohorts of patient may bring addition light to the role of VOCs on severe COVID-19.

Management of patients in the second wave was influenced by accumulated knowledge about the efficacy of several therapeutic options. Importantly, less use of antibiotics [[53]], and more use of corticosteroids were informed by randomized controlled trials [[54]]. In our cohort, only 25% of the patients received corticosteroid in the first wave, while 98% received it in the second wave, which may have had an impact in survival.

The use of noninvasive ventilatory support increased during the second wave, likely due to greater availability, more experience, and less concern about and environmental contamination [[56]]. Use of noninvasive support may have contributed to avoiding intubations in the second wave, in which we found 10% less use of invasive mechanical ventilation compared to the first wave. Currently, there is evidence that noninvasive ventilatory support is effective in decreasing intubation rates, but not mortality, compared to conventional oxygen in COVID-19 [[57]].

Among intubated patients, we found that the more than 80% of patients received protective ventilation in both waves, which our group previously found to be associated with increased survival in COVID-19 [[2]]. Rescue therapies, including recruiting maneuvers, PEEP titration, and extracorporeal membrane oxygenation were more often used in the second wave. The same observation was seen with prone position, which was used in 57% of the patients in the second wave, compared to 38% in the first wave. Prone position was increasingly used during the COVID-19 pandemic, probably based on better training of health professionals, greater knowledge in the management of patients with COVID-19 pneumonia, and previous evidence of benefit of prone in patients with moderate to severe non-COIVD-19 ARDS [[59]].

Our study has several limitations. Being a retrospective observational study, data were collected from the electronic medical record, which could have incomplete or inaccurate data. It is a single-center study, and our findings may not be generalizable to other hospitals in Brazil and other low-and-middle-income countries. However, it is our belief that this does not have a significant impact on the inferential conclusions related to the comparison between COVID-19 waves. Only 25% of respiratory samples from patients in the second wave were available for variant testing, because many patients were diagnosed in other health services before being transferred to our hospital or diagnosed with antigen tests, limiting our power to estimate the association of variants of concern with clinical outcomes. We compared the first two waves of COVID-19 in Brazil, which happened in different contexts of hospital organization. During the first wave, our hospital had five times more ICU beds compared to the second wave, reflecting greater strain. Our study was unable to measure the impact of updated treatment protocols, better structuring of ICUs and training of health care professionals in our hospital that occurred in the second wave. We included vaccination to our DAG conceptual model, but since vaccination was not available for patients in the first wave, or for younger patients in the second wave, its inclusion might bias our results and impact the interpretation of the results. In order to address this issue, we performed a sensitivity analysis removing vaccination from the model and found similar results. The sample of patients admitted in the second wave was smaller, which limited our statistical power. Finally, since our study only included ICU patients, we could not measure the impact of the Gamma variant on increased risk of development of SARI, which in turn may have been responsible for overall increased risk of death in the second wave. Such limitations may impact the generalizability of our results and the interpretation of the results, particularly the significance of lower mortality observed in our hospital during the second wave, when mortality was higher in official country-wide data.

Our study also has strengths. The hospital covers a large geographical area, with an estimated population of 23 million; we screened all consecutive patients with confirmed COVID-19 admitted to the participating ICUs, minimizing selection bias; we included a large sample size; we had complete follow-up of patients until hospital discharge or transfer to another health service; we had few missing data; and, we assessed objective, hard outcomes. In addition, our hospital followed evidence-based clinical management protocols, which were reviewed as evidence became available. These results corroborate the importance of implementing up to date clinical practice guidelines during health emergencies to provide rational allocation of resources and improve patient outcomes.

Conclusion

In this cohort study conducted across multiple ICUs of the largest academic hospital in Brazil, we observed that patients admitted to the ICU in the second wave of the COVID-19 pandemic were younger and had better prognostic scores compared to those in the first wave. They also required fewer advanced life support therapies and had higher survival rates. Survival rates adjusted for age and severity score were found to be similar between waves. These results contrast with the country-level outcomes and underscore the impact of disease severity at ICU admission, as well as the availability and rational allocation of healthcare resources, health-system strain, and health disparities on the outcome of the pandemic. As the world starts to recover from the impacts of the COVID-19 pandemic, other epidemics continue to threaten populations and healthcare systems. Our results shed light on the interplay between infectious agents' virulence, patient vulnerabilities and resource availability and can inform preparedness strategies to respond to epidemics and provide equitable care to all.

Supporting information

S1 Text

Supplementary methods.

(DOCX)

S1 Fig

Causal diagram in the format of directed acyclic graph (DAG).

SAPS 3: Simplified acute Physiology Score 3. This conceptual model shows clinically relevant variables associated with survival. Arrows indicate a presumed direct causal effect of one variable on another variable. Admission in the second wave is the main predictor, shown in green; variables associated with the outcome, but not associated with the main predictor, are shown in blue; variables associated with both the outcome and the main predictor are shown in red (arrows indicate a suspected direct causal effect of that variable on both the main predictor and the outcome). A multivariate analysis for estimating the direct effect of admission in the second wave on survival should be adjusted for potential confounders, identified in the model as age, SAPS3 and vaccination.(TIF)

S2 Fig

Alternative causal diagram in the format of directed acyclic graph (DAG) used for a sensitivity analysis.

SAPS 3: Simplified acute Physiology Score 3. This conceptual model, used as a sensitivity analysis, shows clinically relevant variables associated with survival, not including vaccination. Arrows indicate a presumed direct causal effect of one variable on another variable. Admission in the second wave is the main predictor, shown in green; variables associated with the outcome, but not associated with the main predictor, are shown in blue; variables associated with both the outcome and the main predictor are shown in red (arrows indicate a suspected direct causal effect of that variable on both the main predictor and the outcome). A multivariate analysis for estimating the direct effect of admission in the second wave on survival should be adjusted for potential confounders, identified in the model as age and SAPS3.(TIF)

S1 Table

Laboratory tests at ICU admission.

Data are presented as median [IQR]: interquartile range; comparisons were made with Mann-Whitney test. Missing year 2020: Arterial lactate for 399 (30%) patients; D-dimer for 316 (24%) patients; Arterial pH for 151 (12%) patients; C-reactive protein for 202 (15%) patients. Missing year 2021: Arterial lactate for 65 (24%) patients; D-dimer for 12 (5%) patients; Arterial pH for 45 (17%) patients; C-reactive protein for 10 (4%) patients.(DOCX)

S2 Table

Patient management on the first 24 h after ICU admission.

Definition of abbreviations: O2: oxygen; RASS: Richmond Agitation-Sedation Scale; COVID-19: Coronavirus Disease 2019. Data are n. (%); comparisons were made with the chi-square test.(DOCX)

S3 Table

Association between admission in the second wave, other relevant covariates, and 60-day survival.

HR: hazard ratio; aHR: adjusted hazard ratio; 95%CI: 95% confidence interval; SAPS 3: Simplified acute Physiology Score 3; HR, aHR obtained with univariate and multivariate Cox models, respectively, and 95%CI and p values obtained in each model. SAPS 3 was missing for 1 patient.(DOCX)

S4 Table

Sensitivity analysis—association between admission in the second wave, other relevant covariates, and 60-day survival.

