Impact of Prefecture-level Intensive Care Unit Congestion on Mortality in Severe COVID- 19 Patients: A Retrospective Observational Study in Japan

Impact of Prefecture-level Intensive Care Unit Congestion on Mortality in Severe COVID- 19 Patients: A Retrospective Observational Study in Japan | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Impact of Prefecture-level Intensive Care Unit Congestion on Mortality in Severe COVID- 19 Patients: A Retrospective Observational Study in Japan Yudai Iwasaki, Takayuki Ogura, Hiroyuki Ohbe, Satoru Hashimoto, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5667123/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 28 Oct, 2025 Read the published version in BMC Infectious Diseases → Version 1 posted 4 You are reading this latest preprint version Abstract Background The COVID-19 pandemic has placed unprecedented pressure on global healthcare systems, severely affecting the intensive care unit (ICU) capacity. Therefore, this study explored the association between prefecture-level ICU congestion and COVID-19 mortality in each prefecture of Japan. Methods This retrospective study analyzed data from the CRoss Icu Searchable Information System, covering all patients with COVID-19 who required mechanical ventilation or extracorporeal membrane oxygenation between January 1, 2020, and March 31, 2023. Prefecture-level ICU congestion was calculated as the total ventilator days over 2 weeks for severely ill patients with COVID-19, divided by the maximum potential ventilator days in the prefecture. Prefecture-level ICU congestion in each region was visualized by plotting time-series graphs capturing the temporal progression of congestion levels. A mixed-effects logistic regression model was fitted to evaluate the association between prefecture-level ICU congestion and mortality. Results A total number of 10,046 patients were included in this study, and the mortality rate was 23.2%. Congestion levels varied by time and prefecture, ranging from 0 to over 1.5, indicating a strain beyond capacity. Median congestion level (interquartile range [IQR]) at admission was 0.3 (0.1, 0.6), and increased congestion level was statistically associated with increased mortality (odds ratio: 1.14; 95% confidential interval: 1.08–1.21). Conclusions Increased prefecture-level ICU congestion may be associated with increased COVID-19 mortality, warranting further investigation. Japan prefectures COVID-19 mortality ICU congestion ECMO Mixed effect model Figures Figure 1 Figure 2 Figure 3 Introduction The COVID-19 pandemic has placed unprecedented pressure on global healthcare systems, exposing weaknesses in public health infrastructure and resource allocation [ 1 , 2 ]. Intensive care units (ICUs) have faced severe strain owing to surges in critically ill patients, pushing their capacity beyond limits in many regions [ 1 ]. Studies have demonstrated that a lack of ICU preparedness and surge capacity is closely associated with increased mortality rates, highlighting the critical importance of ICU resources during health crises [ 3 ]. ICU capacity is a vital component in managing critically ill patients, particularly during a pandemic. A global study reported that countries significantly increased ICU bed numbers to mitigate mortality from COVID-19; however, rapid scaling often fell short of demand, resulting in delays in critical care [ 3 , 4 ]. Hospital ICU congestion has been directly linked to poor patient outcomes, including increased mortality, primarily due to delayed admission and inadequate staffing ratios [ 5 , 6 ]. Variations in ICU infrastructure, such as bed number, intensivist availability, and nurse-to-patient ratio, significantly influence patient outcomes in various regions [ 7 , 8 ]. Japan experienced ICU strains during the pandemic, particularly in prefectures with fewer ICU beds and medical personnel. The distribution of ICU resources across prefectures in Japan has become a critical issue, as regions with limited capacity experienced COVID-19 mortality [ 9 ]. Although studies have addressed the ICU resource challenges in Japan [ 10 ], analyses of the specific impact of regional ICU congestion on COVID-19 mortality rates remain insufficient. Therefore, we hypothesized that regional ICU congestion increased the patients' mortality. This study aimed to analyze ICU capacity strains at the prefecture level in Japan, with administrative divisions organized into 47 prefectures, and their effects on patient outcomes during the COVID-19 pandemic. Methods Study design and data source This study was a retrospective cohort analysis using data from CRoss Icu Searchable Information System (CRISIS), the most extensive national repository for patients with severe COVID-19 data in Japan (UMIN000041450, Registration date: 2020/08/18) [ 11 ]. CRISIS was created by the Japan Extracorporeal Membrane Oxygenation Network specifically for the COVID-19 outbreak, facilitating nationwide tracking of real-time data from ICUs during the pandemic; it encompasses over 6,600 ICU beds, representing approximately 90% of all ICU beds across Japan [ 12 ]. CRISIS collected data on every COVID-19 patient who required mechanical ventilation or extracorporeal membrane oxygenation (ECMO) from January 2020 to March 2023 across all participating hospitals. This study was approved by the Hiroshima University Epidemiological Research Ethics Review Committee (approval numbers E2022-0118) and registered with the University Hospital Medical Information Network Clinical Trials Registry, Japan (UMIN trial ID: UMIN000041450). In accordance with Japanese government guidelines, the Institutional Review Board exempted this study from the requirement for written informed consent to protect participant anonymity. Additionally, this study adheres to the standards set by the REporting of Studies Conducted using Observational Routinely Collected Health Data statement [ 13 ]. Study population From January 1, 2020, to March 31, 2023, we included all patients with COVID-19 registered in the CRISIS database who required mechanical ventilation. Patients who lacked complete records for age, sex, height, weight, start and end dates of mechanical ventilation, and clinical prognosis were excluded. Geographic information in Japan Japan is divided into 47 prefectures, which serve as the primary administrative units. These prefectures are grouped into eight larger geographical regions based on their location and historical ties: Hokkaido, Tohoku, Kanto, Chubu, Kinkii, Chugoku, Shikoku, and Kyushu. Each region comprises multiple prefectures; for example, the Tohoku region has six prefectures in northeastern Japan, while the Kanto region includes Tokyo and six surrounding prefectures (Supplemental Table 1). These regional groupings were used to represent broader geographical trends in this analysis. Visualization of each prefecture capacity The number of COVID-19 beds per prefecture was determined by aggregating the number of beds registered at each facility in the CRISIS database. These counts were then normalized per 100,000 inhabitants based on population data derived from the results of the Japanese census [ 14 ]. The analysis involved visualizing the distribution of COVID-19 beds per 100,000 inhabitants across prefecture-population categories: “Low”, “Middle”, and “High”, representing the lower, middle, and upper thirds of the population distribution, respectively. Each prefecture's name was plotted on the x-axis, and the number of COVID-19 beds per 100,000 people was plotted on the y-axis. Points in the scatter plot were color-coded based on the population category, reflecting the population size of each prefecture. The scatter plot visually compares how bed numbers vary with population size across prefectures. The data were then segregated by population category to compute the number of COVID-19 beds per 100,000 individuals in each category. Definition of the level of ICU congestion Calculations of prefecture-level ICU congestion were performed for the early and late stages of each month. The congestion metric was derived from the methodologies used in previous studies [ 15 ]. The following formula was employed to calculate ICU congestion: Prefecture-level ICU congestion = total number of ventilator days for severely ill patients with COVID-19 admitted during the 2 weeks in a prefecture/maximum possible ventilator days for patients with COVID-19 patients with mechanical ventilation in a prefecture (calculated as the number of beds multiplied by 14 days). The number of beds available for COVID-19 patients in a prefecture was calculated by adding the number of beds available for COVID-19 patients in each hospital belonging to a prefecture. The number of beds available for COVID-19 patients requiring ventilation in each hospital was sourced from the registered information initially entered when the CRISIS database was first created. The congestion level for each patient was determined by extracting prefecture-level ICU congestion during their corresponding hospitalization period and assigned as the “congestion level at admission.” Visualization of discrepancy in prefecture-level ICU congestion We visualized prefecture-level ICU congestion in each region by plotting time-series graphs capturing the temporal progression of congestion levels. Initially, we calculated the national average congestion by aggregating and averaging the “ICU congestion” variable across all regions for each date, providing a benchmark against which regional performances can be compared. Subsequently, we generated individual plots for each region using the ‘ggplot2’ package in R. Each plot displayed the region-specific congestion trends along with the national average, depicted as a solid black line, to contextualize the regional data within the broader national scenario. Hokkaido was grouped with the Tohoku region (six northeastern prefectures) for visualization purposes, as Hokkaido alone constitutes a single prefecture, whereas other regions consist of multiple prefectures. To visualize the discrepancies in congestion levels among prefectures in Japan, we created a thematic map using geographic and congestion data. To create the map, we selected two months when the number of patients with severe COVID-19 was higher than that in other periods in this database. The map is plotted with shaded areas representing each prefecture and colored according to the average congestion levels for the month. A color gradient was applied to denote congestion, ranging from green, which indicates low congestion, to red, which indicates high congestion. This gradient effectively illustrates the range of congestion levels from 0 to 1.5. Data collection and main outcome This dataset encompasses various patient details, including age, sex, body mass index (BMI), ECMO usage, prone position, duration of mechanical ventilation, and mortality. BMI