Impact of staffing model conversion from a mandatory critical care consultation model to a closed unit model in the medical intensive care unit.

BACKGROUND: The intensive care unit (ICU) staffing model affects clinical outcomes of critically ill patients. However, the benefits of a closed unit model have not been extensively compared to those of a mandatory critical care consultation model.
METHODS: This retrospective before-after study included patients admitted to the medical ICU. Anthropometric data, admission reason, Acute Physiology and Chronic Health Evaluation II score, Eastern Cooperative Oncology Group grade, survival status, length of stay (LOS) in the ICU, duration of mechanical ventilator care, and occurrence of ventilator-associated pneumonia (VAP) were recorded. The staffing model of the medical ICU was changed from a mandatory critical care consultation model to a closed unit model in September 2017, and indices before and after the conversion were compared.
RESULTS: A total of 1,526 patients were included in the analysis. The mean age was 64.5 years, and 954 (62.5%) patients were men. The mean LOS in the ICU among survivors was shorter in the closed unit model than in the mandatory critical care consultation model by multiple regression analysis (5.5 vs. 6.7 days; p = 0.005). Central venous catheter insertion (38.5% vs. 51.9%; p < 0.001) and VAP (3.5% vs. 8.6%; p < 0.001) were less frequent in the closed unit model group than in the mandatory critical care consultation model group. After adjusting for confounders, the closed unit model group had decreased ICU mortality (adjusted odds ratio 0.65; p < 0.001) and shortened LOS in the ICU compared to the mandatory critical care consultation model group.
CONCLUSION: The closed unit model was superior to the mandatory critical care consultation model in terms of ICU mortality and LOS among ICU survivors.

Introduction

The intensive care unit (ICU) is one of the most specialized units in hospitals. Critically ill patients have various comorbidities and need critical support, such as mechanical ventilation or renal replacement therapy (RRT), which requires skillful workmanship and extensive knowledge. Decisions in the ICU need to be accurate and prompt to respond to rapid changes in deteriorating patients. Hence, it is widely recommended that ICU physicians be experienced clinicians in critical care medicine.

Intensivists are board-certified experts in providing care for critically ill patients. As critical care medicine has become a distinct specialty, the need for specialized critical care physicians continues to grow worldwide. In South Korea, the subspecialty system for critical care medicine started in 2008 [1], and there were over 1,500 intensivists by 2019. However, there is an unmet need due to their high-demand and uneven distribution [2].

Although many institutions have ICUs, the staffing models of each ICU differ according to the number of available intensivists and the economic and cultural situations. Pronovost et al. classified ICU staffing models of intensivists into four groups: 1) a closed unit where the intensivist is the patient’s primary attending physician, 2) a mandatory critical care consultation model where every patient admitted to the ICU receives a critical care consultation, 3) an elective critical care consultation model where the intensivist is involved only when needed, and 4) a model with no intensivist available. The former two groups were further classified as high-intensity staffing models, and the latter two groups were classified as low-intensity staffing models [3].

Many studies on the impact of the staffing model of intensivists in the ICU have routinely compared the high-intensity and low-intensity staffing models, revealing that a high-intensity staffing model was associated with reduced ICU and hospital mortality compared to a low-intensity model [4-14]. Several meta-analyses have shown similar results [3, 15, 16]. However, some studies have reported contradictory results [17, 18]. This subject has rarely been approached in Asian ICUs, and further studies are required to seek a plausible explanation for the differences in outcomes according to changes in the ICU staffing models. This study aimed to compare the closed unit model to the mandatory critical care consultation model and evaluate the superiority of the closed unit model in terms of outcomes in critically ill Asian patients.

Methods

1. Study design and participants

This retrospective before-after cohort study included patients admitted to the medical ICU of a university-affiliated teaching hospital. Patients aged > 19 years and who were admitted to the medical ICU between January 2016 and August 2018 were included. Until August 2017, all patients admitted to the medical ICU received mandatory consultation by a board-certified intensivist. Two intensivists board certified in Internal Medicine and Pulmonology consulted and supervised all patients in the medical ICU daily, but the original primary attending physician continued to be primary charge. From September 2017, the ICU staffing model was changed from a mandatory critical care consultation model to a closed unit model. Patient care in the medical ICU was formally transferred to an intensivist. The dedicated intensivist was present in the ICU during the weekday daytime and was responsible for all patient care, including admission, management, and discharge. The same two intensivists were involved in patient management in both models. We compared the indices before and after the conversion to evaluate the advantages of the closed unit model and compared it to the mandatory critical care consultation model.

This study was approved by the Institutional Review Board (IRB) of Seoul National University Hospital (IRB No.1807-140-961). The requirement for informed consent was waived because of the retrospective design of this study.

2. Data collection

The following variables were recorded after reviewing the medical records: age, sex, Acute Physiology and Chronic Health Evaluation (APACHE) II score, Eastern Cooperative Oncology Group (ECOG) performance status grade, primary reason for ICU admission, cardiopulmonary resuscitation (CPR), referral to palliative care within 24 hours of ICU admission, treatments during the ICU stay, occurrence of ventilator-associated pneumonia (VAP) and delirium during ICU stay, ICU readmission (readmission within 48 hours of ICU discharge), survival status, length of stay (LOS) in the ICU and duration of mechanical ventilator (MV) care.

3. Statistical analysis

Participants were divided into two groups according to the staffing model and the baseline characteristics of the groups were compared. To determine the independent effect of staffing model on ICU mortality, we performed multiple logistic regression analysis by adjusting for age, sex, APACHE II score, ECOG grade, and reasons for ICU admission. The independent impact of the staffing model on the LOS in the ICU and duration of MV care were evaluated using multiple regression analysis.

Subgroup analyses were conducted according to the five reasons for admission: respiratory diagnosis, cardiovascular diagnosis, acute kidney injury, sepsis, and neurologic diagnosis. ICU mortality, LOS in the ICU, and duration of MV care were compared.

