Article, Emergency Medicine

Predictive value of capnography for severity of acute gastroenteritis in the emergency department

a b s t r a c t

Objective: This study first aims to assess the utility of ETCO2 levels in evaluating the severity of dehydration in adult patients that present to the ED with Acute gastroenteritis. AGE. Second, it intends to evaluate the correlation between ETCO2 and several metabolic parameters: creatinine, pH, bicarbonate (HCO3), and bases excessive (BE). Method: This prospective study was conducted with AGE patients in the ED of a training and research hospital between June 2018 and April 2019 after approval of the local ethical-committee. The two groups were defined according to the severity of AGE: mild and non-mild groups. For both groups, ETCO2 levels were measured and recorded on admission of the patients.

Results: 87 patients were included in the analyses. The median of EtCO2 values was found as lower in non-mild group than mild group; 30 (25-35) & 39 (33-34), respectively (p b 0.001). In ROC analysis for distinguishing be- tween the both groups, the AUC value was found to be 0.988 and the best cut-off level was found as 33.5 with 95% sensitivity and 93% specificity. In addition, strong negative correlation between ETCO2 and creatinine (p b 0.001, r: -0.771) were found.

Conclusion: ETCO2 levels decreased in the non-mild group of AGE patients; it could be useful to distinguish the mild group from the non-mild group. ETCO2 could be a reliable marker in predicting AKI in the management of AGE patients.

(C) 2019

Introduction

Acute gastroenteritis (AGE) is an inflammation of a part of the gas- trointestinal system, such as the stomach, Small intestine, or colon; the AGE inflammation is b14 days and is most often caused by infectious agents, such as viruses, bacteria, and fungi. AGE patients usually pre- sented with abdominal pain, cramping, nausea, vomiting, and, espe- cially, diarrhea [1]. In adult patients, although AGE is generally self- limited with simple oral rehydration therapy, it could cause serious morbidity and mortality, especially in severe AGE cases [2,3]. The most important clinical predictor of mortality and morbidity is dehydration level with acute Renal damage and metabolic acidosis due to the dehy- dration in patients with AGE [4]. Therefore, determining the dehydra- tion levels, the renal damage, and the acidosis in the early period is

* Corresponding author.

E-mail address: [email protected] (H. Uzunosmanoglu).

crucial for AGE treatment. Although several classifications and scores are used to evaluate the dehydration level in AGE patients, there is still no reliable, simple, and applicable clinical decision score or marker to evaluate the dehydration level at admission time in the emergency department (ED) [5-7].

End-tidal carbon dioxide (ETCO2) is an important indicator of meta- bolic situation that can be measured non-invasively. In several diseases that cause metabolic acidosis, it has been shown that the ETCO2 levels decreased [8]. The reason for this decrease is that increasing metabolic acidosis stimulates respiration and increases minute ventilation. In sev- eral recent studies, it has been shown that low ETCO2 levels are related to mortality in various diseases causing metabolic acidosis, such as shock, sepsis, Diabetic ketoacidosis , and metabolic disturbances [9-13]. However, the number of studies that have evaluated the severity of dehydration or metabolic situations in AGE patients is limited, espe- cially in adult patients [14,15].

Therefore, this study first aims to assess the utility of ETCO2 levels in evaluating the severity of dehydration in adult patients that present to

https://doi.org/10.1016/j.ajem.2019.158404

0735-6757/(C) 2019

1160 H. Uzunosmanoglu et al. / American Journal of Emergency Medicine 38 (2020) 11591162

the ED with AGE. Second, it intends to evaluate the correlation between ETCO2 and several metabolic parameters: creatinine, pH, bicarbonate (HCO3), and bases excessive (BE).

Materials and methods

Study design

This prospective diagnostic/prognostic study was conducted with AGE patients in the ED of a training and research hospital between June 2018 and April 2019 after receiving approval from a local ethical committee. In addition, written informed consent was obtained from all of the patients or from legally authorized relatives.