HR: hazard ratio; aHR: adjusted hazard ratio; 95%CI: 95% confidence interval; SAPS 3: Simplified acute Physiology Score 3; HR, aHR obtained with univariate and multivariate Cox models, respectively, and 95%CI and p values obtained in each model. SAPS 3 was missing for 1 patient.(DOCX)

Decision Letter 0

Ngah Veranyuy Academic Editor

21 Nov 2023

PONE-D-23-21314TEMPORAL TRENDS OF SEVERITY AND OUTCOMES OF CRITICALLY ILL PATIENTS WITH COVID-19 AFTER THE EMERGENCE OF VARIANTS OF CONCERN: A COMPARISON OF TWO COHORTSPLOS ONE

Dear Dr. Freitas,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE's publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Jan 05 2024 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Veranyuy Ngah, MSc

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at 

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne%5fformatting%5fsample%5fmain%5fbody.pdf and 

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Competing Interests section: 

"Dr. Ferreira reports personal fees from Medtronic, outside the submitted work; Dr. Costa reports personal fees from Timpel, personal fees from Magnamed, outside the submitted work; Dr. Ho reports personal fees from Pan‐American Health Organization, outside the submitted work. The other authors have no conflict of interest to disclose."

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: ""This does not alter our adherence to  PLOS ONE policies on sharing data and materials." (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared. 

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

3. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study's minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

• 4. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.
• 5. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information. 

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

***

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

***

3. Have the authors made all data underlying the findings in their manuscript fully available?

The http://www.plosone.org/static/policies.action#sharing requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

***

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

***

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Summary

Thank you for the opportunity to provide a peer review for this interesting article on "Temporal trends of severity and outcomes of critically ill patients with COVID-19 after the emergence of variants of concern". The article compares critically ill patients' outcomes between two cohorts in the first two waves of COVID-19, comprising a sample of 1583 patients (1315 in the first and 268 in the second wave) from the largest academic hospital in Brazil. Results indicated that in the second wave, admitted COVID-19 ICU patients were younger and had better prognostic scores with higher survival rates compared to those in the first wave. However, the survival rates in the model adjusted for age and severity score were similar between waves; therefore, the authors conclude that factors such as resource allocation and health disparities could have impacted the excess mortality in many countries in the second wave.

Overall Impression

The study addresses an important research question and presents valuable insights into the outcomes of critically ill COVID-19 patients during the first two waves of the pandemic in Brazil. As is typically the case with the peer review process, I believe there are areas where the manuscript could be strengthened. With some revisions and additional contextual information, this manuscript has the potential to make a significant contribution to the field of pandemic healthcare management. To forecast the essence of my review, these recommendations can be interpreted as minor-to-major in nature. Although the effort required to address these recommendations is minimal, the flow-on effects on manuscript quality will be excellent. Please feel free to address these points in your revision, and don't hesitate to reach out if you need further clarification or assistance. Thank you for your contribution to the scientific community.

Specific Comments:

Abstract

Consider providing a brief context for readers unfamiliar with the Brazilian healthcare system and the significance of studying outcomes across different waves of the pandemic.

Introduction

Provide references to the claim made on P5 L101.

Methods

I like how the authors provided a very good description of the study population, highlighting the selection criteria. However, additional information is needed to clarify the specific criteria that were used to define critically ill patients. Detailing parameters like respiratory rate, oxygen saturation, and comorbidities would be helpful for readers to gauge the severity of the cases.

Kindly add references to the Patient Care subsection P7 L143.

On P7 L143, the authors mention the main changes in treatment protocol between 2020 and 2021. Can they also comment on the use of anticoagulants at prophylactic doses to prevent thromboembolism and the nonuse of antibiotics in patients without suspected bacterial infection as reported by Falavigna et al (https://www.jbmede.com.br/index.php/jbme/article/download/69/50).

In the Statistical Analysis subsection P8 L180, several essential details are missing that are crucial for a comprehensive understanding of the statistical analysis. The description mentions that Kaplan-Meier curves were plotted for each of the pandemic waves. It's important to clarify if these curves were plotted separately for each group (first wave vs. second wave) or if there was a comparison between the curves. If a comparison was made, the results of the log-rank test, which is commonly used to compare Kaplan-Meier curves between groups, should be reported.

Kindly explain the rationale for choosing the variables to be included in the multivariable Cox model-building process. There is no mention of how assumptions such as the proportional hazards assumption in Cox models were checked. It's vital to confirm that these assumptions hold for the validity of the analysis.

Information on how the goodness-of-fit of the Cox proportional hazards model was assessed is missing. Techniques like likelihood ratio tests, Akaike Information Criterion (AIC), or Bayesian Information Criterion (BIC) are commonly used to assess model fit.

Details about specific packages or functions within R used for the analyses are important for transparency and reproducibility.

Results

I liked the clear presentation of the patients' characteristics from both waves in Table 1, including demographic information, comorbidities, and other relevant factors summarizing the key differences. This will provide a comprehensive understanding of the patient population studied.

In the Baseline Characteristics subsection P10 L212-219, the authors need to provide specific numerical data from Table 1 to support their statements about the subjects. Instead of using phrases like "male sex was predominant" or "more likely," the authors should directly quote the figures from the table to enhance the precision and clarity of the description. For instance:

Instead of saying "male sex was predominant in both periods," specify the exact percentages of male patients in both the first and second waves.

Instead of saying "56% were ≥60 years old" and "43% were ≥60 years old," provide the actual numbers of patients in each age group (percentage in brackets) for both waves.

Instead of saying, "Patients in the first wave were more likely to be receiving mechanical ventilation and vasopressors," provide the specific percentages or counts of patients receiving mechanical ventilation and vasopressors in each wave.

Provide the actual numerical values for mean SAPS3 and SOFA scores in both waves to convey a more precise comparison.

By directly citing the figures from Table 1, readers can have a clear and accurate understanding of the baseline characteristics of the subjects in each wave without having to move back and forth to Table 1.

The same applies to the ventilatory management in the first 24 hours after ICU admission (P12 L225-232) and management in the first 24 hours after ICU admission (P13 L240-247) subsections. See Lalla et al as an example (https://pubmed.ncbi.nlm.nih.gov/35359698/).

On P13 L240, the heading seems a bit vague; the authors can consider changing it to Clinical Management of Patients in the First 24 Hours After ICU Admission.

In the ICU and hospital outcomes subsection P14 L248-257, similar comments to the "Baseline Characteristics" subsection apply. In addition, include p-values or other indicators of statistical significance to assess whether the observed differences are likely due to chance or represent true disparities between the waves.

On P16 L274, the authors mention the multivariable model with vaccination. Adjusting for a variable that is measured in one group but not in another can introduce bias and affect the validity of the results. In this case, adjusting for vaccination as a variable in the Cox model only for the second wave raises some concerns:

(1) Selection Bias: Since vaccination status was only measured in the second wave, it means that the two groups (first wave and second wave) are not directly comparable concerning this variable. The absence of vaccination data in the first wave introduces selection bias, potentially leading to distorted results.

(2) Limitation of generalizability: The results might not be generalizable to populations where vaccination status is differentially available or measured between waves. This limitation could affect the study's broader applicability and relevance.

In the absence of uniform vaccination data, the study could acknowledge this limitation explicitly. Authors should discuss the implications of this limitation on the interpretation of results and potential biases introduced by the unequal measurement of vaccination status between waves.

Discussion

The authors fully explored potential reasons behind the observed differences in patient outcomes between the two waves. They discussed factors such as healthcare infrastructure and public awareness. Limitations of the study were mentioned; however, the authors may want to discuss further how these limitations might have influenced the results and interpretations.

Conclusion

Consider discussing the implications of the study findings for clinical practice. How can the insights gained from this study inform future pandemic preparedness and response strategies?

Reviewer #2: Overall impression and relevance

The study presents a relevant issue concerning variants of concerns and continuous outbreaks of the COVID-19. It is professionally written, and the methods used are detailed and comprehensive. However, from the topic and aim presented in the introduction, one would expect to see detailed analysis of how the different variants of concern influence survival/mortality in both the first wave and the second wave. But the study presents a comparison of survival/mortality of patients based on admission clinical characteristics during the first 24 hours as its primary outcome and assessing the difference between other clinical outcomes between the 1st and 2nd wave as secondary outcomes. The authors should consider changing the topic and aim to align with the methods and results.

