Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Dose-response relationship between diarrhea quantity and mortality in critical care patients: A retrospective cohort study

  • Ryohei Yamamoto,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Department of Healthcare Epidemiology, School of Public Health in the Graduate School of Medicine, Kyoto University, Yoshida-honmachi, Sakyo-ku, Kyoto, Japan

  • Hajime Yamazaki,

    Roles Conceptualization, Formal analysis, Methodology, Project administration, Supervision, Validation, Writing – original draft, Writing – review & editing

    Affiliation Section of Clinical Epidemiology, Department of Community Medicine, Graduate School of Medicine, Kyoto University, Shogoin-kawaramachi, Sakyo-ku, Kyoto, Japan

  • Shungo Yamamoto,

    Roles Conceptualization, Methodology, Supervision, Validation, Writing – review & editing

    Affiliation Department of Healthcare Epidemiology, School of Public Health in the Graduate School of Medicine, Kyoto University, Yoshida-honmachi, Sakyo-ku, Kyoto, Japan

  • Yuna Ueta,

    Roles Conceptualization, Data curation, Investigation, Project administration, Validation, Writing – review & editing

    Affiliation Department of Nutrition Management, Kameda Medical Center, Kamogawa, Japan

  • Ryo Ueno,

    Roles Conceptualization, Formal analysis, Methodology, Validation, Visualization, Writing – review & editing

    Affiliation The Australian and New Zealand Intensive Care Research Centre, Melbourne, VIC, Australia

  • Yosuke Yamamoto

    Roles Conceptualization, Methodology, Project administration, Supervision, Validation, Writing – original draft, Writing – review & editing

    yamamoto.yosuke.5n@kyoto-u.ac.jp

    Affiliation Department of Healthcare Epidemiology, School of Public Health in the Graduate School of Medicine, Kyoto University, Yoshida-honmachi, Sakyo-ku, Kyoto, Japan

Abstract

Background

Previous studies have shown that diarrhea, defined as a dichotomized cutoff, is associated with increased mortality of patients in intensive care units (ICUs). This study aimed to examine the dose-response relationship between the quantity of diarrhea and mortality in ICU patients with newly developed diarrhea.

Methods

We conducted this single-center retrospective cohort study. We consecutively included all adult patients with newly developed diarrhea in the ICU between January 2017 and December 2018. Newly developed diarrhea was defined according to the World Health Organization definition. The consistency of diarrhea was evaluated by the Bristol stool chart scale, and the quantity of diarrhea was assessed on the day when patients newly developed diarrhea. The primary outcome was in-hospital mortality. The risk ratio (RR) and 95% confidence interval (CI) for the association between diarrhea quantity and mortality were estimated using multivariable modified Poisson regression models.

Results

Among the 231 participants, 68.4% were men; the median age was 72 years. The median diarrhea quantity was 401g (interquartile range [IQR] 230‒645g), and in-hospital mortality was 22.9%. More diarrhea at baseline was associated with higher in-hospital mortality; the adjusted RR (95% CI) per 200-g increase was 1.10 (1.01‒1.20), p = 0.029. In sensitivity analyses with near quartile categories of diarrhea quantity (<250g, 250–399g, 400–649g, ≥650g), the adjusted RRs for each respective category were 1.00 (reference), 1.02 (0.51–2.04), 1.29 (0.69–2.43), and 1.77 (0.99–3.21), p for trend = 0.033.

Conclusions

A greater quantity of diarrhea was an independent risk factor for in-hospital mortality. The diarrhea quantity may be an indicator of disease severity in ICU patients.

Introduction

Diarrhea is a common gastrointestinal symptom in the intensive care unit (ICU), with an incidence of 35%–70% [1]. In ICU patients, enteral nutrition (composition, osmolarity, speed, intermittent or continuous, and fiber), drugs (e.g., antibiotics, laxatives), infectious diseases (e.g., Clostridium difficile infection [CDI]), and comorbidity (e.g., anemia, cirrhosis) can cause diarrhea [2]. The effects of diarrhea include increased risk of contamination of devices and wounds, dehydration, electrolyte abnormalities, and malabsorption [35].

