High total carbon dioxide predicts 1-year readmission and death in patients with acute dyspnea

a b s t r a c t

Rationale: Patients with acute dyspnea are a large heterogeneous patient group where initial management is im- portant for outcome.

Objectives: The objective of the study is to investigate if venous blood gas parameters predict 1-year risk of read- mission or death in patients admitted to the emergency department due to acute dyspnea.

Methods: We studied 283 patients with acute dyspnea and followed them up for 1 year regarding incidence of readmission or death.

Measurements and main results: In venous blood obtained immediately upon admission levels of total carbon di- oxide (TCO2), base excess , potential hydrogen (pH), and partial pressure of carbon dioxide (pCO2) were measured. In Cox proportional hazards models, patients belonging to top and bottom quartiles of TCO2, BE, pH, and pCO2 were compared to patients belonging to the 2 central quartiles and assessed for end point. After adjust- ment, top (hazard ratio [HR], 1.48; 95% confidence interval [CI], 1.08-2.04; P = .016) and bottom (HR, 1.54; 95% CI, 1.08-2.18; P = .017) quartiles of BE were associated with increased risk of readmission or death. The strongest predictor was top quartile of TCO2 (HR, 1.68; 95% CI, 1.21-2.35; P = .002). In the combined analysis, top quartile of TCO2 remained significantly related to the end point (HR, 1.59; 95% CI, 1.03-2.45; P = .035), whereas BE be- came nonsignificant. Comorbidities, for example, prevalent chronic obstructive pulmonary disease, did not ex- plain the association. Neither pCO2 nor pH predicted the end point.

Conclusions: A high value of TCO2 appears to be an easily accessible marker for 1-year readmission or death in patients with acute dyspnea and may thus add clinically important information for risk stratification and follow-up strategies.

(C) 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license


  1. Introduction

Patients presenting with acute dyspnea at the emergency depart- ment (ED) is a large and heterogeneous patient group with high mortal- ity and readmission rates [1-8]. Initial ED management, level of care, and follow-up strategies are important factors for the outcome of acute dyspnea patients [9-12]. However, the individual prognosis is dif- ficult to accurately assess. The use of plasma biomarkers for improved determination of prognosis in acute dyspnea has largely focused on pa- tients with congestive heart failure and, to a lesser degree, on patients with chronic obstructive pulmonary disease (COPD). Many biomarker studies include markers of inflammation or cardiac stress [11,13-17], whereas the value of blood gas parameters [18,19] and the role of

? Sources of support: Funding was obtained from the European Research Council (StG-282255), the Swedish Heart and Lung Foundation; Swedish Research Council; the Novo Nordisk Foundation; the Skane University Hospital donation funds; the Medical Fac- ulty, Lund University; the governmental funding of clinical research within the national health services; and the Albert Pahlsson Research Foundation, Region Skane.

* Corresponding author. Tel.: +46 40 33 35 95, +46 704 89 08 98.

E-mail addresses: [email protected], [email protected] (N. Lund).

biomarkers in unselected patients with acute dyspnea on clinical out- come have been poorly studied.

The underlying causes of dyspnea can be difficult to assess in an early setting. Risk stratification of prognosis is central for clinical deci- sions on the level of care, treatment intensity, and urgency of reaching a definitive underlying diagnosis. Most plasma biomarkers in acute dys- pnea used at the ED have diagnostic purposes (eg, troponin T and C- reactive protein), whereas medical history and scores of vital parame- ters are used to assess prognosis, level of care, and treatment intensity. In Sweden, the “Medical emergency triage and Treatment System Adult” (METTS-A) is a standard tool for risk assessment and triage of ED patients and was used during the time of study enrollment [20].

In patients with COPD and acute dyspnea, a high pressure of carbon dioxide in arterial blood is a well-established predictor of poor progno- sis and motivates a high level of care and treatment intensity [8,21]. However, in the general setting of patients with acute dyspnea at the ED, arterial blood gas analysis is usually not performed. Therefore, it cannot be evaluated or used as a routine biomarker for risk stratifica- tion. Increasing evidence points toward that arterial and venous blood gas results can be used interchangeably [18,19,22-24]. In the ED, venous


0735-6757/(C) 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

blood can routinely be drawn and analyzed at a low cost in the same point-of-care equipment as arterial blood with immediate generation of test results [25]. To our knowledge, venous carbon dioxide and acid-base balance have never been evaluated as Prognostic biomarkers in unselected ED patients admitted because of acute dyspnea.