Background

1 The statement on pg. 5, L 102-103 requires backing with references.

Methods

• 2 A good description of the study setting and hospital changes to accommodate COVID-19 patients has been provided.
• 3 The sentence on pg. 6 L 133-134 "admission to the ICU after more than 7 days of invasive ventilatory support" needs clarification as invasive ventilation is only possible in ICU.
• 4 The exclusion of patients who tested PCR positive for COVID-19 but were excluded from the study based on all other exclusion criteria should be clarified. As the flow chart shows, all these patients were recruited from ICU and if the main outcome of the study is survival at 60 days of COVID-19 patients due to all other associated factors, the exclusion of these positive cases would probably cause selection bias in the sample.
• 5 On pg. 7, L154, an elaboration on how data confidentiality was maintained is required.
• 6 Detailed information has been provided on data collection methods and variables collected.

Analysis

• 7 Good explanation on how continuous and binary variables were analyzed.
• 8 On what basis were variables selected for the multivariable analysis? Please include in the analysis

Results

• 9 Good presentation of participant selection in flow diagram
• 10 The tables are well presented with detailed information, however
• 11 Does the statement on pg10, L209 suggest that no patient dies in ICU? If yes, the outcome variable "death" under ICU outcomes?
• 12 Vaccination and variants of concern were only possible to be measured for the second wave. I am not sure if they can be included in the analysis and results as this would be a concern of measurement bias.
• 13 The section on ICU and hospital outcome needs to be more detailed with specific results from table 3 being quoted. Include significant levels too.
• 14 It is not clear which variables were selected from all the baseline characteristics for assessment of association with survival at 60 days. Table 3 in the supplementary only shows 3 variables, one of which was only measured in the second wave (vaccination) and should not be included as this would bias the results.

Discussion

• 15 On Pg 17, L 285 it is unclear how the authors conclude that 60-day mortality was 27% less in the second wave. A HR of 0.61 translates to a survival of 39% more in the second wave.
• 16 Given that this study is comparing waves 1 and 2 and the variant of concerns were only assessed in wave two, the discussion on Gamma variant might not be necessary. However, this can be mentioned as a limitation and a point to note for future comparative studies
• 17 Overall, the discussion is well tailored balancing the findings with international and national findings from other studies.

***

6. PLOS authors have the option to publish the peer review history of their article (https://journals.plos.org/plosone/s/editorial-and-peer-review-process#loc-peer-review-history). If published, this will include your full peer review and any attached files.

If you choose "no", your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our https://www.plos.org/privacy-policy.

Reviewer #1: No

Reviewer #2: No

***

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Attachment

Submitted filename: PONE-D-23-21314_reviewer.pdf

Author response to Decision Letter 0

28 Dec 2023

Editorial office

C.1: Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming.

R1: We have adjusted some details of the manuscript and confirmed that it meets the requirements.

C2: Thank you for stating the following in the Competing Interests section...

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: ""This does not alter our adherence to PLOS ONE policies on sharing data and materials." (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

R2: We confirm that the statement in the Competing Interests section does not alter our adherence to all PLOS ONE policies on sharing data and materials and we included this recommended statement. Below is the updated Competing Interests statement and we included it in our cover letter.

"Dr. Ferreira reports personal fees from Medtronic, outside the submitted work; Dr. Costa reports personal fees from Timpel, personal fees from Magnamed, outside the submitted work; Dr. Ho reports personal fees from Pan‐American Health Organization, outside the submitted work. This does not alter our adherence to PLOS ONE policies on sharing data and materials. The other authors have no conflict of interest to disclose."

C3: In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study's minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

R3: We would like to update the Data Availability statement in our cover letter. It should now read:

All relevant data are within the paper and its Supporting Information files. Brazilian Data privacy regulations prohibit sharing of individual level data to the public and the ethical approval did not cover public sharing of data for unknown purposes. Anonymized data may be shared upon reasonable request upon contact with the corresponding author (daniela.f@hc.fm.usp.br) or the Research Ethics Committee (cappesq.adm@hc.fm.usp.br).

C4: We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

R4: As mentioned about, we have updated our Data Availability Statement in the cover letter.

C5: Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

R5: Yes, we have adjusted the supporting information files and confirmed that it meets the guidelines. We have added captions for our Supporting Information files at the end of the manuscript, and update in-text citations to match accordingly.

Responses to the Reviewer #1

Abstract

CR1.1: Consider providing a brief context for readers unfamiliar with the Brazilian healthcare system and the significance of studying outcomes across different waves of the pandemic.

R1: Thank you for your suggestion. We have included a brief context of the Brazilian healthcare system to the abstract (P3 L52) and the methods section (P6 L123).

Introduction

CR1.2: Provide references to the claim made on P5 L101.

R2: We have added references, as suggested (P5 L112).

Methods

CR1.3: I like how the authors provided a very good description of the study population, highlighting the selection criteria. However, additional information is needed to clarify the specific criteria that were used to define critically ill patients. Detailing parameters like respiratory rate, oxygen saturation, and comorbidities would be helpful for readers to gauge the severity of the cases.

R3: Thank you for your comment. The criteria to define critically ill patients was admission to one of the hospital´s COVID-19 ICUs. ICU admission was regulated by a team of physicians and nurses who received ICU admission requests from the emergency rooms and wards, and outside hospitals. We added this information to the text to improve clarity (P7 L141).

CR1.4: Kindly add references to the Patient Care subsection P7 L143.

On P7 L143, the authors mention the main changes in treatment protocol between 2020 and 2021. Can they also comment on the use of anticoagulants at prophylactic doses to prevent thromboembolism and the nonuse of antibiotics in patients without suspected bacterial infection as reported by Falavigna et al (https://www.jbmede.com.br/index.php/jbme/article/download/69/50).

R4: Thank you for the suggestion, we provided a more detailed description of protocols to the Patient Care subsection and the suggested reference (P7 L158).

CR1.5: In the Statistical Analysis subsection P8 L180, several essential details are missing that are crucial for a comprehensive understanding of the statistical analysis. The description mentions that Kaplan-Meier curves were plotted for each of the pandemic waves. It's important to clarify if these curves were plotted separately for each group (first wave vs. second wave) or if there was a comparison between the curves. If a comparison was made, the results of the log-rank test, which is commonly used to compare Kaplan-Meier curves between groups, should be reported.

R5: We agree that the description of the Kaplan-Meier curves and its corresponding statistical analysis was not sufficiently clear. We did perform a logrank test to compare the survival in the two waves in a unadjusted model, and we also used an unadjusted Cox proportional hazards model to estimate the association between pandemic wave and survival, in order to obtain an unadjusted hazard ratio. In our revised manuscript, we report both results (logrank and unadjusted HR). We also modified our statistical analysis section to improve clarity. Changes to the manuscript in P10 L216; P18 L 332; P18 L339; P18 L 347.

CR1.6: Kindly explain the rationale for choosing the variables to be included in the multivariable Cox model-building process. There is no mention of how assumptions such as the proportional hazards assumption in Cox models were checked. It's vital to confirm that these assumptions hold for the validity of the analysis.

R6: We used a conceptual causal diagram in the format of directed acyclic graph (DAG) to estimate the association between the main predictor (admission in the second wave) and the outcome, adjusting for relevant confounders. Variables in the DAG conceptual model were selected based on prior knowledge. We added this information to the methods section (P10 L223), with an appropriate reference (P10 L221) and in the abstract (P3 L54).