Several studies have shown an association between diarrhea and mortality [614], and this association remained even in ICU patients without CDI [8]. Taito et al. conducted a systematic review and demonstrated that diarrhea was associated with the length of hospital stay and ICU mortality [14]; however, all previous studies defined diarrhea based on dichotomized criteria with respect to consistency and frequency. The European Society of Intensive Care Medicine (ESICM) has adopted dichotomized criteria for the quantity of diarrhea as a component of the definition of diarrhea in the ICU [3]. However, healthcare providers need to make decisions based on continuous conditions rather than dichotomized conditions in practice [15, 16]. For example, more diarrhea may cause worse electrolyte imbalance, nutritional deficit, and hemodynamic instability owing to water loss [17, 18], which leads to changes in clinical management. Moreover, more diarrhea may result in more deaths. To clarify this, it is necessary to quantify the relationship between the quantity of diarrhea and death.

This retrospective cohort study aimed to examine the dose-response relationship between diarrhea in ICU and mortality. We investigate the association between the quantity of diarrhea and mortality in ICU patients with newly developed diarrhea.

Materials and methods

Study design and setting

We conducted this single-center retrospective cohort study at Kameda Medical Center ICU. This study was reviewed and approved by the institutional review board of Kyoto University (R2253) and Kameda medical center (19–145). These committees waived the requirement of informed consent from all participants enrolled in this study because of the retrospective study design. This study was conducted according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [19]. A preprint has previously been published [20].

Study population

From January 2017 to December 2018, we consecutively included all patients aged ≥18 years with newly developed diarrhea in the ICU. We defined newly developed diarrhea in the ICU as three or more loose or liquid stools per day according to the World Health Organization (WHO) definition [21]. To include patients with newly developed diarrhea in the ICU, the following patients were excluded on the day of ICU admission: patients with a stoma, chronic diarrhea (e.g., inflammatory bowel syndrome, short bowel syndrome), post-gastrointestinal surgery, gastrointestinal bleeding, or bacterial and viral enteritis (including Clostridium difficile enteritis and cytomegalovirus enteritis) or those who already had diarrhea on the day of ICU admission. In addition, patients readmitted to the ICU and those who died on the day of admission were excluded.

Data collection

We collected data such as age, sex, admission category (medical or surgery), sepsis defined by sepsis-3 [22], ICU readmission, Charlson Comorbidity Index (CCI) [23], and treatment limitation (limitations in providing ICU-specific life-sustaining therapies such as mechanical ventilation, cardiopulmonary resuscitation) from electronic health record reviews. Other data such as Acute Physiology and Chronic Health Evaluation (APACHE) II score [24], Simplified Acute Physiology Score (SAPS) II [25], Sequential Organ Failure Assessment (SOFA) score [26], potential causes of diarrhea (proton pump inhibitor, enteral nutrition, antibiotics, laxative drugs), testing for CDI (glutamate dehydrogenase test, CD toxin, or stool culture), and biopsy-diagnosed cytomegalovirus enteritis were also collected. We refer to CCI with age score as “CCI” and define CCI without age score as “CCI without age score.”

Measurement of diarrhea

We defined diarrhea by the WHO definition (three or more loose or liquid stools per day). Stool data of ICU patients were collected from electronic health records. Nurses routinely checked the presence or absence of stools every 2–4 h during ICU stay. In all ICU patients, the consistency and quantity of all stool samples were assessed by a nurse. The Bristol Stool Chart Scale (BSCS) was used to evaluate the consistency of each stool sample [27]. The BSCS is a 7-point scale in which stools are scored according to cohesion and surface cracking as follows: 1. separate hard lumps like nuts; 2. sausage shaped but lumpy; 3. like a sausage or snake but with cracks on its surface; 4. like a sausage or snake and smooth and soft; 5. soft blobs with a clear-cut edge; 6. fluffy pieces with ragged edges and mushy; and 7. watery with no solid pieces. This scale has been evaluated for its concordance and is a widely used scale [2834]. The quantity of stool was measured using a weight scale and recorded in the electronic medical records. Since most ICU patients had urinary catheters, contamination of the stool by urine was minimized. The main exposure was the quantity of diarrhea per day on the day of the diarrhea diagnosis. The daily quantity of diarrhea was calculated from calendar days (total quantity from 0:00 to 24:00). These data were collected from electronic health records.