Statistical analyses were performed using SPSS software (version 25.0 for Windows; IBM SPSS Inc., Armonk, NY, USA) and R (version 4.0.0, https://www.R-project.org). All statistical tests were two-sided, and differences were considered statistically significant at p < 0.05.

Results

1. Study population and treatment in the ICU

A total of 1,657 patients (1,076 and 581 patients in the mandatory critical consultation model group and the closed unit model group, respecively) were admitted to the medical ICU between January 2016 and August 2018. Among them, 131 patients (89 [8.3%] and 42 [7.2%] patients in mandatory critical care consultation model group and the closed unit model group, respectively) were excluded due to incomplete medical records; hence, 1,526 patients were included in the final analysis. The patients were categorized into two groups—the mandatory critical care consultation model group (987 [64.7%] patients) and the closed unit model group (539 [35.3%] patients).

The baseline characteristics of the patients are presented in Table 1. The mean age was 64.5 years, and 954 (62.5%) patients were men. Patients in the closed unit model group had higher APACHE II scores (22.7 vs. 21.3; p = 0.008), ECOG grade (3.3 vs. 3.0; p < 0.001) and higher proportion of patients with ECOG grade ≥3 (84.8% vs. 73.0%; p < 0.001) than those in the mandatory critical care consultation model group. Respiratory failure was the most common reason for ICU admission in both groups, but it was more frequent in the closed unit model group than in the mandatory critical care consultation model group (70.1% vs. 65.0%; p = 0.044).

Table 1

Baseline characteristics of the participants.

Characteristic	Mandatory critical care consultation model group (n = 987)	Closed unit model group (n = 539)	p value
Age, years	64.5 ± 14.8	64.9 ± 15.1	0.592
Male sex (n, %)	616 (62.4%)	338 (62.7%)	0.909
APACHE II score	21.3 ± 9.8	22.7 ± 9.5	0.008
ECOG grade	3.0 ± 1.1	3.3 ± 0.9	<0.001
ICU admission diagnosis†
Respiratory, n (%)	642 (65.0%)	378 (70.1%)	0.044
Cardiovascular, n (%)	191 (19.4%)	92 (17.1%)	0.273
Acute kidney injury, n (%)	163 (16.5%)	106 (19.7%)	0.123
Sepsis, n (%)	152 (15.4%)	89 (16.5%)	0.569
Neurologic, n (%)	47 (4.8%)	23 (4.3%)	0.659
Others††, n (%)	123 (12.5%)	74 (13.7%)	0.480
CPR within 24 hours of ICU admission, n (%)	128 / 971 (13.2%)	73 / 528 (13.8%)	0.727
Referred to palliative care within 24 hours of ICU admission, n (%)	84 / 962 (8.7%)	51 / 528 (9.7%)	0.551

* Values are presented as number/total number (%) for categorical variables or mean ± standard deviation for continuous variables.

*APACHE II score, Acute Physiology and Chronic Health Evaluation II score; ECOG, Eastern Cooperative Oncology Group; ICU, intensive care unit; CPR, cardiopulmonary resuscitation.

† Multiple choices were available for ICU admission diagnosis.

†† Others included gastrointestinal bleeding, for close observation after surgery or procedure, psychiatric, poisoning and etc.

During the ICU stay, 1,033 (67.7%) patients were mechanically ventilated, and 515 (33.7%) patients died in the ICU. The use of central venous catheters was less frequent in the closed unit model group than in the mandatory critical care consultation model group (38.5% vs. 51.9%; p < 0.001), but the rates of other treatment options including RRT, tracheostomy, and extracorporeal membrane oxygenation (ECMO) did not differ between the staffing models. The occurrence of VAP was significantly lower in the closed unit model group than in the mandatory critical care consultation model group (3.5% vs. 8.6%; p < 0.001). The difference in ICU mortality was not statistically significant between the staffing models (31.2% vs. 35.2%; p = 0.115), but the overall LOS in the ICU was shorter in the closed unit model group than in the mandatory critical care consultation model group (6.4 vs. 7.3 days; p = 0.024). The rates of ICU readmission did not differ between the staffing models (0.7% vs. 1.5%; p = 0.190) (Table 2).

Table 2

The study participants’ treatments and clinical outcomes.

Treatment and outcome	Mandatory critical care consultation model group (n = 987)	Closed unit model group (n = 539)	p value
Invasive procedures
Mechanical ventilation, n (%)	652 / 987 (66.1%)	381 / 539 (70.7%)	0.065
RRT, n (%)	313 / 948 (33.0%)	153 / 514 (29.8%)	0.203
Tracheostomy, n (%)	150 / 945 (15.9%)	77 / 512 (15.0%)	0.675
ECMO, n (%)	36 / 950 (3.8%)	27 / 520 (5.2%)	0.201
Central venous catheter, n (%)	498 / 959 (51.9%)	198 / 514 (38.5%)	<0.001
VAP, n (%)	81 / 939 (8.6%)	18 / 517 (3.5%)	<0.001
Delirium, n (%)	104 / 937 (11.1%)	50 / 516 (9.7%)	0.404
In-ICU mortality, n (%)	347 (35.2%)	168 (31.2%)	0.115
Overall LOS in ICU, days	7.3 ± 8.1	6.4 ± 7.6	0.024
LOS in ICU among survivors	6.7 ± 6.9	5.5 ± 6.0	0.005
LOS in ICU among non-survivors	8.5 ± 9.8	8.4 ± 10.0	0.841
The overall duration of MV care, days	2.5 ± 4.6	2.2 ± 4.2	0.326
MV care duration among survivors	2.5 ± 4.5	2.2 ± 4.1	0.439
MV care duration among non-survivors	2.6 ± 4.9	2.1 ± 4.4	0.537
ICU readmission rate, n (%)	15 (1.5%)	4 (0.7%)	0.190

*Values are presented as number/total number (%) for categorical variables or mean ± standard deviation for continuous variables.