Study population, groups, and definitions

Patients over 18 years of age with AGE were included in this study. Patients who presented to the ED with acute diarrhea as the primary complaint for going to the ED, with or without other symptoms sugges- tive of AGE (abdominal pain, cramping, nausea, and vomiting) in addi- tion to acute diarrhea, were accepted to eligible for participation. Acute diarrhea was defined as an increased frequency of stools, from three or more times daily to N20 bowel movements in a 24-h period and/or daily stool weight exceeding 200 g [16]. Patients who were preg- nant, who had known chronic kidney diseases, who had known chronic inflammatory bowel diseases, who had final diagnoses that were not AGE, such as Acute gastrointestinal bleeding or acute mesenteric ische- mia, were excluded from this study.

In this study, the two groups were defined according to the severity of AGE: mild and non-mild groups. Patients whose systolic blood pres- sure was b100 mmHg, whose heart rate was higher than 90 beat/min, whose capillary filling time was N2 s, whose respiratory rate was N20 breaths/min, and whose National Early Warning Score was N5 according to the NICE guidelines were considered in the non-mild group [17]. All of the remaining patients were considered in the mild group.

Study protocol

All of the patients who had the inclusion criteria were examined and treated by the emergency physician in charge as a necessity for the rou- tine management process. In addition to this routine management, the investigator physicians measured and recorded the ETCO2 levels via a Masimo Root(R) device, using a nasal cannula during the first evaluation of the patients. Similarly, the investigator physicians recorded all of the demographical and clinical characteristics.

Statistical analysis

Statistical analyses were performed using SPSS 15.0 (Chicago, IL, USA). The categorical values of the patients were expressed as a number and a percentage and were analyzed with a chi-square test. The Kolmogorov-Smirnov test was used to assess the normal distribution of the variables. Continued values were presented as a mean and stan- dard deviation (SD) or median values and an inter quartile range (IQR) of 25%-75%. Non-parametric values were analyzed using the Mann-Whitney U, and the parametric values were analyzed with a student-t-test. To assess the prognostic utility of ETCO2 levels at varying cut-off values for the distinction between the mild and non-mild groups, a receiver-operating characteristic (ROC) curve was generated, and the area under the curve (AUC) was calculated. The 95% of confi- dence intervals (95% CIs) were also calculated when appropriate, and a p-value b0.05 was considered statistically significant.

Sample size calculation

The sample size was estimated with G*Power for Mac OS X (version 3.1.9.2, Universitat Dusseldorf, Germany). The goal was to detect a 5 mmHg difference between the mild and non-mild groups. In addition, an SD value of6.1 ng/ml was considered, according to Yang et al.’s study [15]. Therefore, assuming a two-sided alpha 0.05, a sample size of at least 32 patients in each group to achieve 90% power was anticipated.

Results

In the end, 87 patients were included in the analyses. Of them, 55.2%

(48) of the participants were female. The median age of patients was 58 (IQR25%-75%; 40-74 years). The demographics and the laboratory re- sults of all of the patients are presented in Table 1.

When comparing both groups (mild and non-mild) for several de- mographical/clinical characteristics and laboratory findings, statistically significance differences were found in age, systolic and diastolic blood pressure, heart rate, white blood cell count, pH, HCO3, and creatinine. All of the other data are presented in Table 2.

The ETCO2 values measured on admission to the ED were 39 (33-34) in the mild group and 30 (25-35) in the non-mild group. This difference was statistically significant (p b 0.001) (Table 2).

A ROC analysis was performed, and the AUC was calculated to deter- mine the cut-off level for the ETCO2 level for distinguishing between the mild and non-mild groups. Accordingly, the AUC value was 0.988 (95% CI; 0.973-1.000) (Fig. 1). The best the cut-off value of the ETCO2 for distinguishing between the mild and non-mild groups was found as

33.5; for this level, the sensitivity and specificity values were calculated as 95% and 93%, respectively.