was categorized into five groups: underweight (BMI < 18.5), low normal (BMI 18.5–22.5), normal (BMI 22.5–25), overweight (BMI 25–30), and obese (BMI ≥ 30). Each patient listed in the CRISIS database was monitored until discharge, transfer to another facility, or death. Pandemic waves were determined according to the information given by Ministry of Health, Labour and Welfare, Japan [ 16 ]. The primary outcome was the in-hospital mortality. Statistical analyses Data are reported as either the mean and standard deviation (SD) or the median and interquartile range (IQR) for continuous variables and as count and percentage (n [%]) for categorical variables. COVID-19 beds per 100,000 inhabitants across prefecture–population categories were calculated. The Kruskal–Wallis test was used to assess whether there were differences in the number of beds across the three population categories. To address the non-linear relationship between prefecture-level ICU congestion levels at admission and patient outcomes, we used a generalized additive model with a logistic link function [ 17 ]. Our model incorporates a spline function for ICU congestion using a cubic spline basis to flexibly model its effects on the log odds of death [ 18 ]. In addition to ICU congestion, the model was controlled for several covariates, including sex, age, body mass index category, ECMO use, prone positioning, and pandemic waves. After confirming the relationship between prefecture-level ICU congestion at admission and mortality, we used mixed-effects logistic regression models to evaluate the impact of ICU congestion and other factors on the prognosis of death among patients with COVID-19. For the main analysis, we constructed a mixed-effects model, including congestion level, sex, age, BMI category, ECMO use, prone position, and pandemic waves as fixed effects, with random intercepts for each prefecture-level to account for the individual tendencies in each prefecture. Continuous variables, such as age and congestion level, were standardized, mean-centered, and scaled by their standard deviations to facilitate model convergence and interpretability. To enhance the robustness of our findings, we conducted a sensitivity analysis using an extended mixed-effects model. This model included all the predictors of the main model, with the addition of random intercepts at both facility and prefecture levels. Both models were fitted using the ‘glmer’ function from the lme4 package in R [ 19 ]. All statistical tests were two-tailed. The primary outcome of mortality was analyzed using a mixed-effects logistic regression to estimate the odds ratios (ORs) for the likelihood of death, accompanied by 95% confidence intervals (CIs). The significance level for all tests was set at 0.05, with p-values < 0.05 considered statistically significant. All statistical analyses were conducted using R (version 4.3.1) (2023-06-16). Results Patient flow chart and patients characteristics From January 1, 2020, to March 31, 2023, 12,279 patients were eligible for this study. We excluded 2,233 patients due to missing values for one or more confounding factors, resulting in 2915 missing points and a final analysis of 10,046 patients (Fig. 1 ). Table 1 shows the characteristics of the study cohort. The mean age was 62.9 years, and approximately 75% of the patients were male. Prone positioning was needed in 45% of the study cohort, and ECMO was introduced in 12.4% of the patients. The median congestion level at admission was 0.3 [IQR: 0.1–0.6]. The distribution of patients across pandemic waves showed that the 3rd and 4th waves accounted for the largest proportions, at 24.4% (n = 2,447 and n = 2,453, respectively). The 5th wave also contributed significantly, accounting for 22.8% (n = 2,291) of the total. In contrast, the 1st and 2nd waves had relatively smaller contributions, at 5.5% (n = 548) and 5.2% (n = 522), respectively. Table 1 Patient characteristics Variables Age, yrs, mean (SD) 62.9 (14.5) Sex, male, n (%) 7494 (74.6) Height, cm, median [IQR] 166.0 [159.0, 171.0] Weight, kg, median [IQR] 69.7 [60.0, 80.0] BMI category, n (%) < 18.5 507 (5.0) 18.5–22.4 1967 (19.6) 22.5–24.9 2288 (22.8) 25.0–29.9 1916 (19.1) ≥ 30.0 3368 (33.5) Prone position, n (%) 4518 (45.0) ECMO, n (%) 1241 (12.4) Congestion level at admission, median [IQR] 0.3 [0.1, 0.6] Pandemic wave, n (%) 1st wave 5.5 (548) 2nd wave 5.2 (522) 3rd wave 24.4 (2447) 4th wave 24.4 (2453) 5th wave 22.8 (2291) 6th wave 9.0 (906) 7th wave 4.3 (432) 8th wave 4.4 (447) SD: standard deviation; BMI: Body mass index; IQR: interquartile range; ECMO: extracorporeal membrane oxygenation The difference in regional ICU capacity and trend in prefecture-level ICU congestion Supplemental Fig. 1 shows a scatterplot of the number of COVID-19 ICU beds per 100,000 individuals in each prefecture. We analyzed the availability of COVID-19 ICU beds per 100,000 people across population categories, finding median values as follows: 1.26 beds (IQR: 0.77–1.89) in low-population prefectures, 0.93 beds (IQR: 0.62–1.10) in mid-population prefectures, and 0.939 beds (IQR: 0.808–1.18) in high-population prefectures. In the Kruskal–Wallis rank-sum test, there were no statistical differences in the number of COVID-19 ICU beds among prefecture-population categories ( p = 0.43). Figure 2 shows the prefecture-level ICU congestion trends in each region and nationwide. Congestion level differed at each time and in each prefecture, from 0 to over 1.5. In Hokkaido, Tohoku, and Chugoku regions, the congestion levels did not exceed 1.0 throughout the study period. In other regions, several prefectures’ had congestion levels of > 1.0, meaning that certain regions had to deal with more severely ill patients with COVID-19 than their capacity. Supplementary Fig. 2 shows timeline changes in the total number of patients with severe COVID-19 and the number of deaths. April 2021 and August 2021 had higher numbers of patients with COVID-19 than other periods. Supplementary Fig. 3 shows a thematic map showing congestion levels in each prefecture in April 2021 and August 2021, reflecting the discrepancy in congestion levels. In April 2021, the Kinki region had higher congestion levels than other regions. In August 2021, both Kanto and Kinki regions had higher congestion levels than other regions. In April 2021, congestion level 1 or higher was only observed in the Kinki region, suggesting that the patients were concentrated in this region. Visualization of prefecture-level ICU congestion at admission and mortality and the results of mixed-effects model Figure 3 shows the cubic spline using the generalized additive model, depicting the relationship between the prefecture-level ICU congestion level at admission and the change in log odds of death among patients with severe COVID-19. The curve highlighted in blue indicates a significant increase in the risk of death as congestion levels increase, particularly beyond 1.0. This increase became progressively steep, suggesting a critical threshold beyond which patient outcomes worsened dramatically. Table 2 provides the results of the mixed-effects model, which includes random intercepts for each prefecture-level in the main analysis and for both prefecture and hospital levels in the sensitivity analysis. The mortality rate was 23.2% in this cohort. Scaling the variables was necessary because of convergence issues associated with continuous variable distributions in the mixed-effects model; specifically, the standard deviation of the congestion level in the original dataset was 0.373. Table 2 − 1. Main analysis: Mixed-effects model using each prefecture as random effects Primary outcome: In-hospital death (n = 2,306/10.046 (23.0%)) Variable Odds ratio (95% Confidence interval ) p value (Intercept) 0.18 (0.14–0.25) < 0.001 Congestion level 1.14 (1.06 – 1.22) < 0.001 Sex: male 1.04 (0.93–1.17) 0.48 Age 2.53 (2.35–2.73) < 0.001 BMI category < 18.5 1.11 (0.87–1.43) 0.39 18.5–22.4 0.97 (0.84–1.12) 0.69 25.0–29.9 0.93 (0.82–1.07) 0.34 ≥ 30.0 1.06 (0.90–1.25) 0.49 ECMO use 3.79 (3.28–4.39) < 0.001 Prone position 1.12 (1.01–1.24) 0.03 Pandemic wave 2nd wave 0.86 (0.64–1.16) 0.33 3rd wave 0.99 (0.78–1.25) 0.95 4th wave 0.95 (0.74–1.22) 0.71 5th wave 1.30 (1.01–1.67) 0.04 6th wave 1.08 (0.83–1.41) 0.57 7th wave 1.07 (0.78–1.48) 0.67 8th wave 1.12 (0.81–1.53) 0.50 BMI: body mass index; ECMO: extracorporeal membrane oxygenation BMI category of “22.5–24.9” and the Pandemic wave of “1st wave” were used as the reference. Table 2 2. Sensitivity analysis: Mixed-effects mode using both each prefecture and each hospital as random effects Primary outcome: In-hospital death (n = 2,306/10.046 (23.0%)) Variable Odds ratio (95% Confidence interval ) p value (Intercept) 0.17 (0.13–0.23) < 0.001 Congestion level 1.14 (1.06 – 1.22) < 0.001 Sex: male 1.05 (0.93–1.18) 0.41 Age 2.54 (2.35–2.75) < 0.001 BMI category < 18.5 1.17 (0.91–1.51) 0.21 18.5–22.4 0.98 (0.84–1.13) 0.75 25.0–29.9 0.93 (0.81–1.07) 0.67 ≥ 30.0 1.04 (0.88–1.23) 0.65 ECMO use 4.01 (3.44–4.68) < 0.001 Prone position 1.24 (1.10–1.40) < 0.001 Pandemic wave 2nd wave 0.86 (0.64–1.17) 0.35 3rd wave 0.99 (0.78–1.26) 0.93 4th wave 0.96 (0.75–1.24) 0.78 5th wave 1.27 (0.98–1.64) 0.07 6th wave 1.13 (0.86–1.50) 0.36 7th wave 1.14 (0.71–1.59) 0.45 8th wave 1.09 (0.79–1.52) 0.60 BMI: body mass index; ECMO: extracorporeal membrane oxygenation BMI category of “22.5–24.9” and the Pandemic wave of “1st wave” were used as the reference. Prefecture-level ICU congestion level at admission demonstrated a statistically significant association with mortality (OR: 1.14; 95% CI: 1.08–1.21). An increase in the congestion level from 0–100% was associated with an absolute increase in the mortality rate of 6.85 percentage points, from the original 23% to nearly 30%. Sensitivity analysis revealed similar results, suggesting that our analyses were robust. Discussion Our retrospective analysis of the CRISIS database revealed no statistically significant differences in ICU bed availability across the population categories in Japan's prefectures. However, considerable regional differences in ICU congestion were observed. The number of patients with COVID-19 and associated deaths peaked in April and August 2021. Spline curve analysis demonstrated a marked increase in mortality when prefecture-level ICU congestion surpassed a level of 1.0, and the mixed-effects model further confirmed that high prefecture-level ICU congestion was associated with increased mortality, which was robust even after adjusting for hospital-level factors. Both the spline regression and mixed-effects models revealed a consistent trend between ICU congestion and mortality despite differing assumptions about linearity. Spline regression allowed us to model the non-linear relationship, whereas the mixed-effects model provided a more detailed analysis accounting for prefectural differences. Together, these