*RRT, renal replacement therapy; ECMO, extracorporeal membrane oxygenation; VAP, ventilator-associated pneumonia; ICU, intensive care unit; LOS, length of stay; MV, mechanical ventilator.

2. Factors associated with all-cause ICU mortality

After multiple logistic regression adjusted for age, sex, APACHE II score, and ECOG grade, conversion to the closed unit model decreased ICU mortality by 35% (p < 0.001). A high APACHE II score, ECOG grade, and ICU admission for acute kidney injury or sepsis were independent risk factors for ICU mortality (Table 3).

Table 3

Independent predictors of in-ICU mortality by multiple logistic regression analysis.

Independent variables	Univariate analysis		Multivariate analysis
	Odds ratio	p value	Odds ratio	p value
	(95% CI)		(95% CI)
Age	0.99 (0.99–1.00)	0.059	0.99 (0.98–0.99)	0.001
Sex	1.06 (0.85–1.32)	0.579
APACHE II score	1.08 (1.06–1.09)	<0.001	1.07 (1.05–1.08)	<0.001
(increase in 1 point)
ECOG grade	1.55 (1.38–1.74)	<0.001	1.37 (1.21–1.56)	<0.001
(increase in 1 point)
Closed unit model	0.84 (0.67–1.04)	0.115	0.65 (0.51–0.83)	<0.001
ICU admission diagnosis
Respiratory	1.14 (0.91–1.43)	0.262
Cardiovascular	1.18 (0.90–1.54)	0.237
Acute kidney injury	2.24 (1.71–2.93)	<0.001	1.67 (1.25–2.23)	<0.001
Sepsis	2.74 (2.07–3.63)	<0.001	1.83 (1.34–2.49)	<0.001
Neurologic	1.33 (0.80–2.16)	0.259

*ICU, intensive care unit; APACHE II score, Acute Physiology and Chronic Health Evaluation II score; ECOG, Eastern Cooperative Oncology Group.

3. Factors associated with LOS in the ICU among survivors

We analyzed the association between the staffing model and LOS in the ICU among the 1,011 ICU survivors. After adjusting for age, sex, APACHE II score, and ECOG grade, patients in the closed unit model group had shorter LOS in the ICU by 1.88 days than those in the mandatory critical care consultation model group (p < 0.001). Patients admitted for respiratory failure or sepsis stayed longer in the ICU than other patients (p < 0.001 and p = 0.022, respectively) (Table 4).

Table 4

Independent predictors of ICU length of stay among survivors by multiple regression analysis.

Independent variables	Univariate analysis		Multivariate analysis
	β-coefficient	p value	β-coefficient	p value
Age	-0.03	0.022	-0.06	<0.001
Sex	0.42	0.322
APACHE II score	0.16	<0.001	0.16	<0.001
ECOG grade	0.78	<0.001	0.56	0.004
Closed unit model	-1.20	0.005	-1.88	<0.001
ICU admission diagnosis
Respiratory	2.24	<0.001	2.35	<0.001
Cardiovascular	0.30	0.579
Acute kidney injury	0.31	0.604
Sepsis	1.40	0.034	1.46	0.022
Neurologic	1.68	0.106

*ICU: Intensive Care Unit, APACHE II score: Acute Physiology and Chronic Health Evaluation II score, RRT: Renal Replacement Therapy.

4. Differences in outcomes according to the ICU admission diagnosis

The benefits of the closed unit model were more prominent among patients admitted to the ICU for respiratory or cardiovascular diseases. Among the patients with ICU admission diagnosis of respiratory failure, ICU mortality was lower (30.4% vs. 37.2%; p = 0.028), and the LOS in the ICU was shorter (6.3 ± 6.6 vs. 7.5 ± 7.1; p = 0.027) in the closed unit model group than in the mandatory critical care consultation model group. Similar findings were observed in patients admitted for cardiovascular failure. The LOS in the ICU and the duration of MV care were shorter (4.4 ± 5.4 vs. 7.4 ± 7.6; p = 0.01 and 1.1 ± 3.1 vs. 3.0 ± 4.0; p = 0.002, respectively) in the closed unit model group than in the mandatory critical care consultation model group. A statistically significant difference in outcome was not observed in patients with an admission diagnosis of acute kidney injury, sepsis, or neurological diseases (Table 5).

Table 5

Patient outcomes changes in the five subgroups according to the staffing model of ICU.

Patient outcome	Mandatory critical care consultation model	Closed unit model	p value
Overall patients	n = 987	n = 539
Mortality, n (%)	347 (35.2%)	168 (31.2%)	0.115
LOS in ICU of survivors, days	6.7 ± 6.9	5.5 ± 6.0	0.005
Duration of MV care of survivors, days	2.5 ± 4.5	2.2 ± 4.1	0.439
Patients with admission diagnosis of respiratory disease	n = 642	n = 378
Mortality, n (%)	239 (37.2%)	115 (30.4%)	0.028
LOS in ICU of survivors, days	7.5 ± 7.1	6.3 ± 6.6	0.027
Duration of MV care of survivors, days	3.5 ± 5.2	2.9 ± 4.7	0.187
Patients with admission diagnosis of cardiovascular disease	n = 191	n = 92
Mortality, n (%)	67 (35.1%)	37 (40.2%)	0.401
LOS in ICU of survivors, days	7.4 ± 7.6	4.4 ± 5.4	0.010
Duration of MV care of survivors, days	3.0 ± 4.0	1.1 ± 3.1	0.002
Patients with admission diagnosis of acute kidney injury	n = 163	n = 106
Mortality, n (%)	84 (51.5%)	49 (46.2%)	0.395
LOS in ICU of survivors, days	6.3 ± 4.7	6.8 ± 6.8	0.549
Duration of MV care of survivors, days	1.7 ± 5.2	2.6 ± 5.1	0.333
Patients with admission diagnosis of sepsis	n = 152	n = 89
Mortality, n (%)	84 (55.3%)	46 (51.7%)	0.591
LOS in ICU of survivors, days	7.7 ± 6.9	7.1 ± 7.1	0.644
Duration of MV care of survivors, days	2.1 ± 3.7	2.4 ± 3.7	0.753
Patients with admission diagnosis of neurologic disease	n = 47	n = 23
Mortality, n (%)	17 (36.2%)	11 (47.8%)	0.350
LOS in ICU of survivors, days	9.3 ± 9.2	4.3 ± 3.2	0.074
Duration of MV care of survivors, days	3.5 ± 9.2	1.0 ± 2.0	0.405

*Values are presented as number (%) for categorical variables or mean ± standard deviation for continuous variables.