When the correlation between the ETCO2 and several laboratory findings was evaluated, a strong positive correlation between ETCO2 and pH (p b 0.001, r: 0.782) and a strong negative correlation between ETCO2 and creatinine (p b 0.001, r: -0.771) were found. All of the other correlation data were shown in Table 3.

Table 1

Demographic data of all patients.

Gender, n (%)

Female 48 (55.2)

Age, years median (IQR 2575) 58 (40-74) Co-morbidity, n (%)

Hypertension 49 (56.3)

Diabetes mellitus 24 (27.6)

Coronary heart disease 28 (32.2)

Chronic lung disease 17 (19.7)

Vital signs, median (IQR 2575)

Systolic blood pressure, mmHg 116 (105-130)

Diastolic blood pressure, mmHg 70 (64-78)

Pulse/minute 108 (92-115)

Temperature, C? 37.4 (36.9-38.4)

Oxygen saturation 94 (91-95)

NEWS score median (IQR 2575) 4 (2-6)

ETCO2 median (IQR 2575) mmHg 34 (30-39)

Laboratory results median (IQR 2575)

WBC (103/uL) 12.7 (9.3-15.4)

Hemoglobin g/dL 13.6 (12.4-14.7)

Platelet x103/uL 206 (145-275)

Urea mg/dL 44.9 (27.8-84.7)

Creatinine mg/dL 1.09 (0.86-1.58)

Potassium mmol/L 4.33 (3.94-4.86)

Sodium mmol/L 137 (134-139)

Blood gas analysis median (IQR 2575)

PH 7.33 (7.30-7.41)

HCO3 mEq/L 21.8 (19.2-24)

BE mmol/L -2.4 [(-4)-(-0.5)]

WBC: White Blood Cell, PH: blood acid level, HCO3: sodium bicarbonate, BE: Base excess NEWS: National Early Warning Score, ETCO2: End-tidal carbon dioxide.

H. Uzunosmanoglu et al. / American Journal of Emergency Medicine 38 (2020) 11591162

1161

Table 2

Patients characteristics and laboratory findings according to Severity of disease.

mmHg

Vital signs median (IQR 2575) Systolic blood pressure, mmHg

Diastolic blood pressure, mmHg

128 (118-138.5) 108 (100-115) b0.001

73 (66-80) 66 (55-74) 0.002

Fig. 1. ROC analysis for ETCO2 levels between mild and non-mild groups.

Yang et al. assessed the correlation between ETCO2 levels and severity of AGE [15]. The authors defined the severity of AGE by using a clinical dehydration scale (CDS). As a result, they concluded that, whereas a cor-

Mild groups (n =

Non mild groups (n

P

44)

=

value

43)

Age median (IQR 2575)

42.5 (32-60.5)

70 (51-82)

b0.001

Gender n (%)

Female

26 (59.1%)

22 (51.2%)

0.4

Co-morbidity n (%)

Hypertension

18 (36.7%)

31 (63.3%)

0.003

Diabetes mellitus

5 (11.4%)

19 (44.2%)

0.001

Coronary heart disease

7 (15.9%)

21 (48.8%)

0.001

Chronic lung disease

2 (4.5%)

15 (34.9%)

b0.001

NEWS score median (IQR

2 (0-3)

6 (5-7)

b0.001

2575)

Laboratory median (IQR 2575)

WBC (103/uL)

11.3 (8.1-13)

14.9 (12.6-16.6)

b0.001

Hemoglobin g/dL

13.65 (12.6-14.5)

13.4 (12.9-15.2)

0.8

Platelet x103/uL

220 (173.7-287.5)

203 (134-274)

0.2

Urea mg/dL

28.9 (24-36.4)

84.7 (64.2-111.8)

b0.001

Creatinine mg/dL

0.87 (0.7-1)

1.58 (1.34-1.88)

b0.001

Potassium mmol/L

4.03 (3.82-4.34)