We used the Schoenfeld residual method to test the proportional hazards assumption in Cox models. We have added this information to the methods section and bellow we showed the test. Changes to the manuscript in P10 L226.

CR1.7: Information on how the goodness-of-fit of the Cox proportional hazards model was assessed is missing. Techniques like likelihood ratio tests, Akaike Information Criterion (AIC), or Bayesian Information Criterion (BIC) are commonly used to assess model fit.

R7: Thank you for bringing up this point. Because the specification of our final model was comprised solely of potential confounders previously defined according to our DAG conceptual model, we did not have to select between different models. We also did not include any interaction terms or other higher order terms in our model and did not have to select. We added a sentence to the methods section to clarify this point P10 L227.

CR1.8: Details about specific packages or functions within R used for the analyses are important for transparency and reproducibility.

R8: We agree, we used the packages survival and survminer. We added the references to the text (P11 L234).

Results

CR1.9: I liked the clear presentation of the patients' characteristics from both waves in Table 1, including demographic information, comorbidities, and other relevant factors summarizing the key differences. This will provide a comprehensive understanding of the patient population studied.

In the Baseline Characteristics subsection P10 L212-219, the authors need to provide specific numerical data from Table 1 to support their statements about the subjects. Instead of using phrases like "male sex was predominant" or "more likely," the authors should directly quote the figures from the table to enhance the precision and clarity of the description. For instance:

Instead of saying "male sex was predominant in both periods," specify the exact percentages of male patients in both the first and second waves.

Instead of saying "56% were ≥60 years old" and "43% were ≥60 years old," provide the actual numbers of patients in each age group (percentage in brackets) for both waves.

Instead of saying, "Patients in the first wave were more likely to be receiving mechanical ventilation and vasopressors," provide the specific percentages or counts of patients receiving mechanical ventilation and vasopressors in each wave.

Provide the actual numerical values for mean SAPS3 and SOFA scores in both waves to convey a more precise comparison.

By directly citing the figures from Table 1, readers can have a clear and accurate understanding of the baseline characteristics of the subjects in each wave without having to move back and forth to Table 1.

The same applies to the ventilatory management in the first 24 hours after ICU admission (P12 L225-232) and management in the first 24 hours after ICU admission (P13 L240-247) subsections. See Lalla et al as an example (https://pubmed.ncbi.nlm.nih.gov/35359698/).

R9: Thank you for the suggestion. We added specific numerical data to the text (P13 L252-258; P14 L277-280; P16 L299-303). We suppressed less relevant details from the text, since they are already shown in the respective tables to avoid excess numerical data repeated in the text. Additionally, we changed the heading of the section on the ventilatory management according to the suggested heading for the section on clinical management (P14 L274).

CR1.10: On P13 L240, the heading seems a bit vague; the authors can consider changing it to Clinical Management of Patients in the First 24 Hours After ICU Admission.

R10: We agree and changed the heading, and we specified presentation of the Clinical Management of Patients in the First 24 Hours (P15 L294).

CR1.11: In the ICU and hospital outcomes subsection P14 L248-257, similar comments to the "Baseline Characteristics" subsection apply. In addition, include p-values or other indicators of statistical significance to assess whether the observed differences are likely due to chance or represent true disparities between the waves.

R11: Yes, we added numerical data and p-values and once again suppressed less relevant details from the text, to avoid excess repetition (P16 L308-317).

CR1.12: On P16 L274, the authors mention the multivariable model with vaccination. Adjusting for a variable that is measured in one group but not in another can introduce bias and affect the validity of the results. In this case, adjusting for vaccination as a variable in the Cox model only for the second wave raises some concerns:

(1) Selection Bias: Since vaccination status was only measured in the second wave, it means that the two groups (first wave and second wave) are not directly comparable concerning this variable. The absence of vaccination data in the first wave introduces selection bias, potentially leading to distorted results.

(2) Limitation of generalizability: The results might not be generalizable to populations where vaccination status is differentially available or measured between waves. This limitation could affect the study's broader applicability and relevance.

In the absence of uniform vaccination data, the study could acknowledge this limitation explicitly. Authors should discuss the implications of this limitation on the interpretation of results and potential biases introduced by the unequal measurement of vaccination status between waves.

R12: We share the reviewer's concerns about the interpretation of the results with vaccination to the DAG model. We debated whether to include it or not to the model. Including vaccination incurs in all the problems mentioned by the reviewer. However, leaving vaccination out of the model might interfere with testing the main hypothesis of the study, given that we were comparing survival between the two waves, and vaccination was available to some patients in the second wave and was expected to reduce mortality. In order to mitigate the risk of bias, we planned and performed a sensitivity analysis without vaccination in the model, and the results of the multivariable model were very similar, but we failed to add it to the original manuscript. It is now added to the methods (P10 L224) and results sections (P18 L339).

We also agree that the limitations need to be explicitly acknowledged, we added a paragraph about vaccination to the discussion (P21 L401) and modified the limitations paragraph to address this concern (P23 L464 and P24 L473).

Discussion

CR1.13: The authors fully explored potential reasons behind the observed differences in patient outcomes between the two waves. They discussed factors such as healthcare infrastructure and public awareness. Limitations of the study were mentioned; however, the authors may want to discuss further how these limitations might have influenced the results and interpretations.

R13: We agree and have added a sentence of the impact of the limitations on data interpretation at the end of the paragraph (P24 L473).

Conclusion

CR1.14: Consider discussing the implications of the study findings for clinical practice. How can the insights gained from this study inform future pandemic preparedness and response strategies?

R14: Thank you for your suggestion. We added a sentence about the importance of our findings to show that implementing evidence-based protocols can impact patient outcomes to the last paragraph of the discussion and added additional comments about pandemic preparedness to the conclusion paragraph (P24 L482 and P25 L499).

Responses to the Reviewer #2

Overall impression and relevance

CR2: The study presents a relevant issue concerning variants of concerns and continuous outbreaks of the COVID-19. It is professionally written, and the methods used are detailed and comprehensive. However, from the topic and aim presented in the introduction, one would expect to see detailed analysis of how the different variants of concern influence survival/mortality in both the first wave and the second wave. But the study presents a comparison of survival/mortality of patients based on admission clinical characteristics during the first 24 hours as its primary outcome and assessing the difference between other clinical outcomes between the 1st and 2nd wave as secondary outcomes. The authors should consider changing the topic and aim to align with the methods and results.

R: Thank you for your comment. Indeed, we were interested in the impact of the emergence of VoCs in the survival of COVID-19 critically ill patients and our hypothesis was that patients admitted in the second wave, which was driven by the emergence of VoCs, had worse clinical outcomes. In the background, we discuss VoCs, but also other factors potentially associated with worse outcomes, such as health care system strain. On the last paragraph of the introduction, we state that our aim was to "compare characteristics, clinical management, and outcomes of critically ill patients hospitalized in the first and second waves of COVID-19 in a large academic hospital in Brazil. We hypothesized that in-hospital mortality would be greater in the second wave." We did test for VoCs for patients in the second wave. There were missing data, but 94% of the samples detected the Gamma variant. As a result, year of admission, our main predictor of survival in the unadjusted analysis, is a proxy of VoCs since all patients in the first wave had the original strain and almost all patients in the second wave were contaminated with the Gamma variant. Baseline clinical characteristics were added only to the adjusted model to account for confounding, but the main analysis focuses on wave of admission (and therefore, VoCs). Then, we believe that our aims are aligned with the results. In order to better clarify that intention, we modified the title (I believe the reviewer meant title instead of topic), mentioning that we were interested in comparing the two waves, which are driven by different variants (P1 L5).