*ICU, intensive care unit; LOS, length of stay; MV, mechanical ventilator.

Discussion

Our study revealed that the closed unit model decreased ICU mortality and shortened LOS of critically ill patients compared to the mandatory critical care consultation model. Although many studies have suggested the superiority of the high-intensity staffing models over the low-intensity models, most studies have been conducted in the United States [4–9, 17, 18]. Only a few studies have been conducted in Asia, but with a limited number of surgical ICU patients [10] or postoperative patients [11]. A Turkish study showed an improved survival rate after conversion to a closed unit model, but the sample size was relatively small (<40% of this study population) and a detailed description of the participants’ characteristics was not provided [12]. A retrospective Japanese study reported better survival in patients with sepsis in the closed unit model than in the open unit model among 35 heterogeneous ICUs, but detailed description of each ICU closed unit was not available and the results could not be applied to the general ICU population [13].

One of the meaningful findings of our study is that the closed unit model was associated with better outcomes than the mandatory critical care consultation model involving the same intensivists. Most of the studies reporting the beneficial effect of a closed unit model compared it to an open unit model. Further, they did not indicate whether the open unit model was a mandatory or elective critical care consultation model, or a no intensivist model [4, 5, 10–12, 14]. It was also unclear whether there were changes in the intensivists involved in patient care. This is also a limitation of the studies that include multiple institutes [7–9, 13, 17, 18].

Although the patients had higher APACHE II scores at admission, the closed unit model led to improved outcomes in critically ill patients. Although speculative, the active admission triage of the intensivists in the closed unit model might have resulted in the admission of patients with higher severity. The improved ICU outcomes in the closed unit model might be due to the decrease in ICU care complications. Although the invasive treatment/procedures performed in both groups were not different, the frequency of central venous catheter insertion was significantly lower in the closed unit model group than in the mandatory critical care consultation model group. This may have led to a lower rate of ICU-acquired infections. This assumption was further supported by the significantly lower VAP, another frequent ICU-acquired infection, in the closed unit model group than in the mandatory critical care consultation model group. Other studies have also shown a reduced VAP rate after conversion to a closed unit model [19, 20].

Subgroup analysis in this study showed that the benefits of the closed unit model were mostly found in patients with respiratory or cardiovascular failure. Timely application and handling of equipments by specialists, including mechanical ventilators, noninvasive ventilators, and ECMO, may be associated with the beneficial results in these subgroups [21-24]. Early access to diagnostic tools such as bronchoscopy, echocardiography, and ultrasonography may also have played a significant role [25-28].

Our study has several limitations. First, due to the before-after observational design of this study, it was possible that factors other than the staffing model may have affected the outcome. Further, the outcome changes might have been due to advances in medicine rather than changes in staffing models [29]. However, the same two intensivists participated in the treatment of enrolled patients in both models. In addition, the proportion of patients referred to palliative care within 24 hours of ICU admission, a possible surrogate for inappropriate admission, was not different between the groups. Second, caution is needed when interpreting the results of a single-center study. ICU conditions differ greatly from one nation to another and also within one nation. However, this study demonstrated that changes in the staffing model can improve outcomes in the medical ICU of an Asian country and identified subgroups that might benefit most from the changes. Third, the number of patients in the mandatory consultation model group was twice as high as that than in the closed unit model group. This was due to differences in time period before and after changing the staffing model. Although equal-sized groups have maximal statistical power, we believe that the smaller size of the group in our study (n = 539) was large enough to detect clinically significant differences between the two groups.

In conclusion, the closed unit model proved to be superior to the mandatory critical care consultation model in terms of ICU mortality and LOS in the ICU. The beneficial effects of the closed unit model were more prominent in patients admitted for respiratory and cardiovascular failure.

7 Jul 2021

PONE-D-21-19587

Impact of staffing model conversion from a mandatory critical care consultation model to a closed unit model in the medical intensive care unit

PLOS ONE

Dear Dr. Lee,

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 address the issues and revise accordingly.

Please submit your revised manuscript by Aug 19 2021 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:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled '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: http://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,

Robert Jeenchen Chen, MD, MPH

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_formatting_sample_main_body.pdf and

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

2. We note that you have indicated that data from this study are available upon request. PLOS only allows data to be available upon request if there are legal or ethical restrictions on sharing data publicly. For information on unacceptable data access restrictions, please see http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions.

In your revised cover letter, please address the following prompts:

a) If there are ethical or legal restrictions on sharing a de-identified data set, please explain them in detail (e.g., data contain potentially identifying or sensitive patient information) and who has imposed them (e.g., an ethics committee). Please also provide contact information for a data access committee, ethics committee, or other institutional body to which data requests may be sent.

b) If there are no restrictions, please upload the minimal anonymized data set necessary to replicate your study findings as either Supporting Information files or to a stable, public repository and provide us with the relevant URLs, DOIs, or accession numbers. Please see http://www.bmj.com/content/340/bmj.c181.long for guidelines on how to de-identify and prepare clinical data for publication. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories.

We will update your Data Availability statement on your behalf to reflect the information you provide.