methods confirmed that elevated ICU congestion was a significant predictor of mortality in patients with severe COVID-19. Previous studies have explored the association between ICU capacity strain and mortality, showing that ICU capacity and strain affected COVID-19 mortality in the United States and the United Kingdom [ 1 , 7 ]. However, these studies had short observation periods, and the differences in viral strains were not considered. Compared with previous studies, our study period was extended by several years, allowing for adjustments in congestion levels during each pandemic phase. The trend of increased mortality associated with ICU congestion was consistent with the findings of previous studies. Multiple models, including mixed-effects models, have been used to analyze hospital and regional differences [ 1 , 3 ]. Gibbons et al. constructed separate models to compare mortality across US census divisions while adjusting for other covariates. Although the number of beds per 100,000 people did not differ among the prefectures, the congestion levels differed between the rural and urban areas of each prefecture. Previous analysis revealed the capacity of the Japanese healthcare system to accommodate critically ill patients with COVID-19 at conventional, contingency, and crisis surge levels, emphasizing that a high number of small ICUs with < 10 beds is one of the limiting factors in accepting a large number of critically ill patients. Additionally, these small medical institutions lack cooperation in functionally complementing one another to continue providing medical services in their local areas [ 20 ]. The discrepancies in regional ICU congestion may be attributable to circumstances in Japan and the small number of hospitals in rural areas. The increased mortality rates can be attributed to the additional burden on medical staff and facilities caused by high congestion [ 21 , 22 ]. A prospective observational study in France reported that transferring selected patients with COVID-19 from overwhelmed regions to areas with greater capacity may have improved patient access to ICU care without compromising the short-term mortality risk of the transferred patients [ 23 ]. Our findings underscore the importance of analyses that address the heterogeneity in health system responses over time and across different prefectures. Policies could focus on alleviating congestion, improving facility staffing levels a, and expanding ICU capacity. Additionally, developing systems to transport critically ill patients might help mitigate the impact of environmental factors contributing to increased mortality. Our study has several limitations that should be considered when interpreting the findings. Firstly, the retrospective nature of the data may introduce bias due to unobserved or unrecorded characteristics, which could influence the estimation of associations. Additionally, the mixed-effects model used in this study did not include interaction factors to account for potential combined effects of covariates such as age and comorbidities, leaving room for unaccounted confounding factors. Secondly, the COVID-19 pandemic progressed, subsequent surges with potentially new variants changed the characteristics of the patient population, including the level of vaccination, which may have influenced mortality [ 24 , 25 ]. Thirdly, real-time changes in the number of beds for COVID-19 as well as the number of physicians and nurses per bed were unmeasured factors affecting COVID-19 mortality. Additionally, the study did not account for whether hospitals operated as high-intensity or low-intensity ICUs, and this aspect could provide further insights into the association between limited resources and COVID-19 mortality. Finally, although all patients in the current study needed MV or ECMO, we did not consider the severity indicators of COVID-19-related respiratory failure to adjust the surge index, unlike a previous study [ 3 ]. Despite these limitations, our findings suggest that ICU congestion may be associated with an increased risk of mortality in Japan. As ICU congestion rises, there appears to be a trend toward worse outcomes in patients with severe COVID-19, underscoring the potential value of improving ICU resource management to mitigate adverse effects. Our findings can support healthcare decision-makers in refining contingency plans and improving hospital preparedness for future health emergencies. Abbreviations ICU, Intensive Care Unit CRISIS, CRoss Icu Searchable Information System BMI, body mass index ECMO, extracorporeal membrane oxygenation IQR, interquartile range OR, odds ratio CI, confidence interval Declarations Ethical approval statement This study was approved by the Hiroshima University Epidemiological Research Ethics Review Committee (approval numbers E2022-0118). In accordance with Japanese government guidelines, the Institutional Review Board exempted this study from the requirement for written informed consent to protect participant anonymity. Consent for publication Not applicable Availability of data and materials The data that support the findings of this study are available from the corresponding author, TO, upon reasonable request. Competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Funding source This research was supported by Japan Agency for Medical Research and Development(AMED) under Grant Number JP23fk0108654. Authors’ contributions Yudai Iwasaki: Conceptualization, Methodology, Investigation, Writing - original draft Takayuki Ogura: Conceptualization, Writing - review & editing Hiroyuki Ohbe: Methodology, Writing - review & editing Satoru Hashimoto: Conceptualization, Data curation, Writing - review & editing Shigeki Kushimoto: Conceptualization, Writing - review & editing Shinichiro Ohshimo: Conceptualization, Data curation, Writing - review & editing Nobuaki Shime: Conceptualization, Data curation, Writing - review & editing, Project administration Shinhiro Takeda: Writing - review & editing, Supervision All authors read and approved the final manuscript. Acknowledgments Editorial support in the form of medical writing, assembling tables, creating high-resolution images based on authors’ detailed directions, collating author comments, copyediting, fact-checking, and referencing was provided by Editage, Cactus Communications. References Wilcox ME, Rowan KM, Harrison DA. Does an unprecedented strain in ICU capacity, as experienced during the COVID-19 pandemic, affect patient outcomes? Crit Care Med. 2022;50:e548-56. https://doi.org/10.1097/CCM.0000000000005464. Haldane V, Jung AS, De Foo C, Bonk M, Jamieson M, Wu S, et al. Strengthening the basics: public health responses to prevent the next pandemic. BMJ 2021;375:e067510. https://doi.org/10.1136/bmj-2021-067510. Gibbons PW, Kim J, Cash RE, He S, Lai D, Christian Renne B, et al. Influence of ICU surge and capacity on COVID mortality across US states and regions during the COVID-19 pandemic. J Intensive Care Med. 2023;38:562-5. https://doi.org/10.1177/08850666231157338. Phua J, Kulkarni AP, Mizota T, Hashemian SMR, Lee WY, Permpikul C, et al. Critical care bed capacity in Asian countries and regions before and during the COVID-19 pandemic: an observational study. Lancet Reg Health West Pac. 2024;44:100982. https://doi.org/10.1016/j.lanwpc.2023.100982. Kadri SS, Sun J, Lawandi A, Strich JR, Busch LM, Keller M, et al. Association between caseload surges and COVID-19 survival in 558 US hospitals in the United States from March to August 2020. Ann Intern Med. 2021;174:1240-51. https://doi.org/10.7326/M21-1213. Kiekkas P, Tzenalis A, Gklava V, Stefanopoulos N, Voyagis G, Aretha D. Delayed admission to the intensive care unit and mortality of critically ill adults: Systematic review and meta-analysis. Biomed Res Int. 2022;2022:4083494. https://doi.org/10.1155/2022/4083494. Janke AT, Mei H, Rothenberg C, Becher RD, Lin Z, Venkatesh AK. Analysis of hospital resource availability and COVID-19 mortality in the United States. J Hosp Med. 2021;16:211-4. https://doi.org/10.12788/jhm.3539. Li Z, Ma X, Gao S, Li Q, Luo H, Sun J, et al. Association between hospital and ICU structural factors and patient outcomes in China: A secondary analysis of National Clinical Improvement System Data in 2019. Crit Care 2022;26:24. https://doi.org/10.1186/s13054-022-03892-7. Ohbe H, Sasabuchi Y, Matsui H, Yasunaga H. Impact of the COVID-19 pandemic on critical care utilization in Japan: A nationwide inpatient database study. J Intensive Care 2022;10:51. https://doi.org/10.1186/s40560-022-00645-0. Kokudo N, Sugiyama H. Hospital capacity during the COVID-19 pandemic. Glob Health Med. 2021;3(2):56-59. doi:10.35772/ghm.2021.01031 Ohshimo S, Liu K, Ogura T, Iwashita Y, Kushimoto S, Shime N, et al. Trends in survival during the pandemic in patients with critical COVID-19 receiving mechanical ventilation with or without ECMO: Analysis of Japanese national registry data. Crit Care 2022;26:354. https://doi.org/10.1186/s13054-022-04187-7. Kato F, Bunya N, Nakayama R, Narimatsu E, Ohshimo S, Shime N, et al. Multisystem inflammatory syndrome in adults with COVID-19 requiring mechanical ventilation: A retrospective cohort study. Acute Med Surg. 2023;10:e885. https://doi.org/10.1002/ams2.885. Benchimol EI, Smeeth L, Guttmann A, Harron K, Hemkens LG, Moher D, et al. Reporting of studies conducted using observational routinely collected health data (RECORD) statement. Z Evid Fortbild Qual Gesundhwes 2016;115-116:33-48. https://doi.org/10.1016/j.zefq.2016.07.010. Portal Site of Official Statistics of Japan. Year. Portal Site of Official Statistics of Japan. [cited 2024 Aug 2]. https://www.e-stat.go.jp/en. Accessed dd month yyyy. Bravata DM, Perkins AJ, Myers LJ, Arling G, Zhang Y, Zillich AJ, et al. Association of intensive care unit patient load and demand with mortality rates in the US Department of Veterans Affairs hospitals during the COVID-19 pandemic. JAMA Netw Open 2021;4:e2034266. https://doi.org/10.1001/jamanetworkopen.2020.34266. Ministry of Health, Labour and Welfare. 2023. New Coronavirus Infection Control Advisory Board (121st) Document 3-7-2. [cited 2024 Aug 2] https://www.mhlw.go.jp/content/10900000/001088930.pdf. Accessed dd month yyyy. Hastie T, Tibshirani R. Generalized additive models for medical research. Stat Methods Med Res. 1995 Sep;4:187-96. https://doi.org/10.1177/096228029500400302. Wood SN. Generalized additive models: An introduction to R, 2nd ed. Place of publication: Chapman & Hall/CRC; 2017. Gilbert JB. Modeling item-level heterogeneous treatment effects: A tutorial with the glmer function from the lme4 package in R. Behav Res Methods 2024;56(5):5055-67. https://doi.org/10.3758/s13428-023-02245-8. Yamamoto T, Ozaki M, Kasugai D, Burnham G. Assessment of critical care surge capacity during the COVID-19 pandemic in Japan. Health Secur. 