4.84 (4.31-5)

b0.001

Sodium mmol/L

138 (137-140)

135 (131-137)

b0.001

Blood gas median (IQR 2575) PH

7.40 (7.37-7.42)

7.30 (7.29-7.32)

b0.001

HCO3 mEq/L

22.4 (20-24.1)

21 (17.8-23.8)

0.013

BE mmol/L

-0.75

-3.6 (-4.6/-2.8)

b0.001

(-2.35/+0.5)

ETCO2 median (IQR 2575)

39 (36-41)

30 (28-32)

b0.001

relation was found between ETCO2 and CDS, this correlation was weak (r:

Pulse/minute 93 (84.5-103) 115 (111-121) b0.001

Temperature, C? 36.9 (36.6-37.1) 38.2 (37.7-38.9) b0.001

Oxygen saturation 95 (94-96.7) 92 (90-94) b0.001

WBC: White Blood Cell, PH: blood acid level, HCO3: sodium bicarbonate, BE: Base excess, NEWS: National Early Warning Score, ETCO2: End-tidal carbon dioxide.

Discussion

The present study, which evaluated the prognostic value of ETCO2 on detecting the severity in patients presented to the ED with AGE, had two important findings. First, it was found that initial ETCO2 levels were significantly lower in the non-mild AGE patients than in the mild group. Considering that the cut-off value of ETCO2in distinguishing the non-mild groups was lower than the normal range of ETCO2 (35-45 mmHg) with high sensitivity and specificity, it is believed that ETCO2 values may be clinically useful in predicting non-mild patients during admission to the ED. Second, it was found that ETCO2 levels cor- related with pH, HCO3, BE, and Creatinine levels. For physicians, one of the most important clinical questions in the management of AGE pa- tients is development of pre-renal Acute kidney injury and/or met- abolic acidosis. Therefore, in considering the strong correlation between ETCO2 and creatinine (r: 0.771), it may be helpful in predicting AKI and it may decrease unnecessary laboratory examining. In contrast, when con- sidering the middle correlation between ETCO2 and HCO3 (r: 0.42) for predicting metabolic acidosis, it seems to be an unreliable marker.

In the medical literature, the majority of previously studies focused on

the pediatric population. To the best of our knowledge, there is no study conducted with adult AGE patients. In addition, the majority of these pe- diatric studies generally focused on assessing the correlation between ETCO2 and HCO3 levels rather than severity score based on clinical evalu- ation. In a prospective cohort study conducted with 169 children with vomiting and/or diarrhea, Freedman et al. evaluated whether ETCO2 is useful to predict severity of dehydration level by using ROC analysis. As a conclusion, the authors reported that the usefulness of ETCO2 was lim- ited (the AUC value was found as 0.34) [18]. In another pediatric study,

-0.20). Finally, in a study that aimed to develop a triage tool for ED to predict acidosis in children with AGE, Madati et al. evaluated ETCO2 as a potential parameter of their triage Decision tool. However, the useful- ness of ETCO2 as a parameter of that the triage decision tool was limited for predicting acidosis or the final disposition [19]. It seems that this study’s results were more optimistic than the results of both previous studies. This study found that the AUC value of ETCO2 values as 0.94 for distinguishing the non-mild group and mild group. There are two main reasons that can lead to these different results. First, use of a clinical deci- sion tool including more parameters was a more objective way to deter- mine the severity and so such a tool may have caused the groups in our study to be separated more clearly. Second, our study’s population consisted of adult patients, unlike the previous ones.

The most important reason of low ETCO2 levels in the non-mild group may be that the dehydration-induced metabolic acidosis and this increasing metabolic acidosis stimulated hyperventilation for the compensation of acidosis. Previous studies clearly revealed the relation- ship between low ETCO2 levels and metabolic acidosis. In Fearon et al.’s study conducted with 42 children with DKA, a strong correlation be- tween the HCO3 and ETCO2 levels (r2 = 0.80) was found [9]. The other studies that evaluated the relationship between ETCO2 and HCO3 in patients with AGE found significant correlations between these parameters; however, coefficients of these correlations were var- iable in these studies. For example, Nagler et al. and Freedman et al.