[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

Reviewer #3: Yes

**********

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

Reviewer #1: I Don't Know

Reviewer #2: Yes

Reviewer #3: Yes

**********

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

The PLOS Data policy 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: Yes

Reviewer #3: 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

Reviewer #3: 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: Lee and colleagues present a restrospective analysis of different clinical outcomes (mainly mortality, length of stay (LOS), rates of central venous catheterization and ventilator associated pneumonia (VAP)) in patients treated in a medical ICU in either a closed unit model (intensivist is the primary physician in charge) or a mandatory critical care consultation model (primary attending physician is in charge + mandatory consultations of an intensivist). Main findings a reduced LOS, central venous catheterization- and VAP-rate, and - in the multivariate analysis - a reduced mortality in the closed unit model compared with the mandatory critical care consultation model. Of course, this study carries with it the innate limitations of retrospective studies, which the authors acknowledge. The manuscript is overall well-written, some minor errors should be corrected.

Reviewer #2: I read with great interest this manuscript on the impact of the ICU staffing model conversion. This is a highly relevant article in the critical care world and I am in agreement with the authors on superiority of the closed unit model and on the unmet need for board-certified intensivists.

I felt the manuscript contained sound statistical analysis and was well-written and succinct in communication of pertinent information. Most of the questions/suggestions that I jotted down during review of the paper were answered and addressed in subsequent sections of the manuscript. Examples include definitions of four staffing models, discussion of limitations, subgroup analysis and characteristics, predictors of mortality, etc.

My final comments are outlined below, but my recommendation is to accept this manuscript for publication with minor revisions. I commend the authors on their rigorous work in this important study.

1. I agree that that ICU characteristics are highly variable among different sites within the same country and internationally. As such, intensivist training and board certification processes differ as well. Since the authors report best outcomes for patients with respiratory and cardiovascular conditions, would you comment on any additional training background of the intensivists in this study? In the US, the pathway to becoming an intensivist is available through critical care fellowship and board certification to anesthesiologists, surgeons, emergency physicians and internal medicine physicians, with the latter group often also co-trained in pulmonology. In other words, these different subgroups of intensivists bring additional skill sets to the ICU.

2. Was the comparison data on rate of ICU re-admissions available between the two staffing models? (i.e. % of patients downgraded from the ICU who were re-admitted to ICU during the same hospital stay). It would be an interesting outcome to track.

3. The “before” and “after” staffing model group sizes are at almost 2:1 ratio (1.8, to be exact). While there were adjustments in statistical analysis, this disparity would be worthwhile to include in the limitations section and discuss its potential effects on results.

4. Discussion section, Limitations, 2nd to last paragraph: “Factors other than staffing model may have affected the outcomes”. Can you provide some examples, either from the current study or from pertinent literature ? Any confounding variables?

5. Discussion section, 1st sentence: “shortened LOS of critically ill patients than the mandatory critical care consultation model”—“than” should be changed to “compared to”.

Reviewer #3: Dear editor, here you receive my review regarding the manuscript entitled ” Impact of staffing model conversion from a mandatory critical care consultation model to a closed unit model in the medical intensive care unit ”, with number PONE-D-21-19587.

The authors describe and present the results of a retrospective before and after cohort study, with patients included who were admitted to the medical ICU in a university teaching hospital. Between January 2016 and August 2017 patients received care on a mandatory consultation basis. From September 2017 till August 2018 patients were received medical care from a closed format unit model. Various indices were compared. The article is well written and easy to understand.

The need for personal informed consent from patients was waived by the medical ethical committee due to the retrospective design of the study.

I have a few comments to make.

There are many improvements made. n abstract and discussion important results are presented, i.e. CVL associated blood stream infections, catheter associated urinary tract infections and ventilator associated pneumonia and the influence on LOS and MV duration.

References are ok.

Page 1 the abstract: Here the order in presentation of either models is confusing, i.e., in the 3rd sentence of results first closed unit model (CUM) and second the mandatory critical care consultation model ( MCCCM) is mentioned. However the order of presentation is changed in the following sentence hen results of CVC and VAP are presented with opponent results. Please choose an order in presentation in the whole article? This will improve reading and understanding what is different. For instance, do I correctly understand that less CVC were used and there were more VAP’s in MCCCM?

The observation periods differ between the 2 models with 7 months more in the mandatory critical care consultation model. Please explain and discuss? This resulted in a different number of patients between groups, i.e. 987 vs 539 patients.

How many patients were not include in these 2 periods? In other words , what was the ratio or sample size of this cohort presented/studied in comparison with the total number of patients who were admitted tot ICU?

What was exactly the assumption or hypothesis before starting the study?

Ad Introduction, Line13 please write Pronovost instead of Pronovist

Ad Methods: L9 …during day and was responsible…? Is here missing maybe “during day and night, and was…” Do you mean 7 x 24 hours closed format/dedication, and weekend included?

Were there any differences regarding the time of discharge, i.e. outside time schedule 8:00-18:00 hours, with a possible association in (repeated) admission within 48 hours after discharge?

Ad statistical analysis: L7… to the five reasons… Here please describe in short what is meant? For instance: In short, ……..are the 5 organ related failures plus others?

Ad Results L7: patients in CUM had higher APPACHE II scores. Please explain? Could it be possible that with more dedication of the intensivists present there was more time to score the APACHE better, which could have led to the this difference?

Ad Table 1 ECOG-scale is presented as significantly different between groups, but with a mean of 3.0 vs 3.3. Statistically different, but I doubt whether of any clinical importance. Is it possible to divide between patients with ECOG≤2 and ≥3 for either group?

Table 1 Others with a remark, with stated “etc” What is the quantity of etc and what is exactly meant by etc? In total it is >12.5%

So LOS and MV duration are in favor of MCCCM. This is important for your ICU and conform previous study results. Whereas, when looking at patients with neurologic disorders (Table 5) the total number of patients may have not been enough to reach significance (p=0.074) in favor of MCCCM. Please explain and describe?

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). 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 Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Reviewer #3: Yes: P. Bruins

[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.

15 Aug 2021

Dear Dr. Emily Chenette,

Editor-in-Chief of PLoS One

We wish to thank the editor and reviewers for their response to our work and their constructive comments. We have addressed the concerns that were raised through further editing of the manuscript and believe it has significantly improved as a consequence of this process. We have highlighted all changes in the revised manuscript and have addressed all referees’ comments in the point-by-point responses below. We greatly appreciate the opportunity to submit this revised review manuscript and hope it is suitable for publication in PLoS One.