Outcome measurement

The primary outcome was in-hospital mortality. Secondary outcomes included ICU, 28-day, and 90-day mortality; ICU-free days at the 28-day [35]; and hospital-free days at the 90-day [36]. In these free days, we used the event-free survival day (the number of event-free days was considered zero for patients who died in the time frame) to measure these outcomes.

Statistical analyses

Patient characteristics are described as median and interquartile range (IQR). Modified Poisson regression models were used to estimate risk ratios (RRs) and 95% confidence intervals (CIs) for the association between the quantity of diarrhea (per 200-g increase) and in-hospital mortality [37, 38]. The reason for using the unit of 200 g is that the ESICM uses a 200–250 g cutoff for diarrhea quantity [3]. The multivariable analysis was adjusted for CCI, SOFA score, and serum albumin levels. These covariates were selected a priori based on clinical plausibility and previous studies [2, 8, 13]. We performed multiple imputations for missing values using multiple imputations by chained equation (MICE) with 50 iterations that generated 100 datasets with imputed missing values [39, 40].

To perform sensitivity analyses, we tested several modified Poisson regression models to assess the robustness of the primary analysis. First, we adjusted for the following covariates: model 1 for age, sex, CCI without age score, SOFA score, and serum albumin; model 2 for CCI, APACHE II score, and serum albumin; model 3 for CCI, SAPS II score, and serum albumin; model 4 for CCI, SOFA score, serum albumin, and enteral nutrition; and model 5 for CCI, SOFA score, serum albumin, enteral nutrition, and laxatives. Second, we conducted a complete case analysis. Third, because CDI and cytomegalovirus enteritis affects mortality, we performed a further analysis excluding patients diagnosed with them after ICU admission. Fourth, to evaluate the influence of urine and stool contamination, we performed an additional analysis limited to only patients without any risk of urine contamination (anuric patients or patients with any forms of urinary catheter and/or nephrostomy). Finally, instead of using continuous data (the actual quantity of diarrhea), we used categorical data (near quartiles of the quantity of diarrhea) for the primary model. We calculated P for trends across the median values of each near quantile category [41].

We applied the same analyses as those for the primary outcome to assess the association between the quantity of diarrhea (per 200-g increase) and the following secondary outcomes: ICU mortality, 28-day mortality, and 90-day mortality.

We reported 95% CIs as an informal measure of uncertainty and avoided using terms such as statistical significance according to the recommendations of the American Statistical Association [42]. The analyses were performed using R software, version 4.0.3 (The R Foundation for Statistical Computing, Vienna, Austria; https://www.R-project.org/).

Results

Patient characteristics

During the study period, 1579 patients were admitted to the ICU, and 334 adult patients with newly developed diarrhea were included in this study (S1 Table). Among those patients, 103 were excluded. Finally, 231 patients were included in the analysis (shown in Fig 1).

thumbnail
Fig 1. Flow diagram of the sample selection.

In our ICU, nurses routinely assess the consistency and quantity of all stools; therefore, no patients were excluded due to missing stool information. GI: Gastrointestinal, CDI: Clostridioides difficile infection, CMV: cytomegalovirus infection.