2021;19:479-87. https://doi.org/10.1089/hs.2020.0227. Nakamura A, Kotani K, Hatakeyama S, Obayashi S, Nagai R. Regional variations in coronavirus disease 2019 mortality in Japan: An ecological study. JMA J. 2023;6(4):397-403. https://doi.org/10.31662/jmaj.2023-0052. Sen-Crowe B, Sutherland M, McKenney M, Elkbuli A. A closer look into global hospital beds capacity and resource shortages during the COVID-19 pandemic. J Surg Res. 2021;260:56-63. https://doi.org/10.1016/j.jss.2020.11.062. Sanchez M-A, Vuagnat A, Grimaud O, Leray E, Philippe J-M, Lescure F-X, et al. Impact of ICU transfers on the mortality rate of patients with COVID-19: Insights from a comprehensive national database in France. Ann Intensive Care 2021;11:151. https://doi.org/10.1186/s13613-021-00933-2. Demoule A, Fartoukh M, Louis G, Azoulay E, Nemlaghi S, Jullien E, et al. ICU strain and outcome in patients with COVID-19: A multicenter retrospective observational study. PLOS ONE 2022;17:e0271358. https://doi.org/10.1371/journal.pone.0271358. Taylor K, Rivere E, Jagneaux T, LeBoeuf G, Estela K, Pierce C, et al. Clinical characteristics and outcomes of SARS-Cov-2 B.1.1.529 infections in hospitalized patients and multi-surge comparison in Louisiana. PLOS ONE 2022;17:e0268853. https://doi.org/10.1371/journal.pone.0268853. Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterial.docx Cite Share Download PDF Status: Published Journal Publication published 28 Oct, 2025 Read the published version in BMC Infectious Diseases → Version 1 posted Editorial decision: Revision requested 24 Dec, 2024 Editor assigned by journal 22 Dec, 2024 Submission checks completed at journal 22 Dec, 2024 First submitted to journal 18 Dec, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5667123","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":394163030,"identity":"ebb8989b-3083-4f86-9ca7-07b8a330d747","order_by":0,"name":"Yudai Iwasaki","email":"","orcid":"","institution":"Imperial Foundation Saiseikai Utsunomiya Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yudai","middleName":"","lastName":"Iwasaki","suffix":""},{"id":394163031,"identity":"bb820fbe-a550-4574-a98c-d262e0202d9e","order_by":1,"name":"Takayuki Ogura","email":"data:image/png;base64,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","orcid":"","institution":"Imperial Foundation Saiseikai Utsunomiya Hospital","correspondingAuthor":true,"prefix":"","firstName":"Takayuki","middleName":"","lastName":"Ogura","suffix":""},{"id":394163032,"identity":"86e729f0-22b4-476d-89d0-09d382f05dba","order_by":2,"name":"Hiroyuki Ohbe","email":"","orcid":"","institution":"Tohoku University Graduate School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Hiroyuki","middleName":"","lastName":"Ohbe","suffix":""},{"id":394163033,"identity":"04d342ae-db3e-44f6-80da-59800594aabe","order_by":3,"name":"Satoru Hashimoto","email":"","orcid":"","institution":"Non-profit Organization Japan ECMO Network","correspondingAuthor":false,"prefix":"","firstName":"Satoru","middleName":"","lastName":"Hashimoto","suffix":""},{"id":394163035,"identity":"acbeb0fa-a24d-44d6-8304-201870349012","order_by":4,"name":"Shigeki Kushimoto","email":"","orcid":"","institution":"Tohoku University Graduate School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Shigeki","middleName":"","lastName":"Kushimoto","suffix":""},{"id":394163036,"identity":"dd887af7-0e1c-4281-bf28-f429c0b7bfb2","order_by":5,"name":"Shinichiro Ohshimo","email":"","orcid":"","institution":"Hiroshima University","correspondingAuthor":false,"prefix":"","firstName":"Shinichiro","middleName":"","lastName":"Ohshimo","suffix":""},{"id":394163037,"identity":"e6bcdf16-2fe1-41ba-b19a-29d136f6387e","order_by":6,"name":"Nobuaki Shime","email":"","orcid":"","institution":"Hiroshima University","correspondingAuthor":false,"prefix":"","firstName":"Nobuaki","middleName":"","lastName":"Shime","suffix":""},{"id":394163038,"identity":"578ddc7e-4405-40ea-8cfa-132b6df95eb0","order_by":7,"name":"Shinhiro Takeda","email":"","orcid":"","institution":"Non-profit Organization Japan ECMO Network","correspondingAuthor":false,"prefix":"","firstName":"Shinhiro","middleName":"","lastName":"Takeda","suffix":""}],"badges":[],"createdAt":"2024-12-18 07:38:43","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5667123/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5667123/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12879-025-11755-z","type":"published","date":"2025-10-28T15:57:15+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":73338990,"identity":"bbc0db5d-87a7-4fb4-9031-721943306a79","added_by":"auto","created_at":"2025-01-09 04:56:56","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":432226,"visible":true,"origin":"","legend":"\u003cp\u003ePatient flow char\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-5667123/v1/071af6639509d1432b34cba0.png"},{"id":73337677,"identity":"e07133b7-1b31-450c-bc41-ec6150014ee8","added_by":"auto","created_at":"2025-01-09 04:48:56","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":3571330,"visible":true,"origin":"","legend":"\u003cp\u003eCongestion trends in each region\u003c/p\u003e\n\u003cp\u003eThe bold black line shows the congestion level for Japan as a whole. This figurereveals the congestion trends of Hokkaido and Tohoku, Kanto, Chubu, Kinki, Chugoku, Shikoku, and Kyusyu regions in Japan.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-5667123/v1/214fc82565aeacd855797276.png"},{"id":73337673,"identity":"7f1c8aee-97a0-42a9-956a-c255a3f5ebf9","added_by":"auto","created_at":"2025-01-09 04:48:56","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":4493,"visible":true,"origin":"","legend":"\u003cp\u003eCubic spline using the generalized additive model\u003c/p\u003e\n\u003cp\u003eThe figure depicts the relationship between prefecture-level ICU congestion levels at admission and the change in log odds of death among severe COVID-19 patients. The shaded area represents the 95% confidence interval, emphasizing the variability and precision of the estimates.\u003c/p\u003e","description":"","filename":"fig.png","url":"https://assets-eu.researchsquare.com/files/rs-5667123/v1/ac258dd552be77111dcb075b.png"},{"id":95040536,"identity":"2f477286-6c68-4040-9e40-84e4d738d695","added_by":"auto","created_at":"2025-11-03 16:09:34","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4106290,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5667123/v1/1a24e49c-03a7-4fa1-a5ed-acc33047d0a1.pdf"},{"id":73338991,"identity":"5c70d38f-d58c-478b-99f5-c9ddcd4f4769","added_by":"auto","created_at":"2025-01-09 04:56:56","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":764173,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-5667123/v1/6e49ac19f5e92cb733405061.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Impact of Prefecture-level Intensive Care Unit Congestion on Mortality in Severe COVID- 19 Patients: A Retrospective Observational Study in Japan","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe COVID-19 pandemic has placed unprecedented pressure on global healthcare systems, exposing weaknesses in public health infrastructure and resource allocation [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Intensive care units (ICUs) have faced severe strain owing to surges in critically ill patients, pushing their capacity beyond limits in many regions [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Studies have demonstrated that a lack of ICU preparedness and surge capacity is closely associated with increased mortality rates, highlighting the critical importance of ICU resources during health crises [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eICU capacity is a vital component in managing critically ill patients, particularly during a pandemic. A global study reported that countries significantly increased ICU bed numbers to mitigate mortality from COVID-19; however, rapid scaling often fell short of demand, resulting in delays in critical care [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Hospital ICU congestion has been directly linked to poor patient outcomes, including increased mortality, primarily due to delayed admission and inadequate staffing ratios [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Variations in ICU infrastructure, such as bed number, intensivist availability, and nurse-to-patient ratio, significantly influence patient outcomes in various regions [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Japan experienced ICU strains during the pandemic, particularly in prefectures with fewer ICU beds and medical personnel. The distribution of ICU resources across prefectures in Japan has become a critical issue, as regions with limited capacity experienced COVID-19 mortality [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Although studies have addressed the ICU resource challenges in Japan [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], analyses of the specific impact of regional ICU congestion on COVID-19 mortality rates remain insufficient.\u003c/p\u003e \u003cp\u003eTherefore, we hypothesized that regional ICU congestion increased the patients' mortality. This study aimed to analyze ICU capacity strains at the prefecture level in Japan, with administrative divisions organized into 47 prefectures, and their effects on patient outcomes during the COVID-19 pandemic.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy design and data source\u003c/h2\u003e \u003cp\u003eThis study was a retrospective cohort analysis using data from CRoss Icu Searchable Information System (CRISIS), the most extensive national repository for patients with severe COVID-19 data in Japan (UMIN000041450, Registration date: 2020/08/18) [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. CRISIS was created by the Japan Extracorporeal Membrane Oxygenation Network specifically for the COVID-19 outbreak, facilitating nationwide tracking of real-time data from ICUs during the pandemic; it encompasses over 6,600 ICU beds, representing approximately 90% of all ICU beds across Japan [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eCRISIS collected data on every COVID-19 patient who required mechanical ventilation or extracorporeal membrane oxygenation (ECMO) from January 2020 to March 2023 across all participating hospitals. This study was approved by the Hiroshima University Epidemiological Research Ethics Review Committee (approval numbers E2022-0118) and registered with the University Hospital Medical Information Network Clinical Trials Registry, Japan (UMIN trial ID: UMIN000041450). In accordance with Japanese government guidelines, the Institutional Review Board exempted this study from the requirement for written informed consent to protect participant anonymity. Additionally, this study adheres to the standards set by the REporting of Studies Conducted using Observational Routinely Collected Health Data statement [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eStudy population\u003c/h3\u003e\n\u003cp\u003eFrom January 1, 2020, to March 31, 2023, we included all patients with COVID-19 registered in the CRISIS database who required mechanical ventilation. Patients who lacked complete records for age, sex, height, weight, start and end dates of mechanical ventilation, and clinical prognosis were excluded.