Reviewer #1:

Lee and colleagues present a retrospective analysis of different clinical outcomes (mainly mortality, length of stay (LOS), rates of central venous catheterization and ventilator associated pneumonia (VAP) in patients treated in a medical ICU in either a closed unit model (intensivist is the primary physician in charge) or a mandatory critical care consultation model (primary attending physician is in charge + mandatory consultations of an intensivist). Main findings a reduced LOS, central venous catheterization- and VAP-rate, and - in the multivariate analysis - a reduced mortality in the closed unit model compared with the mandatory critical care consultation model. Of course, this study carries with it the innate limitations of retrospective studies, which the authors acknowledge. The manuscript is overall well-written, some minor errors should be corrected.

� We thank the reviewer for this valuable comment. As suggested, we have gone through the revised article and corrected minor errors.

Introduction 13th line: Corrected “Pronovist” to “Pronovost”

Discussion 2nd line: Corrected “than” to “compared to”

Method: Study design and participants 10th line: Corrected “during day” to “during the weekday daytime”

Reviewer #2:

I read with great interest this manuscript on the impact of the ICU staffing model conversion. This is a highly relevant article in the critical care world and I am in agreement with the authors on superiority of the closed unit model and on the unmet need for board-certified intensivists.

I felt the manuscript contained sound statistical analysis and was well-written and succinct in communication of pertinent information. Most of the questions/suggestions that I jotted down during review of the paper were answered and addressed in subsequent sections of the manuscript. Examples include definitions of four staffing models, discussion of limitations, subgroup analysis and characteristics, predictors of mortality, etc.

My final comments are outlined below, but my recommendation is to accept this manuscript for publication with minor revisions. I commend the authors on their rigorous work in this important study.

1. I agree that that ICU characteristics are highly variable among different sites within the same country and internationally. As such, intensivist training and board certification processes differ as well. Since the authors report best outcomes for patients with respiratory and cardiovascular conditions, would you comment on any additional training background of the intensivists in this study? In the US, the pathway to becoming an intensivist is available through critical care fellowship and board certification to anesthesiologists, surgeons, emergency physicians and internal medicine physicians, with the latter group often also co-trained in pulmonology. In other words, these different subgroups of intensivists bring additional skill sets to the ICU.

� We thank the reviewer for this insightful comment. The pathway to becoming an intensivist is similar in South Korea. Anesthesiologists, surgeons, emergency physicians, neurologists, and internal medicine physicians can become an intensivist after adequate training through critical care fellowship. Most internal medicine physicians receive training in pulmonology before critical care fellowship. Both intensivists in this study were board certified in Internal Medicine and Pulmonology and we added this information in 10th line of Methods: Study design and participants.

2. Was the comparison data on rate of ICU re-admissions available between the two staffing models? (i.e. % of patients downgraded from the ICU who were re-admitted to ICU during the same hospital stay). It would be an interesting outcome to track.

� We agree with the reviewer that ICU readmission is one of the important outcomes. We assessed rate of readmission (defined as readmission within 48 hours of ICU discharge), and found that although the rate of readmission was lower in the closed unit model (0.7% vs. 1.5%; p = 0.190), the number was too small to show any statistical significance. We added this result to the manuscript: 26-27 lines of Results: Study population and treatment in the ICU and Table 2.

3. The “before” and “after” staffing model group sizes are at almost 2:1 ratio (1.8, to be exact). While there were adjustments in statistical analysis, this disparity would be worthwhile to include in the limitations section and discuss its potential effects on results.

� We thank the reviewer for this valuable comment. The duration of “before” staffing model was about several months longer than the “after” staffing model. This difference was the result of the timing of our study enrollment. We planned this study as soon as we believed we have gathered enough data to compare the different staffing models.

We understand the reviewer’s concern and added this as a limitation of this study (50-54th lines of Discussion).

4. Discussion section, Limitations, 2nd to last paragraph: “Factors other than staffing model may have affected the outcomes”. Can you provide some examples, either from the current study or from pertinent literature? Any confounding variables?

� We thank the reviewer for this valuable comment. Although there were no confounding variables found during the retrospective analysis, the before-after observational design of this study in itself has its limitations. We cannot completely exclude the possibility that outcome improvements may be from advances in medicine (Curr Anesthesiol Rep 2013;3:65-72) rather than changes in staffing models. For example, decreased use of central venous catheter in this study may be the result of studies that have shown usual care is not inferior to early, goal-directed therapy in septic shock. (N Engl J Med 2017;376:2223-2234). We have added this explanation in the discussion section (41-42th lines of Discussion).

5. Discussion section, 1st sentence: “shortened LOS of critically ill patients than the mandatory critical care consultation model”—“than” should be changed to “compared to”.

� Thank you for correcting our mistake. Changes were made in 2nd line of Discussion as suggested.

Reviewer #3:

Dear editor, here you receive my review regarding the manuscript entitled ” Impact of staffing model conversion from a mandatory critical care consultation model to a closed unit model in the medical intensive care unit ”, with number PONE-D-21-19587.

The authors describe and present the results of a retrospective before and after cohort study, with patients included who were admitted to the medical ICU in a university teaching hospital. Between January 2016 and August 2017 patients received care on a mandatory consultation basis. From September 2017 till August 2018 patients were received medical care from a closed format unit model. Various indices were compared. The article is well written and easy to understand.

The need for personal informed consent from patients was waived by the medical ethical committee due to the retrospective design of the study.

I have a few comments to make.

There are many improvements made. n abstract and discussion important results are presented, i.e. CVL associated blood stream infections, catheter associated urinary tract infections and ventilator associated pneumonia and the influence on LOS and MV duration.

References are ok.