https://doi.org/10.1371/journal.pone.0280806.g001

The median age of patients was 72 (IQR [64, 80]) years, 158/231 (68.4%) patients were men, median CCI was two (IQR [1, 3]), median APACHE II score was 21 (IQR [14, 28]), and median SOFA score was 9 (IQR [6, 12]). Patients admitted for nonoperative reasons were the most prevalent (162/231, 70.1%). Sepsis was diagnosed in 121 patients (52.4%). Antimicrobials and laxative drugs as possible causes of diarrhea were administered to 214/231 (92.6%) and 119/231 (51.5%) patients, respectively. Overall, 2/231 (0.9%) patients were diagnosed with CDI in the ICU, and two (0.9%) patients were diagnosed with CMV by colonoscopic biopsy in the ICU. None of the patients used probiotics or synbiotics. Some forms of urinary catheter and/or nephrostomy catheters were inserted in 218/231 (94.4%) patients (216 urinary catheters and 2 nephrostomy catheters), and 5/231 (2.2%) patients were anuric. The median number of days from ICU admission to newly developed diarrhea was 3 (IQR [2, 6]), and the median quantity of diarrhea was 401 (IQR [230.5, 645]) g. The median consecutive days of diarrhea was 1 day (IQR [1, 2]), and the median total number of days of diarrhea in the ICU was 2 days (IQR [1, 4]).

Other patient characteristics on ICU admission are summarized in Table 1. Three patients had missing values for the severity score because arterial blood gas was not measured. There were no missing measurements for other variables, including the quantity of diarrhea.

Association between the quantity of diarrhea and outcomes

Table 2 presents mortality, length of stay, and free day survival. Two and 16 patients were lost to 28-day and 90-day follow-ups, respectively. In the unadjusted analysis, the quantity of diarrhea was associated with increased in-hospital mortality (unadjusted RR per 200 g increased: 1.10 [95% CI 1.01–1.19], p = 0.03). After adjusting for CCI, SOFA score, and serum albumin level, this association remained (adjusted RR per 200-g increase: 1.10 [95% CI 1.01–1.20], p = 0.03) (Table 3).

thumbnail
Table 2. Mortality, length of stay, and free day survival of the study patients.

https://doi.org/10.1371/journal.pone.0280806.t002

thumbnail
Table 3. Association between the quantity of diarrhea and in-hospital mortality.

https://doi.org/10.1371/journal.pone.0280806.t003

Sensitivity analyses for the primary analysis

The association between the quantity of diarrhea and in-hospital mortality remained similar in various multivariable analysis models and other sensitivity analyses (Table 3). We also performed a sensitivity analysis using the categories of the quantity of diarrhea. With no established criteria to distinguish the quantity of diarrhea, we used near-quantile-defined categories of the quantity of diarrhea. The quartiles of diarrhea were 230 g in the 25th percentile, 401 g in the 50th percentile, and 645 g in the 75th percentile. Therefore, the patients were divided into the following categories: mild (<250 g), moderate (250–399 g), severe (400–649 g), and very severe (≥650 g). In-hospital mortality was 19.7% (12/61) for mild, 19.2% (10/52) for moderate, 21.3% (13/61) for severe, and 31.6% (18/57) for very severe. Multivariable-modified Poisson regression analysis using these categories, with the same adjustments as in the primary model, showed a trend toward increased in-hospital mortality as the quantity of diarrhea increased (P for trend = 0.033) (Table 3).

Secondary analyses

For secondary analyses, a similar association was observed between the quantity of diarrhea and ICU 28-day and 90-day mortalities (Table 4). Multivariable analysis showed a similar trend of higher mortality with higher quantities of diarrhea.

thumbnail
Table 4. Association between the 200-g increase in the quantity of diarrhea and secondary outcomes.

https://doi.org/10.1371/journal.pone.0280806.t004

Discussion

In this retrospective study, we investigated the association between the quantity of diarrhea and in-hospital mortality in 231 patients with newly developed diarrhea in the ICU. Multivariable analysis revealed that diarrhea quantity was an independent predictor of in-hospital mortality. This association was consistent across several sensitivity analyses. Similarly, the greater the quantity of diarrhea, the higher the ICU 28-day and 90-day mortalities. To the best of our knowledge, this is the first study to show an association between the quantity of diarrhea and mortality.