Table 3

Correlation coefficients and levels of significance between pH, HCO3, BE, creatinine and ETCO2.

Variables

Correlation coefficient

p-Value

pH

0.782

b0.001

HCO3 mEq/L

0.427

b0.001

BE mmol/L

0.692

b0.001

Creatinine mg/dL

-0.771

b0.001

PH: blood acid level, HCO3: sodium bicarbonate, BE: Base excess, ETCO2: End-tidal carbon dioxide.

1162 H. Uzunosmanoglu et al. / American Journal of Emergency Medicine 38 (2020) 11591162

reported strong correlation between ETCO2 and HCO3 levels (r values were 0.80 and 0.78, respectively) in their studies conducted in AGE chil- dren [14,18]. In contrast, Garcia et al. and Yang et al. reported low coef- ficients of correlation levels in their studies; r values were reported as

0.61 and 0.32, respectively [15,20]. In our study, it was found there

was a moderate correlation (r: 0.42) between ETCO2 and HCO3 levels. Finally, when considering the variable results of the studies, it could be stated that it is unclear whether ETCO2 levels predicted the metabolic acidosis in patients with AGE.

Previous studies that were conducted with pediatric population did not evaluate the relationship between ETCO2 and creatinine levels. In our study, different from the previous studies, this relationship was evaluated; a strong negative correlation (-0.772) was found between ETCO2 and the creatinine levels. This data could be helpful for physicians to detect low-risk patients for AKI, decreasing unnecessary laboratory examinations and saving time. In addition, it can be also used to detect high-risk patients that may require more urgent attention and labora- tory tests in ED.

Limitations

This study had some limitations. First, this study was conducted in a single center. Although the sample size was calculated, the study popu- lation was relatively small. Therefore, the main findings of this study cannot be generalized to the general population. In addition, though the study attempted to standardize the definition of AGE, the etiological search of AGE did not work.

Conclusion

In conclusion, this study showed that the ETCO2 levels decreased in the non-mild group of AGE patients; it could be useful to distinguish the mild group from the non-mild group. In addition, because of the strong negative correlation between the ETCO2 and creatinine levels, ETCO2 could be a reliable marker in predicting AKI in the management of AGE patients. However, the ETCO2 level is not a reliable marker in predicting the metabolic acidosis in these patient groups.

Acknowledgements

The authors declared that there is no acknowledgement.

Authorship

Conceived and designed the experiments; SKC and HU. Performed the experiments; HU, EE, SD. Analyzed and interpreted the data; SKC, YC, EE. Contributed reagents, materials, analysis tools or data; SKC, SD, EE, HU, YC. Wrote the paper; SKC, HU, SD.

Funding

There is no funding.

Declaration of competing interest

The authors declared that there is no conflict of interest.

Informed consent

The patient informed consent were signed by all participations.

Ethical approval

The local ethic committee was approved this study. (Kecioren Train- ing and Research Hospital, local ethical committee).

Human rights

During the study period, the research was conducted in accordance with the tenets of the Helsinki Declaration and Human Rights.