* Page 1 the abstract: Here the order in presentation of either models is confusing, i.e., in the 3rd sentence of results first closed unit model (CUM) and second the mandatory critical care consultation model (MCCCM) is mentioned. However the order of presentation is changed in the following sentence hen results of CVC and VAP are presented with opponent results. Please choose an order in presentation in the whole article? This will improve reading and understanding what is different. For instance, do I correctly understand that less CVC were used and there were more VAP’s in MCCCM?

� We apologize for the confusion. Both central venous catheter insertion and VAP were less frequent in the closed unit model (CUM). We modified the order of the mentioned models for better understanding, in both the abstract and Results section (14-16th lines of Abstract and 17-19th lines of Results: Study population and treatment in ICU).

* The observation periods differ between the 2 models with 7 months more in the mandatory critical care consultation model. Please explain and discuss? This resulted in a different number of patients between groups, i.e. 987 vs 539 patients.

� We thank the reviewer for this valuable comment. The duration of “before” staffing model was about several months longer than the “after” staffing model and therefore, almost twice as more patients were enrolled in the mandatory consultation model compared to the closed unit model. This difference was the result of the timing of our study enrollment. We planned this study as soon as we believed we have gathered enough data to compare the different staffing models.

We understand the reviewer’s concern and added this as a limitation of this study. “Third, there are almost twice as more patients in the mandatory consultation model compared to the closed unit model. This was due to differences in the duration of the before and after staffing models. While equal-sized groups have maximal statistical power, we believe that the smaller group of our study (n=539) was still large enough sample to detect clinical significant differences between the two groups.” (50-54th lines of Discussion).

* How many patients were not include in these 2 periods? In other words , what was the ratio or sample size of this cohort presented/studied in comparison with the total number of patients who were admitted to ICU?

� In total, 1,076 patients in mandatory critical care consultation model group and 581 patients in closed unit model group were admitted to ICU during study period. Of them, 89 (8.3%) patients in mandatory critical care consultation model group and 42 (7.2%) patients in closed unit model group) were excluded due to incomplete medical records. The proportion of excluded patients in each group does not seem to differ significantly. We added this to the manuscript in 1-4th lines of Results: Study population and treatment in the ICU.

* What was exactly the assumption or hypothesis before starting the study?

� The aim of this study was to compare closed unit model to mandatory critical care consultation model and to evaluate the superiority of the closed unit model in the outcomes of critically ill Asian patients. We have shown in a large-scale Asian population that closed unit model is associated with lower ICU mortality and shorted ICU LOS. We have added this comment in the 27-30th lines of Introduction.

* Ad Introduction, Line13 please write Pronovost instead of Pronovist

� Thank you for correcting our mistake. Changes were made in 13th line of Introduction as suggested.

* Ad Methods: L9 …during day and was responsible…? Is here missing maybe “during day and night, and was…” Do you mean 7 x 24 hours closed format/dedication, and weekend included?

� We apologize for the confusion. The intensivists of this study stayed in the ICU only during the weekday daytime, and not during the weekends and night hours. We have corrected 'during day' to 'during the weekday daytime'.

* Were there any differences regarding the time of discharge, i.e. outside time schedule 8:00-18:00 hours, with a possible association in (repeated) admission within 48 hours after discharge?

� We thank the reviewer for this comment. Unfortunately, data regarding the time of discharge was not available for analysis. The rate of readmission (defined as readmission within 48 hours of ICU discharge) was not significantly different between the two groups. We added this result to the manuscript (26-27 lines of Results: Study population and treatment in the ICU and Table 2).

* Ad statistical analysis: L7… to the five reasons… Here please describe in short what is meant? For instance: In short, ……..are the 5 organ related failures plus others?

� We clarified the five reasons in the Methods section as “the reasons of ICU admission: respiratory diagnosis, cardiovascular diagnosis, acute kidney injury, sepsis, and neurologic diagnosis.”

* Ad Results L7: patients in CUM had higher APACHE II scores. Please explain? Could it be possible that with more dedication of the intensivists present there was more time to score the APACHE better, which could have led to the this difference?

� The APACHE II scores used in the study was measured by the charge nurse on duty. Therefore, differences in APACHE II scores cannot be explained by dedicated intensivists better scoring the patients. It is more likely that active admission triage of the intensivists in the closed unit model led to the admission of patients with higher severity. We added this explanation in the 21-24th lines of Discussion.

* Ad Table 1 ECOG-scale is presented as significantly different between groups, but with a mean of 3.0 vs 3.3. Statistically different, but I doubt whether of any clinical importance. Is it possible to divide between patients with ECOG≤2 and ≥3 for either group?

� We thank the reviewer for this valuable comment. Proportion of patients with ECOG grade ≥3 was also significantly higher in the closed unit model group compared to the mandatory critical care consultation group (84.8% vs. 73.0%; p < 0.001) and we have added this information in the Results section lines 10-11. However, taking into account that mean ECOG grade was used in the multiple regression analysis, we ask the reviewer to consider leaving ECOG grade in Table 1 as it is.

* Table 1 Others with a remark, with stated “etc” What is the quantity of etc and what is exactly meant by etc? In total it is >12.5%

� In the medical records, the admission diagnosis of the patients admitted to the ICU was coded into 6 categories (respiratory, cardiovascular, acute kidney injury, sepsis, neurologic, and others). Others included gastrointestinal bleeding, for close observation after surgery or procedure, psychiatric, poisoning and etc. We have added this description under Table 1. We apologize for not being able to describe the ‘others’ category in detail due to the retrospective design of this study.

* So LOS and MV duration are in favor of MCCCM. This is important for your ICU and conform previous study results. Whereas, when looking at patients with neurologic disorders (Table 5) the total number of patients may have not been enough to reach significance (p=0.074) in favor of MCCCM. Please explain and describe?