In diagnosing diarrhea, the quantity, as well as the frequency and consistency, can help us predict outcomes. In this study, we showed that mortality increased with increasing diarrhea quantity, according to the adjusted RR in patients with newly developed diarrhea in the ICU. Previous studies have reported an association between the presence of diarrhea and mortality; however, no studies have examined whether mortality increases with a greater quantity of diarrhea [14, 44]. A systematic review of 12 studies, most of which used the definition of diarrhea as three or more loose or liquid stools, showed an association between diarrhea and mortality, but with high heterogeneity (RR: 1.43; 95% CI: 1.03–1.98; I2 = 86.7%; n = 11,866) [14]. A recent prospective study evaluating the association between the presence or absence of diarrhea, as defined by the WHO definition of at least 3 liquid bowel movements per day, found that mortality was not associated with the presence of diarrhea [1]. One possible reason for the inconsistent results of these studies may be that diarrhea is defined by consistency and frequency, without taking quantity into account, which might lead to classifying small quantities of clinically unimportant bowel movements as diarrhea. More diarrhea leads to worse electrolyte imbalance, nutritional deficit, and hemodynamic instability owing to water loss [17, 18], so the quantity of diarrhea should be evaluated for ICU patients.

The reason for the higher mortality rate among patients with a greater quantity of diarrhea remains unclear. Patients with CDI or cytomegalovirus enteritis, which are known to cause diarrhea, have been reported to have higher mortality, but they were excluded from our study. Indeed, diarrhea can cause dehydration, electrolyte abnormalities, metabolic acidosis, malnutrition, device contamination, and wound contamination [45]. However, since dehydration and electrolyte abnormalities are carefully corrected in the ICU, it is questionable to assume that diarrhea directly contributes to mortality.

Possible explanations for the relationship between diarrhea and mortality are as follows. First, diarrhea can be a sign of gastrointestinal organ failure that is associated with a high risk of mortality [4, 14, 46]. Patients with diarrhea have higher severity scores than those without diarrhea [2, 6, 810, 14]. In our study, most patients received treatments that could cause diarrhea, such as enteral nutrition and antimicrobials. These interventions are part of the treatment regimen for critically ill patients. In addition, approximately 60% of patients were on ventilation and used vasopressors, which means that patients with diarrhea have a higher severity of illness. In our analysis, we adjusted for the SOFA score, an organ disorder score that does not include gastrointestinal function, and showed that diarrhea is a risk factor for mortality independent of other organ disorders. The quantity of diarrhea may be a candidate when adjusting for organ dysfunction. Second, diarrhea can be a sign of a disorder of the gut microbiota, which is called dysbiosis. This dysbiosis is believed to increase vulnerability to nosocomial infections, sepsis, organ failure, and mortality [47, 48]. The development of diarrhea might be associated with dysbiosis in the gut microbiota of ICU patients [49]. However, our data and analyses are not sufficiently conclusive to prove them. Further research is needed to test these hypotheses.

This study had several limitations. First, the measurement of diarrhea was not completely accurate. If diarrhea spills out of the diaper, it may not be measured. In this case, this may have led to an underestimation of the quantity of diarrhea. However, we believe that this measurement of the quantity of diarrhea reflects real clinical practice. Second, the inter-rater reliability of BSCS was not confirmed in our study. The reliability of BSCS has been studied and widely used [2732, 50], and our nurses were trained to measure BSCS in clinical practice, which should have minimized the inter-rater variability. Third, eight patients were voiding freely without the use of urine catheters. Measuring the weight of the diaper could be the sum of stool and urine in those patients; however, our sensitivity analysis showed that the increase in the risk of mortality remained after limiting the analysis to only patients without risk of urine contamination. Fourth, mechanistic data on developing diarrhea in ICUs were lacking. To elucidate mechanisms, studies are needed to assess the relationship between changes in the microbiome over time during ICU stays with newly developed diarrhea, and to evaluate whether interventions to correct the microbiome improve diarrhea. Finally, this was a retrospective single-center study. Prospective multicenter studies that consider diarrhea quantity are needed to improve the measurement validity of diarrhea and the generalizability of these findings.

Conclusion

The greater quantity of diarrhea was associated with higher mortality in ICU patients with newly developed diarrhea. The quantity of diarrhea may be considered an indicator of disease severity in ICU patients. Further research is needed to determine if there is a causal relationship between the quantity of diarrhea and death.