\u003c/p\u003e\n\u003ch3\u003eGeographic information in Japan\u003c/h3\u003e\n\u003cp\u003eJapan is divided into 47 prefectures, which serve as the primary administrative units. These prefectures are grouped into eight larger geographical regions based on their location and historical ties: Hokkaido, Tohoku, Kanto, Chubu, Kinkii, Chugoku, Shikoku, and Kyushu. Each region comprises multiple prefectures; for example, the Tohoku region has six prefectures in northeastern Japan, while the Kanto region includes Tokyo and six surrounding prefectures (Supplemental Table\u0026nbsp;1). These regional groupings were used to represent broader geographical trends in this analysis.\u003c/p\u003e\n\u003ch3\u003eVisualization of each prefecture capacity\u003c/h3\u003e\n\u003cp\u003eThe number of COVID-19 beds per prefecture was determined by aggregating the number of beds registered at each facility in the CRISIS database. These counts were then normalized per 100,000 inhabitants based on population data derived from the results of the Japanese census [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe analysis involved visualizing the distribution of COVID-19 beds per 100,000 inhabitants across prefecture-population categories: \u0026ldquo;Low\u0026rdquo;, \u0026ldquo;Middle\u0026rdquo;, and \u0026ldquo;High\u0026rdquo;, representing the lower, middle, and upper thirds of the population distribution, respectively. Each prefecture's name was plotted on the x-axis, and the number of COVID-19 beds per 100,000 people was plotted on the y-axis. Points in the scatter plot were color-coded based on the population category, reflecting the population size of each prefecture. The scatter plot visually compares how bed numbers vary with population size across prefectures. The data were then segregated by population category to compute the number of COVID-19 beds per 100,000 individuals in each category.\u003c/p\u003e\n\u003ch3\u003eDefinition of the level of ICU congestion\u003c/h3\u003e\n\u003cp\u003eCalculations of prefecture-level ICU congestion were performed for the early and late stages of each month. The congestion metric was derived from the methodologies used in previous studies [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. The following formula was employed to calculate ICU congestion:\u003c/p\u003e \u003cp\u003ePrefecture-level ICU congestion\u0026thinsp;=\u0026thinsp;total number of ventilator days for severely ill patients with COVID-19 admitted during the 2 weeks in a prefecture/maximum possible ventilator days for patients with COVID-19 patients with mechanical ventilation in a prefecture (calculated as the number of beds multiplied by 14 days).\u003c/p\u003e \u003cp\u003eThe number of beds available for COVID-19 patients in a prefecture was calculated by adding the number of beds available for COVID-19 patients in each hospital belonging to a prefecture. The number of beds available for COVID-19 patients requiring ventilation in each hospital was sourced from the registered information initially entered when the CRISIS database was first created. The congestion level for each patient was determined by extracting prefecture-level ICU congestion during their corresponding hospitalization period and assigned as the \u0026ldquo;congestion level at admission.\u0026rdquo;\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eVisualization of discrepancy in prefecture-level ICU congestion\u003c/h2\u003e \u003cp\u003eWe visualized prefecture-level ICU congestion in each region by plotting time-series graphs capturing the temporal progression of congestion levels. Initially, we calculated the national average congestion by aggregating and averaging the \u0026ldquo;ICU congestion\u0026rdquo; variable across all regions for each date, providing a benchmark against which regional performances can be compared. Subsequently, we generated individual plots for each region using the \u0026lsquo;ggplot2\u0026rsquo; package in R. Each plot displayed the region-specific congestion trends along with the national average, depicted as a solid black line, to contextualize the regional data within the broader national scenario. Hokkaido was grouped with the Tohoku region (six northeastern prefectures) for visualization purposes, as Hokkaido alone constitutes a single prefecture, whereas other regions consist of multiple prefectures.\u003c/p\u003e \u003cp\u003eTo visualize the discrepancies in congestion levels among prefectures in Japan, we created a thematic map using geographic and congestion data. To create the map, we selected two months when the number of patients with severe COVID-19 was higher than that in other periods in this database. The map is plotted with shaded areas representing each prefecture and colored according to the average congestion levels for the month. A color gradient was applied to denote congestion, ranging from green, which indicates low congestion, to red, which indicates high congestion. This gradient effectively illustrates the range of congestion levels from 0 to 1.5.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eData collection and main outcome\u003c/h3\u003e\n\u003cp\u003eThis dataset encompasses various patient details, including age, sex, body mass index (BMI), ECMO usage, prone position, duration of mechanical ventilation, and mortality. BMI was categorized into five groups: underweight (BMI\u0026thinsp;\u0026lt;\u0026thinsp;18.5), low normal (BMI 18.5\u0026ndash;22.5), normal (BMI 22.5\u0026ndash;25), overweight (BMI 25\u0026ndash;30), and obese (BMI\u0026thinsp;\u0026ge;\u0026thinsp;30). Each patient listed in the CRISIS database was monitored until discharge, transfer to another facility, or death. Pandemic waves were determined according to the information given by Ministry of Health, Labour and Welfare, Japan [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. The primary outcome was the in-hospital mortality.\u003c/p\u003e\n\u003ch3\u003eStatistical analyses\u003c/h3\u003e\n\u003cp\u003eData are reported as either the mean and standard deviation (SD) or the median and interquartile range (IQR) for continuous variables and as count and percentage (n [%]) for categorical variables.\u003c/p\u003e \u003cp\u003eCOVID-19 beds per 100,000 inhabitants across prefecture\u0026ndash;population categories were calculated. The Kruskal\u0026ndash;Wallis test was used to assess whether there were differences in the number of beds across the three population categories. To address the non-linear relationship between prefecture-level ICU congestion levels at admission and patient outcomes, we used a generalized additive model with a logistic link function [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Our model incorporates a spline function for ICU congestion using a cubic spline basis to flexibly model its effects on the log odds of death [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. In addition to ICU congestion, the model was controlled for several covariates, including sex, age, body mass index category, ECMO use, prone positioning, and pandemic waves. After confirming the relationship between prefecture-level ICU congestion at admission and mortality, we used mixed-effects logistic regression models to evaluate the impact of ICU congestion and other factors on the prognosis of death among patients with COVID-19.\u003c/p\u003e \u003cp\u003eFor the main analysis, we constructed a mixed-effects model, including congestion level, sex, age, BMI category, ECMO use, prone position, and pandemic waves as fixed effects, with random intercepts for each prefecture-level to account for the individual tendencies in each prefecture. Continuous variables, such as age and congestion level, were standardized, mean-centered, and scaled by their standard deviations to facilitate model convergence and interpretability. To enhance the robustness of our findings, we conducted a sensitivity analysis using an extended mixed-effects model. This model included all the predictors of the main model, with the addition of random intercepts at both facility and prefecture levels. Both models were fitted using the \u0026lsquo;glmer\u0026rsquo; function from the lme4 package in R [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. All statistical tests were two-tailed. The primary outcome of mortality was analyzed using a mixed-effects logistic regression to estimate the odds ratios (ORs) for the likelihood of death, accompanied by 95% confidence intervals (CIs). The significance level for all tests was set at 0.05, with p-values\u0026thinsp;\u0026lt;\u0026thinsp;0.05 considered statistically significant. All statistical analyses were conducted using R (version 4.3.1) (2023-06-16).\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003ePatient flow chart and patients characteristics\u003c/h2\u003e \u003cp\u003eFrom January 1, 2020, to March 31, 2023, 12,279 patients were eligible for this study. We excluded 2,233 patients due to missing values for one or more confounding factors, resulting in 2915 missing points and a final analysis of 10,046 patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e shows the characteristics of the study cohort. The mean age was 62.9 years, and approximately 75% of the patients were male. Prone positioning was needed in 45% of the study cohort, and ECMO was introduced in 12.4% of the patients. The median congestion level at admission was 0.3 [IQR: 0.1\u0026ndash;0.6]. The distribution of patients across pandemic waves showed that the 3rd and 4th waves accounted for the largest proportions, at 24.4% (n\u0026thinsp;=\u0026thinsp;2,447 and n\u0026thinsp;=\u0026thinsp;2,453, respectively). The 5th wave also contributed significantly, accounting for 22.8% (n\u0026thinsp;=\u0026thinsp;2,291) of the total. In contrast, the 1st and 2nd waves had relatively smaller contributions, at 5.5% (n\u0026thinsp;=\u0026thinsp;548) and 5.2% (n\u0026thinsp;=\u0026thinsp;522), respectively.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePatient characteristics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge, yrs, mean (SD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e62.9 (14.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex, male, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e7494 (74.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHeight, cm, median [IQR]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e166.0 [159.0, 171.0]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWeight, kg, median [IQR]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e69.7 [60.0, 80.0]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI category, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt; 18.