� LOS and MV duration was mostly shorter in the closed unit model compared to the mandatory critical care consultation model. We agree with the reviewer that statistical significance could not be obtained in several specific admission diagnosis groups (including neurologic diagnosis) due to the lack of sample size. In total, LOS in ICU of survivors was significantly shorter in the closed unit model (5.5 days vs. 6.7 days; p = 0.005) compared to the mandatory critical care consultation model group and these results are consistent with other studies (Multz et al. Am J Respir Crit Care Med 1998;157:1468-73, Ogura et al. J Intensive Care 2018;6:57).

Submitted filename: Closed unit Revision response Final (PLoS One) 0717.docx

Click here for additional data file.

10 Sep 2021

PONE-D-21-19587R1Impact of staffing model conversion from a mandatory critical care consultation model to a closed unit model in the medical intensive care unitPLOS ONE

Dear Dr. Lee,

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 revise accordingly.

Please submit your revised manuscript by Oct 25 2021 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.

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

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,

Robert Jeenchen Chen, MD, MPH

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

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

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: I Don't Know

Reviewer #2: Yes

**********

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

The PLOS Data policy 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: Yes

**********

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: I would like to thank the authors for addressing the reviewers' comments. In my opinion, a final check for linguistic flaws (e. g. "twice as more patients" in the new paragraph of the limitations section schould be changed to "twice as many") ought to be the last step before acceptance.

Reviewer #2: Thank you for submitting your revisions and addressing all recommendations by reviewers. No further suggestions on my end.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). 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 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.

24 Sep 2021

Dear Dr. Emily Chenette,

Editor-in-Chief of PLoS One

We wish to thank the editor and reviewers for their response to our work and their constructive comments. We have addressed the concerns that were raised through further editing of the manuscript and believe it has significantly improved as a consequence of this process. We have highlighted all changes in the revised manuscript and have addressed all referees’ comments in the point-by-point responses below. We greatly appreciate the opportunity to submit this revised review manuscript and hope it is suitable for publication in PLoS One.

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

Reviewer #1: No

� We apologize that the data cannot be shared publicly due to potentially identifying or sensitive information. Requests for data access may be sent to Seoul National University Institutional Data Access / Ethics Committee (contact via Tel: 82-2-2072-0694) for researchers who meet the criteria for access to confidential data.

6. Review Comments to the Author

Reviewer #1: I would like to thank the authors for addressing the reviewers' comments. In my opinion, a final check for linguistic flaws (e. g. "twice as more patients" in the new paragraph of the limitations section schould be changed to "twice as many") ought to be the last step before acceptance.

� We thank the reviewer for this valuable comment. We performed grammatical corrections on the revised text, and as a result, the following sentences were corrected.

Introduction 27-30th line: Corrected “The aim of this study was to compare closed unit model to mandatory critical care consultation model and to evaluate the superiority of the closed unit model in the outcomes of critically ill Asian patients.” to “This study aimed to compare the closed unit model to the mandatory critical care consultation model and evaluate the superiority of the closed unit model in terms of outcomes in critically ill Asian patients.”

Result 1st-2nd line: Corrected “1,076 patients in mandatory critical consultation model group and 581 patients in closed unit model group” to “1,076 and 581 patients in the mandatory critical consultation model group and the closed unit model group, respecively”

Result 3rd-5th line: Corrected “89 [8.3%] patients in mandatory critical care consultation model group and 42 [7.2%] patients in closed unit model group” to “89 [8.3%] and 42 [7.2%] patients in mandatory critical care consultation model group and the closed unit model group, respectively”

Discussion 21th line: Corrected “Despite patients having higher APACHE II scores at admission” to “Although the patients had higher APACHE II scores at admission”

Discussion 41-42th line: Corrected “It is impossible to completely exclude the possibility that outcome changes may have been from advances in medicine rather than changes in staffing models” to “Further, the outcome changes might have been due to advances in medicine rather than changes in staffing models”

Discussion 49-54th line: Corrected “Third, there are almost twice as more patients in the mandatory consultation model compared to the closed unit model. This was due to differences in the duration of the before and after staffing models. While equal-sized groups have maximal statistical power, we believe that the smaller group of our study (n=539) was still large enough sample to detect clinical significant differences between the two groups” to “Third, the number of patients in the mandatory consultation model group was twice as high as that than in the closed unit model group. This was due to differences in time period before and after changing the staffing model. Although equal-sized groups have maximal statistical power, we believe that the smaller size of the group in our study (n=539) was large enough to detect clinically significant differences between the two groups”.

Reviewer #2: Thank you for submitting your revisions and addressing all recommendations by reviewers. No further suggestions on my end.

� Thanks for your thoughtful comment.

Submitted filename: Closed unit Revision response Final (PLoS One) 0925.docx

Click here for additional data file.

13 Oct 2021

Impact of staffing model conversion from a mandatory critical care consultation model to a closed unit model in the medical intensive care unit

PONE-D-21-19587R2

Dear Dr. Lee,

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,

Robert Jeenchen Chen, MD, MPH

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 #2: All comments have been addressed

Reviewer #4: 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 #2: Yes

Reviewer #4: Yes

**********

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

Reviewer #2: Yes

Reviewer #4: Yes

**********

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

The PLOS Data policy 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 #2: Yes

Reviewer #4: Yes

**********

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 #2: Yes

Reviewer #4: 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 #2: Thank you for resubmitting your revisions in response to reviewers' comments. I have no additional suggestions.

Reviewer #4: Although the concept that closed ICU are better than open ICU is not novel and it has been well demonstrated in numerous studies over the past 2 decades leading to a closed ICU model at many if not most US hospitals, the authors make an interesting point that this is one of the few to study in detail the two models in Asia. This article is well written and would be of value to the literature as it validates previous findings in a south east Asia where medical training and practices may differ considerably compared to the US and other countries were the previous studies were conducted.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). 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 Privacy Policy.

Reviewer #2: No

Reviewer #4: No

19 Oct 2021

PONE-D-21-19587R2

Impact of staffing model conversion from a mandatory critical care consultation model to a closed unit model in the medical intensive care unit

Dear Dr. Lee:

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

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 plosone@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. Robert Jeenchen Chen

Academic Editor

PLOS ONE