Supporting information

S1 Table. The mortality and length of stay of the entire cohort.

https://doi.org/10.1371/journal.pone.0280806.s001

(DOCX)

Acknowledgments

The authors thank the staff of the ICU of the Kameda Medical Center. We thank the Japanese Society of Education for Physicians and Trainees in the Intensive Care Clinical Trial Group (JSEPTIC-CTG) for their suggestions and comments on the earlier concept of this study.

References

  1. 1. Dionne JC, Mbuagbaw L, Devlin JW, Duprey MS, Cartin-Ceba R, Tsang J, et al. Diarrhea during critical illness: a multicenter cohort study. Intensive Care Medicine. 2022;48(5):570–9. pmid:35411491
  2. 2. Thibault R, Graf S, Clerc A, Delieuvin N, Heidegger C, Pichard C. Diarrhoea in the ICU: respective contribution of feeding and antibiotics. Critical Care. 2013;17(4):R153. pmid:23883438
  3. 3. Reintam Blaser A, Malbrain MLNG, Starkopf J, Fruhwald S, Jakob SM, De Waele J, et al. Gastrointestinal function in intensive care patients: terminology, definitions and management. Recommendations of the ESICM Working Group on Abdominal Problems. Intensive Care Medicine. 2012;38(3):384–94. pmid:22310869
  4. 4. Reintam A, Parm P, Kitus R, Kern H, Starkopf J. Gastrointestinal symptoms in intensive care patients. Acta Anaesthesiol Scand. 2009;53(3):318–24. pmid:19243317
  5. 5. Hiesmayr M, Schindler K, Pernicka E, Schuh C, Schoeniger-Hekele A, Bauer P, et al. Decreased food intake is a risk factor for mortality in hospitalised patients: the NutritionDay survey 2006. Clin Nutr. 2009;28(5):484–91. pmid:19573957
  6. 6. Borges SL, Pinheiro BDV, Pace FHDL, Chebli JMF. Diarréia nosocomial em unidade de terapia intensiva: incidência e fatores de risco. Arquivos de Gastroenterologia. 2008;45(2):117–23.
  7. 7. Ozgur M, Girgin NK, Akalin H, Iscimen R, Sinirtas M, Kahveci F. A retrospective evaluation of the incidence and risk factors of nosocomial diarrhea in critically ill adult patients. Acta Medica. 2016;32:741–6.
  8. 8. Tirlapur N, Puthucheary ZA, Cooper JA, Sanders J, Coen PG, Moonesinghe SR, et al. Diarrhoea in the critically ill is common, associated with poor outcome and rarely due to Clostridium difficile. Scientific Reports. 2016;6(1):24691. pmid:27094447
  9. 9. Wu TJ, Liu ZJ, Zhao YM, Yang CL, Zhang CY. [Clinical analysis of the factors related to diarrhea in intensive care unit]. Zhongguo Wei Zhong Bing Ji Jiu Yi Xue. 2004;16(12):747–9. pmid:15585152
  10. 10. Jakob SM, Bütikofer L, Berger D, Coslovsky M, Takala J. A randomized controlled pilot study to evaluate the effect of an enteral formulation designed to improve gastrointestinal tolerance in the critically ill patient—the SPIRIT trial. Critical Care. 2017;21(1). pmid:28599662
  11. 11. Vieira LV, Pedrosa LAC, Souza VS, Paula CA, Rocha R. Incidence of diarrhea and associated risk factors in patients with traumatic brain injury and enteral nutrition. Metab Brain Dis. 2018;33(5):1755–60. pmid:30014176
  12. 12. Rodrigues de Barros J, Fraga Lobo IM, Melo Soares F, Ferreira de Almeida DSS. Factors associated with diarrhea in a unit of intensive therapy: cohort study. NUTRICION CLINICA Y DIETETICA HOSPITALARIA. 2018;38(2):122–7.
  13. 13. Atasever AG, Ozcan PE, Kasali K, Abdullah T, Orhun G, Senturk E. The frequency, risk factors, and complications of gastrointestinal dysfunction during enteral nutrition in critically ill patients. Therapeutics and Clinical Risk Management. 2018;Volume 14:385–91. pmid:29503558