Background

CR2.1 The statement on pg. 5, L 102-103 requires backing with references.

R1: Thank you for your comment, we added the references (P5 L 112).

Methods

CR2.2 A good description of the study setting and hospital changes to accommodate COVID-19 patients has been provided.

R2: Thank you for your comment.

CR2.3 The sentence on pg. 6 L 133-134 "admission to the ICU after more than 7 days of invasive ventilatory support" needs clarification as invasive ventilation is only possible in ICU. Do you mean admission to the COVID-19 ICU after more than 7 days of invasive ventilatory support? Because invasive ventilation is possible only when a patient is in ICU. Also, why would you exclude these patients unless they were confirmed negative for COVID-19?

R3: We appreciate the comment. However, in our hospital, we sometimes received patients transferred from other hospitals with fewer resources than ours, and occasionally they might have been under mechanical ventilation in the previous hospital for several days and have been received variable ventilatory management, and we believe that they may be very different form our target population, and therefore needed to be excluded. Only 19 patients were excluded for this reason. On the other hand, in our hospital, invasive ventilation was started in the emergency room for patients who arrived with overt respiratory failure, or the wards for patients with sudden deterioration. These patients typically were transferred to the ICU within a few hours and were not excluded from the study. We added a clarification on the methods section (P7 L141 e P7 L148-152).

CR2.4 The exclusion of patients who tested PCR positive for COVID-19 but were excluded from the study based on all other exclusion criteria should be clarified. As the flow chart shows, all these patients were recruited from ICU and if the main outcome of the study is survival at 60 days of COVID-19 patients due to all other associated factors, the exclusion of these positive cases would probably cause selection bias in the sample.

R4: Thank you for your comment. We aimed to include a very broad and representative sample of critically ill patients with COVID-19 in our study. However, we believe that including patients with certain clinical conditions such as terminal disease and palliative care, for example, or prolonged mechanical ventilation provided in a different hospital, would introduce bias into our sample. Moreover, we also excluded patients who were admitted to the ICU for a reason that was not related to COVID-19 (post exploratory laparotomy for abdominal obstruction, for example), had no COVID-19 symptoms, but tested positive in screening tests, since at the time all patients admitted to our hospital, for any reason, were tested. We edited the methods section to clarify this exclusion criteria (P7 L148-152). Only 4 patients were excluded for this reason and less than 7% of the total sample were excluded for any of the exclusion criteria.

CR2.5 On pg. 7, L154, an elaboration on how data confidentiality was maintained is required.

An elaboration of this should be included especially as online information hacking can occur.

R5: We agree, we added more details (P8 L172).

CR2.6: Detailed information has been provided on data collection methods and variables collected.

What about this data from patients in the first wave?

R6: Thank you for your comment. We did not test respiratory samples from patients admitted in the first wave for SARS-CoV-2 variants because at the time, there were no VOCs circulating in Brazil. Patients the first wave we considered infected by the original strain. We added a sentence to the methods section to clarify this point (P9 L200).

Analysis

CR2.7: Good explanation on how continuous and binary variables were analyzed.

R7: Thank you for your comment.

CR2.8: On what basis were variables selected for the multivariable analysis? Please include in the analysis

R8: Thank you for your comment. As we mentioned in our response to reviewer #1, we used a conceptual causal diagram in the format of directed acyclic graph (DAG) to estimate the association between the main predictor admission in the second wave and the outcome, adjusting for relevant confounders. Variables in the DAG conceptual model were selected based on prior knowledge, as is usually the case. We added this information to the methods section (P10 L 223), with an appropriate reference (P10 L 221).

Results

CR2.9: Good presentation of participant selection in flow diagram

R9: Thank you for your comment.

CR2.10: The tables are well presented with detailed information, however.

R10: We believe this comment precedes the next one.

CR2.11: Does the statement on pg10, L209 suggest that no patient dies in ICU? If yes, the outcome variable "death" under ICU outcomes?

R11: Thank you for pointing out this omission, we added that follow up was continued until discharge, death or transfer (P12 L 246).

CR2.12: Vaccination and variants of concern were only possible to be measured for the second wave. I am not sure if they can be included in the analysis and results as this would be a concern of measurement bias.

R12: We share both reviewer's concerns about the interpretation of the results with vaccination to the DAG model. As we mentioned in our response to reviewer #1, we debated whether to include it or not to the model. Including vaccination incurs in all the problems mentioned by the reviewer. However, leaving vaccination out of the model might interfere with testing the main hypothesis of the study, given that we were comparing survival between the two waves, and vaccination was available to some patients in the second wave and was expected to reduce mortality. In order to mitigate the risk of bias, we planned and performed a sensitivity analysis without vaccination in the model, and the results of the multivariable model were very similar, but we failed to add it to the original manuscript. It is now added to the methods (P10 L224) and results sections (P18 L339), and a sentence on the limitations paragraph.

Variants of concern were not used as predictors in the survival analysis, we used it to characterize our study population on the second wave.

CR2.13: The section on ICU and hospital outcome needs to be more detailed with specific results from table 3 being quoted. Include significant levels too.

R13: We agree, we added numerical data and p-values.

CR2.14: It is not clear which variables were selected from all the baseline characteristics for assessment of association with survival at 60 days. Table 3 in the supplementary only shows 3 variables, one of which was only measured in the second wave (vaccination) and should not be included as this would bias the results.

R14: We agree that the original manuscript did not explicitly mentioned the criterion for selecting variables for the survival model. We used a conceptual causal diagram in the format of directed acyclic graph (DAG) to estimate the association between the main predictor (admission in the second wave) and the outcome, adjusting for relevant confounders. Variables in the DAG conceptual model were selected based on prior knowledge, as is usually the case. We added this information to the methods section (P10 L 223), with an appropriate reference (P10 L 221) and in the abstract (P3 L54). The discussion about the inclusion of vaccination in the model and the sensitivity analysis is discussed above.

Discussion

CR2.15: On Pg 17, L 285 it is unclear how the authors conclude that 60-day mortality was 27% less in the second wave. A HR of 0.61 translates to a survival of 39% more in the second wave.

R15: Thank you for your comment. We agree that the sentence is confusing because it mentions mortality instead of survival. All of our analyses are based on survival, not mortality, so mentioning the mortality rate on the first paragraph of Discussion was inappropriate. Changes to the manuscript in P19 L354.

CR2.16: Given that this study is comparing waves 1 and 2 and the variant of concerns were only assessed in wave two, the discussion on Gamma variant might not be necessary. However, this can be mentioned as a limitation and a point to note for future comparative studies.

R16: Thank you for your comment. We agree and added a sentence about future comparative studies to the Discussion section (P22 L421).

CR2.17: Overall, the discussion is well tailored balancing the findings with international and national findings from other studies.

R17: Thank you for your comment.

Attachment

Submitted filename: Response to Reviewers.docx

Decision Letter 1

Reyes Luis Felipe Academic Editor

15 Feb 2024

Temporal trends of severity and outcomes of critically ill patients with COVID-19 after the emergence of variants of concern: A comparison of two waves

PONE-D-23-21314R1

Dear Dr. Freitas,

We're pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you'll receive an e-mail detailing the required amendments. When these have been addressed, you'll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they'll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Luis Felipe Reyes, M.D., Ph.D., MSc.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the "Comments to the Author" section, enter your conflict of interest statement in the "Confidential to Editor" section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

***

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

***

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

***

4. Have the authors made all data underlying the findings in their manuscript fully available?

The http://www.plosone.org/static/policies.action#sharing requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: No

***

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

***

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Thank you for allowing me the opportunity to review the revised manuscript on "Temporal trends of severity and outcomes of critically ill patients with COVID-19 after the emergence of variants of concern: A comparison of two waves." I appreciate the authors' efforts in addressing the major comments and concerns raised in the initial submission.