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e507 (5.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e18.5\u0026ndash;22.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1967 (19.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e22.5\u0026ndash;24.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2288 (22.8)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e25.0\u0026ndash;29.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1916 (19.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge; 30.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3368 (33.5)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProne position, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4518 (45.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eECMO, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1241 (12.4)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCongestion level at admission, median [IQR]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.3 [0.1, 0.6]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePandemic wave, n (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1st wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e5.5 (548)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2nd wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e5.2 (522)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3rd wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e24.4 (2447)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e24.4 (2453)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e22.8 (2291)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9.0 (906)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4.3 (432)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4.4 (447)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"2\"\u003eSD: standard deviation; BMI: Body mass index; IQR: interquartile range; ECMO: extracorporeal membrane oxygenation\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eThe difference in regional ICU capacity and trend in prefecture-level ICU congestion\u003c/b\u003e \u003cb\u003eSupplemental Fig.\u0026nbsp;1\u003c/b\u003e shows a scatterplot of the number of COVID-19 ICU beds per 100,000 individuals in each prefecture. We analyzed the availability of COVID-19 ICU beds per 100,000 people across population categories, finding median values as follows: 1.26 beds (IQR: 0.77\u0026ndash;1.89) in low-population prefectures, 0.93 beds (IQR: 0.62\u0026ndash;1.10) in mid-population prefectures, and 0.939 beds (IQR: 0.808\u0026ndash;1.18) in high-population prefectures. In the Kruskal\u0026ndash;Wallis rank-sum test, there were no statistical differences in the number of COVID-19 ICU beds among prefecture-population categories (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.43).\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e shows the prefecture-level ICU congestion trends in each region and nationwide. Congestion level differed at each time and in each prefecture, from 0 to over 1.5. In Hokkaido, Tohoku, and Chugoku regions, the congestion levels did not exceed 1.0 throughout the study period. In other regions, several prefectures\u0026rsquo; had congestion levels of \u0026gt;\u0026thinsp;1.0, meaning that certain regions had to deal with more severely ill patients with COVID-19 than their capacity. \u003cb\u003eSupplementary Fig.\u0026nbsp;2\u003c/b\u003e shows timeline changes in the total number of patients with severe COVID-19 and the number of deaths. April 2021 and August 2021 had higher numbers of patients with COVID-19 than other periods. \u003cb\u003eSupplementary Fig.\u0026nbsp;3\u003c/b\u003e shows a thematic map showing congestion levels in each prefecture in April 2021 and August 2021, reflecting the discrepancy in congestion levels. In April 2021, the Kinki region had higher congestion levels than other regions. In August 2021, both Kanto and Kinki regions had higher congestion levels than other regions. In April 2021, congestion level 1 or higher was only observed in the Kinki region, suggesting that the patients were concentrated in this region.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eVisualization of prefecture-level ICU congestion at admission and mortality and the results of mixed-effects model\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e shows the cubic spline using the generalized additive model, depicting the relationship between the prefecture-level ICU congestion level at admission and the change in log odds of death among patients with severe COVID-19. The curve highlighted in blue indicates a significant increase in the risk of death as congestion levels increase, particularly beyond 1.0. This increase became progressively steep, suggesting a critical threshold beyond which patient outcomes worsened dramatically.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e2\u003c/span\u003e provides the results of the mixed-effects model, which includes random intercepts for each prefecture-level in the main analysis and for both prefecture and hospital levels in the sensitivity analysis. The mortality rate was 23.2% in this cohort. Scaling the variables was necessary because of convergence issues associated with continuous variable distributions in the mixed-effects model; specifically, the standard deviation of the congestion level in the original dataset was 0.373.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u0026thinsp;\u003cb\u003e\u0026minus;\u0026thinsp;1.\u003c/b\u003e Main analysis: Mixed-effects model using each prefecture as random effects\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003ePrimary outcome: In-hospital death (n\u0026thinsp;=\u0026thinsp;2,306/10.046 (23.0%))\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOdds ratio\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(95% Confidence interval )\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e(Intercept)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.14\u0026ndash;0.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCongestion level\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e1.14\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e(1.06\u003c/b\u003e\u0026ndash;\u003cb\u003e1.22)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex: male\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.93\u0026ndash;1.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.48\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.53\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.35\u0026ndash;2.73)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI category\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt; 18.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.87\u0026ndash;1.43)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.39\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e18.5\u0026ndash;22.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.84\u0026ndash;1.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.69\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e25.0\u0026ndash;29.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.82\u0026ndash;1.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.34\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge; 30.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.90\u0026ndash;1.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.49\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eECMO use\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(3.28\u0026ndash;4.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProne position\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.01\u0026ndash;1.24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePandemic wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2nd wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.64\u0026ndash;1.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.33\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3rd wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.78\u0026ndash;1.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.95\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.74\u0026ndash;1.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.71\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.01\u0026ndash;1.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.04\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.83\u0026ndash;1.41)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.57\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.78\u0026ndash;1.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.81\u0026ndash;1.53)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.50\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eBMI: body mass index; ECMO: extracorporeal membrane oxygenation\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eBMI category of \u0026ldquo;22.5\u0026ndash;24.9\u0026rdquo; and the Pandemic wave of \u0026ldquo;1st wave\u0026rdquo; were used as the reference.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003e2.\u003c/b\u003e Sensitivity analysis: Mixed-effects mode using both each prefecture and each hospital as random effects\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003ePrimary outcome: In-hospital death (n\u0026thinsp;=\u0026thinsp;2,306/10.046 (23.0%))\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOdds ratio\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(95% Confidence interval )\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e(Intercept)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.13\u0026ndash;0.23)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCongestion level\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003e1.14\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e(1.06\u003c/b\u003e\u0026ndash;\u003cb\u003e1.22)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e\u0026lt;\u0026thinsp;0.001\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex: male\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.93\u0026ndash;1.18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.41\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(2.35\u0026ndash;2.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI category\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt; 18.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.91\u0026ndash;1.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.21\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e18.5\u0026ndash;22.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.84\u0026ndash;1.13)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.75\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e25.0\u0026ndash;29.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.81\u0026ndash;1.