  14. 14. Taito S, Kawai Y, Liu K, Ariie T, Tsujimoto Y, Banno M, et al. Diarrhea and patient outcomes in the intensive care unit: Systematic review and meta-analysis. J Crit Care. 2019;53:142–8. pmid:31238230
  15. 15. Altman DG, Royston P. The cost of dichotomising continuous variables. BMJ. 2006;332(7549):1080.1. pmid:16675816
  16. 16. Dawson NV, Weiss R. Dichotomizing Continuous Variables in Statistical Analysis. Medical Decision Making. 2012;32(2):225–6.
  17. 17. Wierdsma NJ, Peters JH, Weijs PJ, Keur MB, Girbes AR, Van Bodegraven AA, et al. Malabsorption and nutritional balance in the ICU: fecal weight as a biomarker: a prospective observational pilot study. Critical Care. 2011;15(6):R264. pmid:22071233
  18. 18. Anigilaje EA. Management of Diarrhoeal Dehydration in Childhood: A Review for Clinicians in Developing Countries. Front Pediatr. 2018;6:28. pmid:29527518
  19. 19. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: Guidelines for Reporting Observational Studies. Annals of Internal Medicine. 2007;147(8):573. pmid:17938396
  20. 20. Yamamoto R, Yamazaki H, Yamamoto S, Ueta Y, Ueno R, Yamamoto Y. Association between diarrhea quantity and in-hospital mortality in intensive care unit patients: A retrospective cohort study. 02 September 2021, PREPRINT (Version 1) available at Research Square [https://doi.org/10.21203/rs.3.rs-859799/v1].
  21. 21. Diarrhoeal disease 2017 [updated 2 May 2017. Available from: https://www.who.int/en/news-room/fact-sheets/detail/diarrhoeal-disease.
  22. 22. Seymour CW, Liu VX, Iwashyna TJ, Brunkhorst FM, Rea TD, Scherag A, et al. Assessment of Clinical Criteria for Sepsis. JAMA. 2016;315(8):762.
  23. 23. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–83. pmid:3558716
  24. 24. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med. 1985;13(10):818–29. pmid:3928249
  25. 25. Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. Jama. 1993;270(24):2957–63. pmid:8254858
  26. 26. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonça A, Bruining H, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22(7):707–10. pmid:8844239
  27. 27. O’Donnell LJ, Virjee J, Heaton KW. Detection of pseudodiarrhoea by simple clinical assessment of intestinal transit rate. BMJ. 1990;300(6722):439–40. pmid:2107897
  28. 28. Cook DJ, Johnstone J, Marshall JC, Lauzier F, Thabane L, Mehta S, et al. Probiotics: Prevention of Severe Pneumonia and Endotracheal Colonization Trial—PROSPECT: a pilot trial. Trials. 2016;17(1). pmid:27480757
  29. 29. Koh H, Lee MJ, Kim MJ, Shin JI, Chung KS. Simple diagnostic approach to childhood fecal retention using the Leech score and Bristol stool form scale in medical practice. J Gastroenterol Hepatol. 2010;25(2):334–8. pmid:19817956
  30. 30. Dionne JC, Sullivan K, Mbuagbaw L, Takaoka A, Duan EH, Alhazzani W, et al. Diarrhoea: interventions, consequences and epidemiology in the intensive care unit (DICE-ICU): a protocol for a prospective multicentre cohort study. BMJ Open. 2019;9(6):e028237. pmid:31248929
  31. 31. Caroff DA, Edelstein PH, Hamilton K, Pegues DA. The Bristol Stool Scale and Its Relationship to Clostridium difficile Infection. Journal of Clinical Microbiology. 2014;52(9):3437–9. pmid:25031446