The methods section now provides a comprehensive description of the statistical methods employed, addressing previous deficiencies.

The results section has been reframed and rewritten in accordance with expected reporting guidelines. This is commendable.

Regarding the discussion section, I noticed a shift from discussing "survival" on line 322 to using the term "mortality" on line 343. To maintain consistency and avoid confusion for readers, I recommend continuing with the term "survival" throughout this section.

Overall, the manuscript is now well-structured and reads smoothly.

Reviewer #2: All coments have been adressed appropraitly by the authors.

The authors have re-analysed their data and provided more detailed explanantions and transparency.

***

7. PLOS authors have the option to publish the peer review history of their article (https://journals.plos.org/plosone/s/editorial-and-peer-review-process#loc-peer-review-history). If published, this will include your full peer review and any attached files.

If you choose "no", your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our https://www.plos.org/privacy-policy.

Reviewer #1: No

Reviewer #2: No

***

Acceptance letter

Reyes Luis Felipe Academic Editor

27 Feb 2024

PONE-D-23-21314R1

PLOS ONE

Dear Dr. Freitas,

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now being handed over to our production team.

At this stage, our production department will prepare your paper for publication. This includes ensuring the following:

* All references, tables, and figures are properly cited

* All relevant supporting information is included in the manuscript submission,

* There are no issues that prevent the paper from being properly typeset

If revisions are needed, the production department will contact you directly to resolve them. If no revisions are needed, you will receive an email when the publication date has been set. At this time, we do not offer pre-publication proofs to authors during production of the accepted work. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few weeks to review your paper and let you know the next and final steps.

Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

If we can help with anything else, please email us at customercare@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Luis Felipe Reyes

Academic Editor

PLOS ONE

We would like to thank the Hospital das Clinicas COVID‐19 crisis committee, healthcare workers and staff for their important work at our hospital during the COVID‐19 pandemic.

Abbreviations

• aHR

• Adjusted hazard ratio

• ARDS

• Acute Respiratory Distress Syndrome

• BMI

• Body mass index
• COVID‐19
• Coronavirus Disease 2019

• DAG

• Directed acyclic graph

• FiO2

• Fraction of inspired oxygen

• HCFMUSP

• Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo

• HFNC

• High-flow nasal cannula

• HR

• Hazard ratio

• IBW

• Ideal body weight

• ICU

• Intensive care unit

• IgG

• Immunoglobulin G

• IgM

• Immunoglobulin M

• IQR

• Interquartile range

• NIV

• Noninvasive ventilation
• PaO2/FiO2
• Partial pressure of arterial oxygen and fraction of inspired oxygen ratio

• PEEP

• Positive end-expiratory pressure

• REDCap

• Research Electronic Data Capture

• RRT

• Renal replacement therapy

• RT-PCR

• Reverse Transcription Polymerase Chain Reaction

• SAPS 3

• Simplified Acute Physiology Score III

• SARI

• Severe acute respiratory infection
• SARS-CoV-2
• Severe Acute Respiratory Syndrome—Coronavirus 2