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.67\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge; 30.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.88\u0026ndash;1.23)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.65\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eECMO use\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(3.44\u0026ndash;4.68)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProne position\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(1.10\u0026ndash;1.40)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePandemic wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2nd wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.64\u0026ndash;1.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.35\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3rd wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.78\u0026ndash;1.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.93\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.75\u0026ndash;1.24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.78\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.98\u0026ndash;1.64)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.07\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.86\u0026ndash;1.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.36\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.71\u0026ndash;1.59)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.45\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8th wave\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(0.79\u0026ndash;1.52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.60\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eBMI: body mass index; ECMO: extracorporeal membrane oxygenation\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eBMI category of \u0026ldquo;22.5\u0026ndash;24.9\u0026rdquo; and the Pandemic wave of \u0026ldquo;1st wave\u0026rdquo; were used as the reference.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003ePrefecture-level ICU congestion level at admission demonstrated a statistically significant association with mortality (OR: 1.14; 95% CI: 1.08\u0026ndash;1.21). An increase in the congestion level from 0\u0026ndash;100% was associated with an absolute increase in the mortality rate of 6.85 percentage points, from the original 23% to nearly 30%. Sensitivity analysis revealed similar results, suggesting that our analyses were robust.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur retrospective analysis of the CRISIS database revealed no statistically significant differences in ICU bed availability across the population categories in Japan's prefectures. However, considerable regional differences in ICU congestion were observed. The number of patients with COVID-19 and associated deaths peaked in April and August 2021. Spline curve analysis demonstrated a marked increase in mortality when prefecture-level ICU congestion surpassed a level of 1.0, and the mixed-effects model further confirmed that high prefecture-level ICU congestion was associated with increased mortality, which was robust even after adjusting for hospital-level factors.\u003c/p\u003e \u003cp\u003eBoth the spline regression and mixed-effects models revealed a consistent trend between ICU congestion and mortality despite differing assumptions about linearity. Spline regression allowed us to model the non-linear relationship, whereas the mixed-effects model provided a more detailed analysis accounting for prefectural differences. Together, these methods confirmed that elevated ICU congestion was a significant predictor of mortality in patients with severe COVID-19.\u003c/p\u003e \u003cp\u003ePrevious studies have explored the association between ICU capacity strain and mortality, showing that ICU capacity and strain affected COVID-19 mortality in the United States and the United Kingdom [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. However, these studies had short observation periods, and the differences in viral strains were not considered. Compared with previous studies, our study period was extended by several years, allowing for adjustments in congestion levels during each pandemic phase.\u003c/p\u003e \u003cp\u003eThe trend of increased mortality associated with ICU congestion was consistent with the findings of previous studies. Multiple models, including mixed-effects models, have been used to analyze hospital and regional differences [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Gibbons et al. constructed separate models to compare mortality across US census divisions while adjusting for other covariates.\u003c/p\u003e \u003cp\u003eAlthough the number of beds per 100,000 people did not differ among the prefectures, the congestion levels differed between the rural and urban areas of each prefecture. Previous analysis revealed the capacity of the Japanese healthcare system to accommodate critically ill patients with COVID-19 at conventional, contingency, and crisis surge levels, emphasizing that a high number of small ICUs with \u0026lt;\u0026thinsp;10 beds is one of the limiting factors in accepting a large number of critically ill patients. Additionally, these small medical institutions lack cooperation in functionally complementing one another to continue providing medical services in their local areas [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. The discrepancies in regional ICU congestion may be attributable to circumstances in Japan and the small number of hospitals in rural areas.\u003c/p\u003e \u003cp\u003eThe increased mortality rates can be attributed to the additional burden on medical staff and facilities caused by high congestion [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. A prospective observational study in France reported that transferring selected patients with COVID-19 from overwhelmed regions to areas with greater capacity may have improved patient access to ICU care without compromising the short-term mortality risk of the transferred patients [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Our findings underscore the importance of analyses that address the heterogeneity in health system responses over time and across different prefectures. Policies could focus on alleviating congestion, improving facility staffing levels a, and expanding ICU capacity. Additionally, developing systems to transport critically ill patients might help mitigate the impact of environmental factors contributing to increased mortality.\u003c/p\u003e \u003cp\u003eOur study has several limitations that should be considered when interpreting the findings.\u003c/p\u003e \u003cp\u003eFirstly, the retrospective nature of the data may introduce bias due to unobserved or unrecorded characteristics, which could influence the estimation of associations. Additionally, the mixed-effects model used in this study did not include interaction factors to account for potential combined effects of covariates such as age and comorbidities, leaving room for unaccounted confounding factors. Secondly, the COVID-19 pandemic progressed, subsequent surges with potentially new variants changed the characteristics of the patient population, including the level of vaccination, which may have influenced mortality [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Thirdly, real-time changes in the number of beds for COVID-19 as well as the number of physicians and nurses per bed were unmeasured factors affecting COVID-19 mortality. Additionally, the study did not account for whether hospitals operated as high-intensity or low-intensity ICUs, and this aspect could provide further insights into the association between limited resources and COVID-19 mortality. Finally, although all patients in the current study needed MV or ECMO, we did not consider the severity indicators of COVID-19-related respiratory failure to adjust the surge index, unlike a previous study [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eDespite these limitations, our findings suggest that ICU congestion may be associated with an increased risk of mortality in Japan. As ICU congestion rises, there appears to be a trend toward worse outcomes in patients with severe COVID-19, underscoring the potential value of improving ICU resource management to mitigate adverse effects. Our findings can support healthcare decision-makers in refining contingency plans and improving hospital preparedness for future health emergencies.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eICU,\u0026nbsp;Intensive Care Unit\u003c/p\u003e\n\u003cp\u003eCRISIS,\u0026nbsp;CRoss Icu Searchable Information System\u003c/p\u003e\n\u003cp\u003eBMI,\u0026nbsp;body mass index\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eECMO,\u0026nbsp;extracorporeal membrane oxygenation\u003c/p\u003e\n\u003cp\u003eIQR,\u0026nbsp;interquartile range\u003c/p\u003e\n\u003cp\u003eOR,\u0026nbsp;odds ratio\u003c/p\u003e\n\u003cp\u003eCI, confidence interval\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthical approval statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Hiroshima University Epidemiological Research Ethics Review Committee (approval numbers E2022-0118). In accordance with Japanese government guidelines, the Institutional Review Board exempted this study from the requirement for written informed consent to protect participant anonymity.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data that support the findings of this study are available from the corresponding author, TO, upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eCompeting interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eFunding source\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research was supported by Japan Agency for Medical Research and Development(AMED) under Grant Number JP23fk0108654.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYudai Iwasaki: Conceptualization, Methodology, Investigation, Writing - original draft\u003c/p\u003e\n\u003cp\u003eTakayuki Ogura: Conceptualization, Writing - review \u0026amp; editing\u003c/p\u003e\n\u003cp\u003eHiroyuki Ohbe: Methodology, Writing - review \u0026amp; editing\u003c/p\u003e\n\u003cp\u003eSatoru Hashimoto: Conceptualization, Data curation, Writing - review \u0026amp; editing\u003c/p\u003e\n\u003cp\u003eShigeki Kushimoto: Conceptualization, Writing - review \u0026amp; editing\u003c/p\u003e\n\u003cp\u003eShinichiro Ohshimo: Conceptualization, Data curation, Writing - review \u0026amp; editing\u003c/p\u003e\n\u003cp\u003eNobuaki Shime: Conceptualization, Data curation, Writing - review \u0026amp; editing, Project administration\u003c/p\u003e\n\u003cp\u003eShinhiro Takeda: Writing - review \u0026amp; editing, Supervision\u003c/p\u003e\n\u003cp\u003eAll authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEditorial support in the form of medical writing, assembling tables, creating high-resolution images based on authors\u0026rsquo; detailed directions, collating author comments, copyediting, fact-checking, and referencing was provided by Editage, Cactus Communications.