  32. 32. Lewis SJ, Heaton KW. Stool Form Scale as a Useful Guide to Intestinal Transit Time. Scandinavian Journal of Gastroenterology. 1997;32(9):920–4. pmid:9299672
  33. 33. Bishop S, Young H, Goldsmith D, Buldock D, Chin M, Bellomo R. Bowel motions in critically ill patients: a pilot observational study. Crit Care Resusc. 2010;12(3):182–5. pmid:21261576
  34. 34. Schmidt SB, Kulig W, Winter R, Vasold AS, Knoll AE, Rollnik JD. The effect of a natural food based tube feeding in minimizing diarrhea in critically ill neurological patients. Clinical Nutrition. 2019;38(1):332–40. pmid:29358002
  35. 35. Agus MSD, Wypij D, Hirshberg EL, Srinivasan V, Faustino EV, Luckett PM, et al. Tight Glycemic Control in Critically Ill Children. New England Journal of Medicine. 2017;376(8):729–41. pmid:28118549
  36. 36. Self WH, Semler MW, Wanderer JP, Wang L, Byrne DW, Collins SP, et al. Balanced Crystalloids versus Saline in Noncritically Ill Adults. New England Journal of Medicine. 2018;378(9):819–28.
  37. 37. Zou G. A Modified Poisson Regression Approach to Prospective Studies with Binary Data. American Journal of Epidemiology. 2004;159(7):702–6. pmid:15033648
  38. 38. Holmberg MJ, Andersen LW. Estimating Risk Ratios and Risk Differences. JAMA. 2020;324(11):1098.
  39. 39. Rubin DB. Multiple imputation for nonresponse in surveys: John Wiley & Sons; 2004.
  40. 40. Groothuis-Oudshoorn K, Van Buuren S. Mice: multivariate imputation by chained equations in R. J Stat Softw. 2011;45(3):1–67.
  41. 41. Bansal S, Buring JE, Rifai N, Mora S, Sacks FM, Ridker PM. Fasting Compared With Nonfasting Triglycerides and Risk of Cardiovascular Events in Women. JAMA. 2007;298(3):309. pmid:17635891
  42. 42. Wasserstein RL, Lazar NA. The ASA Statement on p-Values: Context, Process, and Purpose. The American Statistician. 2016;70(2):129–33.
  43. 43. Knaus WA, Wagner DP, Draper EA, Zimmerman JE, Bergner M, Bastos PG, et al. The APACHE III prognostic system. Risk prediction of hospital mortality for critically ill hospitalized adults. Chest. 1991;100(6):1619–36. pmid:1959406
  44. 44. Lebak KJ, Bliss DZ, Savik K, Patten-Marsh KM. What’s New on Defining Diarrhea in Tube-Feeding Studies? Clinical Nursing Research. 2003;12(2):174–204. pmid:12741669
  45. 45. Reintam Blaser A, Deane AM, Fruhwald S. Diarrhoea in the critically ill. Curr Opin Crit Care. 2015;21(2):142–53. pmid:25692805
  46. 46. Reintam Blaser A, Poeze M, Malbrain MLNG, Björck M, Oudemans-Van Straaten HM, Starkopf J. Gastrointestinal symptoms during the first week of intensive care are associated with poor outcome: a prospective multicentre study. Intensive Care Medicine. 2013;39(5):899–909. pmid:23370829
  47. 47. Kitsios GD, Morowitz MJ, Dickson RP, Huffnagle GB, McVerry BJ, Morris A. Dysbiosis in the intensive care unit: Microbiome science coming to the bedside. J Crit Care. 2017;38:84–91. pmid:27866110
  48. 48. Mcdonald D, Ackermann G, Khailova L, Baird C, Heyland D, Kozar R, et al. Extreme Dysbiosis of the Microbiome in Critical Illness. mSphere. 2016;1(4):e00199–16.
  49. 49. Duan J, Meng X, Liu S, Zhou P, Zeng C, Fu C, et al. Gut Microbiota Composition Associated With Clostridium difficile-Positive Diarrhea and C. difficile Type in ICU Patients. Frontiers in Cellular and Infection Microbiology. 2020;10.
  50. 50. Lacy BE, Mearin F, Chang L, Chey WD, Lembo AJ, Simren M, et al. Bowel Disorders. Gastroenterology. 2016;150(6):1393–407.e5. pmid:27144627