References

  1. Graves NS. Acute gastroenteritis. Prim Care 2013;40(3):727-41. https://doi.org/10. 1016/j.pop.2013.05.006 Sep. [Epub 2013 Jul 19].
  2. Liu L, Johnson HL, Cousens S, Perin J, Scott S, Lawn JE, et al. Global, regional, and na- tional causes of child mortality: an updated systematic analysis for 2010 with time trends since 2000. Lancet 2012;379:2151-61.
  3. Guerrant RL, Kosek M, Lima AAM, Lorntz B, Guyatt HL. Updating the DALYs for diarrhoeal disease. Trends Parasitol 2002;18:191-3.
  4. de Heese HV, Tonin C, Bowie MD, Evans A. Management of metabolic acidosis in acute gastroenteritis. Br Med J 1966;2(5506):144-6 Jul 16.
  5. Bailey B, Gravel J, Goldman RD, Friedman JN, Parkin PC. External validation of the clinical dehydration scale for children with acute gastroenteritis. Acad Emerg Med 2010;17:583-8.
  6. Gorelick MH, Shaw KN, Murphy KO. Validity and reliability of clinical signs in the di- agnosis of dehydration in children. Pediatrics 1997;99:E6.
  7. Yilmaz K, Karabocuoglu M, Citak A, Uzel N. Evaluation of laboratory tests in dehydrated children with acute gastroenteritis. J Paediatr Child Health 2002;38: 226-8.
  8. Hunter CL, Silvestri S, Ralls G, Bright S, Papa L. The sixth vital sign: prehospital end- tidal carbon dioxide predicts in-hospital mortality and metabolic disturbances. Am J Emerg Med 2014;32:160-5.
  9. Fearon DM, Steele DW. End-tidal carbon dioxide predicts the presence and severity of acidosis in children with diabetes. Acad Emerg Med 2002;9:1373-8.
  10. Hunter CL, Silvestri S, Dean M, Falk JL, Papa L. End-tidal carbon dioxide is associated with mortality and lactate in patients with suspected sepsis. Am J Emerg Med 2013; 31:64-71.
  11. McGillicuddy DC, Tang A, Cataldo L, Gusev J, Shapiro NI. Evaluation of end-tidal car- bon dioxide role in predicting elevated SOFA scores and lactic acidosis. Intern Emerg Med 2009;4:41-5.
  12. Kheng CP, Rahman NH. The use of end-tidal carbon dioxide monitoring in patients with hypotension in the emergency department. Int J Emerg Med 2012;5:31-8.
  13. Agus MS, Alexander JL, Mantell PA. Continuous non-invasive end-tidal CO2 monitor- ing in pediatric inpatients with diabetic ketoacidosis. Pediatr Diabetes 2006;7(4): 196-200 Aug.
  14. Nagler J, Wright RO, Krauss B. End-tidal carbon dioxide as a measure of acidosis among children with gastroenteritis. Pediatrics 2006;118:260-7.
  15. Yang HW, Jeon W, Min YG, Lee JS. Usefulness of end-tidal carbon dioxide as an indi- cator of dehydration in pediatric emergency departments: a retrospective observa- tional study. Medicine (Baltimore) 2017;96(35):e7881.
  16. Tintinalli Judith E. Tintinalli’s emergency medicine a comprehensive study guide. In: Kman NE, Werman HA, editors. Disorders presenting primarily with diarrhea. 8th ed. New York: McGraw-Hill; 2016. p. 492-6.
  17. Intravenous fluid therapy in adults in hospital clinical guideline https://www.nice. org.uk/guidance/cg174/resources/intravenous-fluid-therapy-in-adults-in-hospital- pdf-35109752233669 Published: 10 December 2013 [accessed 29 May 2019].
  18. Freedman SB, Johnson DW, Nettel-Aguirre A, Mikrogianakis A, Williamson-Urquhart S, Monfries N, et al. Assessing dehydration employing end-tidal carbon dioxide in children with vomiting and diarrhea. Pediatr Emerg Care 2018;34(8):564-9.
  19. Madati PJ, Bachur R. Development of an emergency department triage tool to predict acidosis among children with gastroenteritis. Pediatr Emerg Care 2008;24(12): 822-30.
  20. Solana Garcia MJ, Lopez Lopez R, Adrian Gutierrez J, Penalba Citores A, Guerrero Soler M, MaranonPardillo R. Use of capnography in acute gastroenteritis. An Pediatr (Barc) 2008;68(4):342-5.

Leave a Reply

Your email address will not be published. Required fields are marked *