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eWilcox ME, Rowan KM, Harrison DA. Does an unprecedented strain in ICU capacity, as experienced during the COVID-19 pandemic, affect patient outcomes? Crit Care Med. 2022;50:e548-56. https://doi.org/10.1097/CCM.0000000000005464.\u003c/li\u003e\n\u003cli\u003eHaldane V, Jung AS, De Foo C, Bonk M, Jamieson M, Wu S, et al. Strengthening the basics: public health responses to prevent the next pandemic. BMJ 2021;375:e067510. https://doi.org/10.1136/bmj-2021-067510.\u003c/li\u003e\n\u003cli\u003eGibbons PW, Kim J, Cash RE, He S, Lai D, Christian Renne B, et al. Influence of ICU surge and capacity on COVID mortality across US states and regions during the COVID-19 pandemic. J Intensive Care Med. 2023;38:562-5. https://doi.org/10.1177/08850666231157338.\u003c/li\u003e\n\u003cli\u003ePhua J, Kulkarni AP, Mizota T, Hashemian SMR, Lee WY, Permpikul C, et al. Critical care bed capacity in Asian countries and regions before and during the COVID-19 pandemic: an observational study. Lancet Reg Health West Pac. 2024;44:100982. https://doi.org/10.1016/j.lanwpc.2023.100982.\u003c/li\u003e\n\u003cli\u003eKadri SS, Sun J, Lawandi A, Strich JR, Busch LM, Keller M, et al. Association between caseload surges and COVID-19 survival in 558 US hospitals in the United States from March to August 2020. Ann Intern Med. 2021;174:1240-51. https://doi.org/10.7326/M21-1213.\u003c/li\u003e\n\u003cli\u003eKiekkas P, Tzenalis A, Gklava V, Stefanopoulos N, Voyagis G, Aretha D. Delayed admission to the intensive care unit and mortality of critically ill adults: Systematic review and meta-analysis. Biomed Res Int. 2022;2022:4083494. https://doi.org/10.1155/2022/4083494.\u003c/li\u003e\n\u003cli\u003eJanke AT, Mei H, Rothenberg C, Becher RD, Lin Z, Venkatesh AK. Analysis of hospital resource availability and COVID-19 mortality in the United States. J Hosp Med. 2021;16:211-4. https://doi.org/10.12788/jhm.3539.\u003c/li\u003e\n\u003cli\u003eLi Z, Ma X, Gao S, Li Q, Luo H, Sun J, et al. Association between hospital and ICU structural factors and patient outcomes in China: A secondary analysis of National Clinical Improvement System Data in 2019. Crit Care 2022;26:24. https://doi.org/10.1186/s13054-022-03892-7.\u003c/li\u003e\n\u003cli\u003eOhbe H, Sasabuchi Y, Matsui H, Yasunaga H. Impact of the COVID-19 pandemic on critical care utilization in Japan: A nationwide inpatient database study. J Intensive Care 2022;10:51. https://doi.org/10.1186/s40560-022-00645-0. \u003c/li\u003e\n\u003cli\u003eKokudo N, Sugiyama H. Hospital capacity during the COVID-19 pandemic. Glob Health Med. 2021;3(2):56-59. doi:10.35772/ghm.2021.01031 \u003c/li\u003e\n\u003cli\u003eOhshimo S, Liu K, Ogura T, Iwashita Y, Kushimoto S, Shime N, et al. Trends in survival during the pandemic in patients with critical COVID-19 receiving mechanical ventilation with or without ECMO: Analysis of Japanese national registry data. Crit Care 2022;26:354. https://doi.org/10.1186/s13054-022-04187-7.\u003c/li\u003e\n\u003cli\u003eKato F, Bunya N, Nakayama R, Narimatsu E, Ohshimo S, Shime N, et al. Multisystem inflammatory syndrome in adults with COVID-19 requiring mechanical ventilation: A retrospective cohort study. Acute Med Surg. 2023;10:e885. https://doi.org/10.1002/ams2.885.\u003c/li\u003e\n\u003cli\u003eBenchimol EI, Smeeth L, Guttmann A, Harron K, Hemkens LG, Moher D, et al. Reporting of studies conducted using observational routinely collected health data (RECORD) statement. Z Evid Fortbild Qual Gesundhwes 2016;115-116:33-48. https://doi.org/10.1016/j.zefq.2016.07.010. \u003c/li\u003e\n\u003cli\u003ePortal Site of Official Statistics of Japan. Year. Portal Site of Official Statistics of Japan. [cited 2024 Aug 2]. https://www.e-stat.go.jp/en. Accessed dd month yyyy.\u003c/li\u003e\n\u003cli\u003eBravata DM, Perkins AJ, Myers LJ, Arling G, Zhang Y, Zillich AJ, et al. Association of intensive care unit patient load and demand with mortality rates in the US Department of Veterans Affairs hospitals during the COVID-19 pandemic. JAMA Netw Open 2021;4:e2034266. https://doi.org/10.1001/jamanetworkopen.2020.34266.\u003c/li\u003e\n\u003cli\u003eMinistry of Health, Labour and Welfare. 2023. New Coronavirus Infection Control Advisory Board (121st) Document 3-7-2. [cited 2024 Aug 2] https://www.mhlw.go.jp/content/10900000/001088930.pdf. Accessed dd month yyyy.\u003c/li\u003e\n\u003cli\u003eHastie T, Tibshirani R. Generalized additive models for medical research. Stat Methods Med Res. 1995 Sep;4:187-96. https://doi.org/10.1177/096228029500400302.\u003c/li\u003e\n\u003cli\u003eWood SN. Generalized additive models: An introduction to R, 2nd ed. Place of publication: Chapman \u0026amp; Hall/CRC; 2017.\u003c/li\u003e\n\u003cli\u003eGilbert JB. Modeling item-level heterogeneous treatment effects: A tutorial with the glmer function from the lme4 package in R. Behav Res Methods 2024;56(5):5055-67. https://doi.org/10.3758/s13428-023-02245-8.\u003c/li\u003e\n\u003cli\u003eYamamoto T, Ozaki M, Kasugai D, Burnham G. Assessment of critical care surge capacity during the COVID-19 pandemic in Japan. Health Secur. 2021;19:479-87. https://doi.org/10.1089/hs.2020.0227.\u003c/li\u003e\n\u003cli\u003eNakamura A, Kotani K, Hatakeyama S, Obayashi S, Nagai R. Regional variations in coronavirus disease 2019 mortality in Japan: An ecological study. JMA J. 2023;6(4):397-403. https://doi.org/10.31662/jmaj.2023-0052.\u003c/li\u003e\n\u003cli\u003eSen-Crowe B, Sutherland M, McKenney M, Elkbuli A. A closer look into global hospital beds capacity and resource shortages during the COVID-19 pandemic. J Surg Res. 2021;260:56-63. https://doi.org/10.1016/j.jss.2020.11.062.\u003c/li\u003e\n\u003cli\u003eSanchez M-A, Vuagnat A, Grimaud O, Leray E, Philippe J-M, Lescure F-X, et al. Impact of ICU transfers on the mortality rate of patients with COVID-19: Insights from a comprehensive national database in France. Ann Intensive Care 2021;11:151. https://doi.org/10.1186/s13613-021-00933-2.\u003c/li\u003e\n\u003cli\u003eDemoule A, Fartoukh M, Louis G, Azoulay E, Nemlaghi S, Jullien E, et al. ICU strain and outcome in patients with COVID-19: A multicenter retrospective observational study. PLOS ONE 2022;17:e0271358. https://doi.org/10.1371/journal.pone.0271358.\u003c/li\u003e\n\u003cli\u003eTaylor K, Rivere E, Jagneaux T, LeBoeuf G, Estela K, Pierce C, et al. Clinical characteristics and outcomes of SARS-Cov-2 B.1.1.529 infections in hospitalized patients and multi-surge comparison in Louisiana. PLOS ONE 2022;17:e0268853. https://doi.org/10.1371/journal.pone.0268853.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-infectious-diseases","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"infd","sideBox":"Learn more about [BMC Infectious Diseases](http://bmcinfectdis.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/infd","title":"BMC Infectious Diseases","twitterHandle":"#bmcinfectdis","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Japan prefectures, COVID-19 mortality, ICU congestion, ECMO, Mixed effect model","lastPublishedDoi":"10.21203/rs.3.rs-5667123/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5667123/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eThe COVID-19 pandemic has placed unprecedented pressure on global healthcare systems, severely affecting the intensive care unit (ICU) capacity. Therefore, this study explored the association between prefecture-level ICU congestion and COVID-19 mortality in each prefecture of Japan.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThis retrospective study analyzed data from the CRoss Icu Searchable Information System, covering all patients with COVID-19 who required mechanical ventilation or extracorporeal membrane oxygenation between January 1, 2020, and March 31, 2023. Prefecture-level ICU congestion was calculated as the total ventilator days over 2 weeks for severely ill patients with COVID-19, divided by the maximum potential ventilator days in the prefecture. Prefecture-level ICU congestion in each region was visualized by plotting time-series graphs capturing the temporal progression of congestion levels. A mixed-effects logistic regression model was fitted to evaluate the association between prefecture-level ICU congestion and mortality.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eA total number of 10,046 patients were included in this study, and the mortality rate was 23.2%. Congestion levels varied by time and prefecture, ranging from 0 to over 1.5, indicating a strain beyond capacity. Median congestion level (interquartile range [IQR]) at admission was 0.3 (0.1, 0.6), and increased congestion level was statistically associated with increased mortality (odds ratio: 1.14; 95% confidential interval: 1.08\u0026ndash;1.21).\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eIncreased prefecture-level ICU congestion may be associated with increased COVID-19 mortality, warranting further investigation.\u003c/p\u003e","manuscriptTitle":"Impact of Prefecture-level Intensive Care Unit Congestion on Mortality in Severe COVID- 19 Patients: A Retrospective Observational Study in Japan","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-09 04:48:51","doi":"10.21203/rs.3.rs-5667123/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-12-24T16:13:31+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-12-23T03:39:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-12-23T03:38:54+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Infectious Diseases","date":"2024-12-18T07:36:23+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-infectious-diseases","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"infd","sideBox":"Learn more about [BMC Infectious Diseases](http://bmcinfectdis.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/infd","title":"BMC Infectious Diseases","twitterHandle":"#bmcinfectdis","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"3457c358-6261-4305-ae35-dcd29e6444fd","owner":[],"postedDate":"January 9th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-11-03T16:05:38+00:00","versionOfRecord":{"articleIdentity":"rs-5667123","link":"https://doi.org/10.1186/s12879-025-11755-z","journal":{"identity":"bmc-infectious-diseases","isVorOnly":false,"title":"BMC Infectious Diseases"},"publishedOn":"2025-10-28 15:57:15","publishedOnDateReadable":"October 28th, 2025"},"versionCreatedAt":"2025-01-09 04:48:51","video":"","vorDoi":"10.1186/s12879-025-11755-z","vorDoiUrl":"https://doi.org/10.1186/s12879-025-11755-z","workflowStages":[]},"version":"v1","identity":"rs-5667123","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5667123","identity":"rs-5667123","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}