• SD

• standard deviation

• SOFA

• Sequential (Sepsis-related) Organ Failure Assessment

• STROBE

• Strengthening The Reporting of Observational Studies in Epidemiology

• VOCs

• Variants of concern

1 Dr. Ferreira reports personal fees from Medtronic, outside the submitted work; Dr. Costa reports personal fees from Timpel, personal fees from Magnamed, outside the submitted work; Dr. Ho reports personal fees from Pan‐American Health Organization, outside the submitted work. This does not alter our adherence to PLOS ONE policies on sharing data and materials. The other authors have no conflict of interest to disclose. World Health Organization (WHO).Coronavirus Disease (COVID-19) Dashboard.Geneva: WHO; (2023). Available from: https://covid19.who.int/. Accessed 2023 July 7. 2 Ferreira JC, Ho YL, Besen BAMP, Malbouisson LMS, Taniguchi LU, Mendes PV, et al. Protective ventilation and outcomes of critically ill patients with COVID-19: a cohort study. Ann Intensive Care. (2021) Jun 7; 11(1):92. doi: 10.1186/s13613-021-00882-w, 34097145 3 Ranzani OT, Bastos LSL, Gelli JGM, Marchesi JF, Baião F, Hamacher S, et al. Characterisation of the first 250 000 hospital admissions for COVID-19 in Brazil: a retrospective analysis of nationwide data. Lancet Respir Med. (2021) Apr 1; 9(4):407–418. doi: 10.1016/S2213-2600(20)30560-9, 33460571 4 World Health Organization (WHO).Tracking SARS-CoV-2 variants. Geneva: WHO; (2023). Available from: https://www.who.int/activities/tracking-SARS-CoV-2-variants. Accessed 2022 Oct 24. 5 Gräf T, Bello G, Naveca FG, Gomes M, Cardoso VLO, da Silva AF, et al. Phylogenetic-based inference reveals distinct transmission dynamics of SARS-CoV-2 lineages Gamma and P.2 in Brazil. iScience. (2022) Apr 15; 25(4):104156. doi: 10.1016/j.isci.2022.104156, 35368908 6 Faria NR, Mellan TA, Whittaker C, Claro IM, Candido DDS, Mishra S, et al. Genomics and epidemiology of the P.1 SARS-CoV-2 lineage in Manaus, Brazil. Science. (2021) May 21; 372(6544):815–821. doi: 10.1126/science.abh2644, 33853970 7 Hoffmann M, Arora P, Groß R, Seidel A, Hörnich BF, Hahn AS, et al. SARS-CoV-2 variants B.1.351 and P.1 escape from neutralizing antibodies. Cell. (2021) Apr 29; 184(9):2384–2393.e12. doi: 10.1016/j.cell.2021.03.036, 33794143 8 Souza WM, Amorim MR, Sesti-Costa R, Coimbra LD, Brunetti NS, Toledo-Teixeira DA, et al. Neutralisation of SARS-CoV-2 lineage P.1 by antibodies elicited through natural SARS-CoV-2 infection or vaccination with an inactivated SARS-CoV-2 vaccine: an immunological study. Lancet Microbe. (2021) Oct 1; 2(10):e527–35. doi: 10.1016/S2666-5247(21)00129-4, 34258603 9 Harvey WT, Carabelli AM, Jackson B, Gupta RK, Thomson EC, Harrison EM, et al. SARS-CoV-2 variants, spike mutations and immune escape. Nat Rev Microbiol. (2021) Jul; 19(7):409–424. doi: 10.1038/s41579-021-00573-0, 34075212 Hahn G, Wu CM, Lee S, Lutz SM, Khurana S, Baden LR, et al. Genome-wide association analysis of COVID-19 mortality risk in SARS-CoV-2 genomes identifies mutation in the SARS-CoV-2 spike protein that colocalizes with P.1 of the Brazilian strain. Genet Epidemiol. (2021) Oct 1; 45(7):685–93. doi: 10.1002/gepi.22421, 34159627 Silva SJR da, Pena L. Collapse of the public health system and the emergence of new variants during the second wave of the COVID-19 pandemic in Brazil. One Health. (2021) Dec; 13:100287. doi: 10.1016/j.onehlt.2021.100287, 34222607 Lin L, Liu Y, Tang X, He D. The Disease Severity and Clinical Outcomes of the SARS-CoV-2 Variants of Concern. Front Public Health. (2021) Nov 30; 9:775224. doi: 10.3389/fpubh.2021.775224, 34917580 Nicolete VC, Rodrigues PT, Fernandes ARJ, Corder RM, Tonini J, Buss LF, et al. Epidemiology of COVID-19 after Emergence of SARS-CoV-2 Gamma Variant, Brazilian Amazon, 2020–2021. Emerg Infect Dis. (2022) Mar 1; 28(3):709–12. doi: 10.3201/eid2803.211993, 34963505 Fundação Oswaldo Cruz (Brasil).A Rede Genômica. Rio de Janeiro: FIOCRUZ, (2022). Portuguese. Available from: http://www.genomahcov.fiocruz.br/. Accessed 2022 Oct 26. Bastos LS, Ranzani OT, Souza TML, Hamacher S, Bozza FA. COVID-19 hospital admissions: Brazil's first and second waves compared. Lancet Respir Med. (2021) Aug; 9(8):e82–e83. doi: 10.1016/S2213-2600(21)00287-3, 34273268 Fundação Oswaldo Cruz (Brasil).COVID-19 Observatory points to the biggest health system collapse in the history of Brazil. Rio de Janeiro: FIOCRUZ, (2022). Available from: https://portal.fiocruz.br/en/news/covid-19-observatory-points-biggest-health-system-collapse-history-brazil. Accessed 2022 Oct 26. Lobo SM, Mello PMVC. Challenges of the coronavirus pandemic for Brazilian intensivists: present and future. Rev Bras Ter Intensiva (2021) Sep 20:S0103–507X2021005001203. Portuguese, English. Challen R, Brooks-Pollock E, Read JM, Dyson L, Tsaneva-Atanasova K, Danon L. Risk of mortality in patients infected with SARS-CoV-2 variant of concern 202012/1: Matched cohort study. BMJ. (2021) Mar 9; 372:n579. doi: 10.1136/bmj.n579, 33687922 Salyer SJ, Maeda J, Sembuche S, Kebede Y, Tshangela A, Moussif M, et al. The first and second waves of the COVID-19 pandemic in Africa: a cross-sectional study. Lancet. (2021) Apr 3; 397(10281):1265–1275. doi: 10.1016/S0140-6736(21)00632-2, 33773118 Davies NG, Jarvis CI, CMMID COVID-19 Working Group, Edmunds WJ, Jewell NP, Diaz-Ordaz K, et al. Increased mortality in community-tested cases of SARS-CoV-2 lineage B.1.1.7. Nature. (2021) May; 593(7858):270–274. doi: 10.1038/s41586-021-03426-1, 33723411 Abdool Karim SS, de Oliveira T. New SARS-CoV-2 variants—Clinical, Public Health, and Vaccine Implications. N Engl J Med. (2021) May 13; 384(19):1866–1868. doi: 10.1056/NEJMc2100362, 33761203 Whittaker C, Ratmann O, Dye C, Sabino EC, Faria NR. Altered demographic profile of hospitalizations during the second COVID-19 wave in Amazonas, Brazil. Lancet Reg Health Am. (2021) Oct; 2:100064. doi: 10.1016/j.lana.2021.100064, 34522912 Freitas ARR, Beckedorff OA, Cavalcanti LP de G, Siqueira AM, Castro DB, Costa CFD, et al. The emergence of novel SARS-CoV-2 variant P.1 in Amazonas (Brazil) was temporally associated with a change in the age and sex profile of COVID-19 mortality: A population based ecological study. Lancet Reg Health Am. (2021) Sep; 1:100021. doi: 10.1016/j.lana.2021.100021, 34514463 Spira B.The Impact of the Highly Virulent SARS-CoV-2 Gamma Variant on Young Adults in the State of São Paulo: Was It Inevitable? Cureus. (2022) Jul 1; 14(7):e26486. de Souza FSH, Hojo-Souza NS, de Oliveira DCR, Silva CMD, Guidoni DL. An overview of Brazilian working age adults vulnerability to COVID-19. Sci Rep. (2022) Feb 18; 12(1):2798. doi: 10.1038/s41598-022-06641-6, 35181692 Ranzani OT, Hitchings MDT, Dorion M, D'Agostini TL, De Paula RC, De Paula OFP, et al. Effectiveness of the CoronaVac vaccine in older adults during a gamma variant associated epidemic of covid-19 in Brazil: Test negative case-control study. BMJ. (2021) Sep 6; 374:n2091. Erratum for: BMJ. 2021 Aug 20;374:n2015. doi: 10.1136/bmj.n2091, 34489240 BRASIL. Ministério da Saúde. Secretaria de Vigilância em Saúde. Departamento de Imunização e Doenças Transmissíveis. Plano Nacional de Operaracionalização da Vacinação contra a COVID-19 [National Plan for the Operationalization of Vaccination against COVID-19].7th ed. p. 1–108. Brasília, (2021). Portuguese. Available from: https://sbim.org.br/images/files/notas-tecnicas/plano-nacional-operacionalizacao-planovacinacaocovid-ed7-20210517-cgpni.pdf. Accessed 2023 Jan 4. Hallal PC, Victora CG. Overcoming Brazil's monumental COVID-19 failure: an urgent call to action. Nat Med. (2021) Jun; 27(6):933. doi: 10.1038/s41591-021-01353-2, 33958796 Falavigna M, Stein C, Amaral JLG, Azevedo LCP, Belli KC, Colpani V, et al. Diretrizes Brasileiras para o tratamento farmacológico de pacientes hospitalizados com COVID-19. Rev Bras Ter Intensiva (2022) Jan; 34(1):1–12. Portuguese, English. Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O'Neal L, et al. The REDCap consortium: Building an international community of software platform partners. J Biomed Inform. (2019) Jul; 95:103208. doi: 10.1016/j.jbi.2019.103208, 31078660 Moreno RP, Metnitz PGH, Almeida E, Jordan B, Bauer P, Campos RA, et al; SAPS 3 Investigators. SAPS 3—From evaluation of the patient to evaluation of the intensive care unit. Part 2: Development of a prognostic model for hospital mortality at ICU admission. Intensive Care Med. (2005) Oct; 31(10):1345–55. Erratum in: Intensive Care Med. 2006 May;32(5):796. Vincent JL, Moreno R, Takala J, Willatts S, de Mendonça A, Bruining H, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med. (1996) Jul; 22(7):707–10. doi: 10.1007/BF01709751, 8844239 von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg. (2014) Dec; 12(12):1495–9. doi: 10.1016/j.ijsu.2014.07.013, 25046131 Lederer DJ, Bell SC, Branson RD, Chalmers JD, Marshall R, Maslove DM, et al. Control of Confounding and Reporting of Results in Causal Inference Studies. Guidance for Authors from Editors of Respiratory, Sleep, and Critical Care Journals. Ann Am Thorac Soc. (2019) Jan; 16(1):22–8. doi: 10.1513/AnnalsATS.201808-564PS, 30230362 R Core Team ((2022)). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Available from: https://posit.cloud. Kassambara A, Kosinski M, Biecek P ((2021)). survminer: Drawing Survival Curves using "ggplot2". R package version 0.4.9. 2021. Available from: https://rpkgs.datanovia.com/survminer/index.html. Therneau TM, Grambsch PM. Modeling Survival Data: Extending the Cox Model. New York: Springer-Verlag. (2000). Therneau TM. A Package for Survival Analysis in R. R package version 3.5–7. (2023). Available from: https://CRAN.R-project.org/package=survival Hoogenboom WS, Pham A, Anand H, Fleysher R, Buczek A, Soby S, et al. Clinical characteristics of the first and second COVID-19 waves in the Bronx, New York: A retrospective cohort study. Lancet Reg Health Am. (2021) Nov; 3:100041. doi: 10.1016/j.lana.2021.100041, 34423331 Dongelmans DA, Termorshuizen F, Brinkman S, Bakhshi-Raiez F, Arbous MS, de Lange DW, et al. Characteristics and outcome of COVID-19 patients admitted to the ICU: a nationwide cohort study on the comparison between the first and the consecutive upsurges of the second wave of the COVID-19 pandemic in the Netherlands. Ann Intensive Care. 2022 (2022) Jan 13; 12(1):5. doi: 10.1186/s13613-021-00978-3, 35024981 Domingo P, Pomar V, Mur I, Castellví I, Corominas H, de Benito N. Not all COVID-19 pandemic waves are alike. Clin Microbiol Infect. (2021) Jul 1; 27(7):1040.e7–1040.e10. doi: 10.1016/j.cmi.2021.04.005, 33887469 Fan G, Yang Z, Lin Q, Zhao S, Yang L, He D. Decreased Case Fatality Rate of COVID‐19 in the Second Wave: A study in 53 countries or regions. Transbound Emerg Dis. (2021) Mar; 68(2):213–215. doi: 10.1111/tbed.13819, 32892500 Iftimie S, Lopez-Azcona AF, Vallverdu I, Hernandez-Flix S, de Febrer G, Parra S, et al. First and second waves of coronavirus disease-19: A comparative study in hospitalized patients in Reus, Spain. PLoS One. (2021) Mar 31; 16(3):e0248029. doi: 10.1371/journal.pone.0248029, 33788866 Prete CA, Buss LF, Whittaker C, Salomon T, Oikawa MK, Pereira RHM, et al. SARS-CoV-2 antibody dynamics in blood donors and COVID-19 epidemiology in eight Brazilian state capitals: A serial cross-sectional study. Elife. (2022) Sep 22; 11:e7823. doi: 10.7554/eLife.78233, 36135358 Brizzi A, Whittaker C, Servo LMS, Hawryluk I, Prete CA, de Souza WM, et al. Spatial and temporal fluctuations in COVID-19 fatality rates in Brazilian hospitals. Nat Med. (2022) Jul; 28(7):1476–1485doi: 10.1038/s41591-022-01807-1, 35538260 Zeiser FA, Donida B, da Costa CA, Ramos G de O, Scherer JN, Barcellos NT, et al. First and second COVID-19 waves in Brazil: A cross-sectional study of patients' characteristics related to hospitalization and in-hospital mortality. ancet Reg Health Am. (2022)Feb; 6:100107. Silva GAE, Jardim BC, Lotufo PA. Mortalidade por COVID-19 padronizada por idade nas capitais das diferentes regiões do Brasil [Age-adjusted COVID-19 mortality in state capitals in different regions of Brazil]. Cad Saude Publica. (2021) Jul 7; 37(6):e00039221. Portuguese. Lázaro APP, Albuquerque PLMM, Meneses GC, Zaranza MS, Batista AB, Aragão NL, et al. Critically ill COVID-19 patients in northeast Brazil: mortality predictors during the first and second waves including SAPS 3. Trans R Soc Trop Med Hyg. (2022) Nov 1; 116(11):1054–1062. doi: 10.1093/trstmh/trac046, 35598042 Paredes MI, Lunn SM, Famulare M, Frisbie LA, Painter I, Burstein R, et al. Associations Between Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Variants and Risk of Coronavirus Disease 2019 (COVID-19) Hospitalization Among Confirmed Cases in Washington State: A Retrospective Cohort Study. Clin Infect Dis. (2022) Aug 24; 75(1):e536–e544. doi: 10.1093/cid/ciac279, 35412591 Nonaka CKV, Gräf T, Barcia CA de L, Costa VF, de Oliveira JL, Passos R da H, et al. SARS-CoV-2 variant of concern P.1 (Gamma) infection in young and middle-aged patients admitted to the intensive care units of a single hospital in Salvador, Northeast Brazil, February 2021. Int J Infect Dis. (2021) Oct; 111:47–54. doi: 10.1016/j.ijid.2021.08.003, 34390857 Oliveira EA, Simões e Silva AC, Oliveira MCL, Colosimo EA, Mak RH, Vasconcelos MA, et al. Comparison of the First and Second Waves of the Coronavirus Disease 2019 Pandemic in Children and Adolescents in a Middle-Income Country: Clinical Impact Associated with Severe Acute Respiratory Syndrome Coronavirus 2 Gamma Lineage. J Pediatr. (2022) May; 244:178–185.e3. doi: 10.1016/j.jpeds.2022.01.001, 35031347 Santos RS, Barros DS, Moraes TMP, Hayashi CY, Ralio RB, Minenelli FF, et al. Clinical characteristics and outcomes of hospitalized patients with COVID-19 in a Brazilian hospital: a retrospective study of the first and second waves. IJID Regions. (2022) Jun; 3:189–95. doi: 10.1016/j.ijregi.2022.04.002, 35720152 Cavalcanti AB, Zampieri FG, Rosa RG, Azevedo LCP, Veiga VC, Avezum A, et al. Hydroxychloroquine with or without Azithromycin in Mild-to-Moderate Covid-19. N Engl J Med. (2020) Nov 19; 383(21):2041–2052. Erratum in: N Engl J Med. 2020 Nov 19;383(21):e119. doi: 10.1056/NEJMoa2019014, 32706953 RECOVERY Collaborative Group; Horby P, Lim WS, Emberson JR, Mafham M, Bell JL, et al. Dexamethasone in Hospitalized Patients with Covid-19. N Engl J Med. (2021) Feb 25; 384(8):693–704. doi: 10.1056/NEJMoa2021436, 32678530 Tomazini BM, Maia IS, Cavalcanti AB, Berwanger O, Rosa RG, Veiga VC, et al. Effect of Dexamethasone on Days Alive and Ventilator-Free in Patients with Moderate or Severe Acute Respiratory Distress Syndrome and COVID-19: The CoDEX Randomized Clinical Trial. JAMA. (2020) Oct 6; 324(13):1307–1316. doi: 10.1001/jama.2020.17021, 32876695 Winslow RL, Zhou J, Windle EF, Nur I, Lall R, Ji C, et al. SARS-CoV-2 environmental contamination from hospitalised patients with COVID-19 receiving aerosol-generating procedures. Thorax. (2022) Mar; 77(3):259–267. doi: 10.1136/thoraxjnl-2021-218035, 34737194 Perkins GD, Ji C, Connolly BA, Couper K, Lall R, Baillie JK, 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) Feb 8; 327(6):546–558. doi: 10.1001/jama.2022.0028, 35072713 Frat JP, Quenot JP, Badie J, Coudroy R, Guitton C, Ehrmann S, et al. Effect of High-Flow Nasal Cannula Oxygen vs Standard Oxygen Therapy on Mortality in Patients With Respiratory Failure Due to COVID-19 The SOHO-COVID Randomized Clinical Trial. JAMA. (2022) Sep 27; 328(12):1212–1222. doi: 10.1001/jama.2022.15613, 36166027 Guérin C, Reignier J, Richard JC, Beuret P, Gacouin A, Boulain T, et al; PROSEVA Study Group. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med. (2013) Jun 6; 368(23):2159–68. doi: 10.1056/NEJMoa1214103, 23688302

By Daniela Helena Machado Freitas; Eduardo Leite Vieira Costa; Natalia Alcantara Zimmermann; Larissa Santos Oliveira Gois; Mirella Vittig Alves Anjos; Felipe Gallego Lima; Pâmela Santos Andrade; Daniel Joelsons; Yeh‐Li Ho; Flávia Cristina Silva Sales; Ester Cerdeira Sabino; Carlos Roberto Ribeiro Carvalho and Juliana Carvalho Ferreira

Reported by Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author