The aim of the study was to investigate if easily accessible venous blood gas analysis of total carbon dioxide (TCO2), base excess (BE), po- tential hydrogen (pH), and partial pressure of carbon dioxide (pCO2) predicts 1-year risk of readmission or death in patients admitted to the ED due to acute dyspnea.

  1. Methods

We studied patients presenting with dyspnea during 2011 at the ED of the University Hospital of Skane in Malmo, Sweden. In 2011, the de- partment had approximately 83000 visits. The hospital is the only emergency hospital in the municipality of approximately 300000

inhabitants. The study was approved by the regional board of ethics in Lund. All patients of 18 years of age or older presenting with acute dys- pnea as the major complaint were eligible for the study. This yielded 5057 visits of patients with acute dyspnea of which 500 were randomly selected for review of patient records. In total, 283 fulfilled study criteria for inclusion in our analyses (Fig. 1). Data collection was performed in 2013 from the medical records of the University Hospital of Skane. Var- iables recorded included sex (male/female), age (years), vital signs and symptoms according to METTS-A [20], pulse oximetry (percentages), respiratory rate (RR, rate/min), pH (pH scale), BE (millimoles per liter) [26,27], pCO2 (kilopascal), TCO2 (millimoles per liter), medical history, and readmission or death within 1 year. In the METTS-A ED triage sys- tem, patients with respiratory complaints are categorized into dyspnea, chest pain, or hyperventilation and ranked into 4 priority levels accord- ing to vital signs. Priority 4 comprises patients with normal vital signs; and priority 1, patients with pathologic vital signs needing immediate medical attention [20]. Additional detail on the methods for making

500 visits of patients presenting with acute dyspnea at the ED during 1st of January to 31st December 2011were randomly selected for review of patient records.

31 visits were excluded due to and no occurrence of dyspnea.

469 visits were further assessed.

34 visits were excluded as the patients left the hospital before physical examination.

435 visits were further assessed.

8 visits were excluded due to incomplete personal identity number and lacking medical records.

427 visits were further assessed.

106 visits were excluded due to missing blood gas parameters.

321 visits were further assessed

19 visits were excluded due to missing respiratory rate data and 5 visits were excluded due to missing pulsoximetry data.

297 visits were further assessed.

14 patients were recurrent visitors and excluded from statistical analysis after first readmission.

This yielded 283 patients with acute dyspnea included in the study.

Fig. 1. Selection of patients for statistical analysis.

the measurements is provided in an online data supplement. The com- bined end point was either a first Hospital readmission regardless of cause or death during the 1-year follow-up period. Date of death was registered from a regional or national population register. Planned re- visits were not included.

Venous blood gas parameters were obtained upon arrival to the ED. The blood gas samples were immediately analyzed on a Radiometer ABL800 Flex (Copenhagen, Denmark) [25]. Among the parameters ana- lyzed, TCO2 represents the total dissolved carbon dioxide in the blood and is constituted to approximately 95% of bicarbonate (HCO3-) and to 5% of carbon dioxide (CO2), and carbonic acid (H2CO3) [28]. We relat- ed the venous blood gas parameters with the risk of readmission or death using Cox proportional hazard models. The following 3 adjust- ments were used: In model 1, we adjusted for age and sex. In model 2, we included additional adjustment for METTS-A. In model 3, we added saturation, RR, and if statistically significant acid base parameters. Finally, we also adjusted model 3 for a history of COPD. Venous blood gas parameters were divided into quartiles. The 2 central quartiles were merged and defined as the reference group and compared to the top and bottom quartiles, respectively. Statistical analysis was per- formed in IBM SPSS (Malmo, Sweden) Statistics version 21. P b .05 was considered significant. Schoenfeld’s test was performed to assure validity of the proportional hazards. In the subset of the population where both venous and arterial blood gas analyses were performed, we correlated blood gas parameters using spearman correlation analysis.

  1. Results

A total of 283 patients with acute dyspnea at the ED were evaluated (Fig. 1). The mean age was 66.1 years (66.1 +- 18.5 years), and females (55.8%, 158/283) were more common than males. Most patients admit- ted for inpatient care was admitted to a general ward (39.9%, 113/283). A large proportion of the patients were directly discharged from the ED (36.7%, 104/283). The remaining were either admitted to an emergency ward (21.2%, 60/283) or an intensive care unit (2.1%, 6/283). During the 1-year follow-up, 74.2% (210/283) of the patients were readmitted or died, of which 67.1% (190/283) had a first readmission and 7.1% (20/ 283) died with no prior readmission. Detailed patient characteristics are shown in Table 1.

The bottom quartile of BE ranged from -26 to -1.0 mmol/L; and the top quartile, from 3.0 to 15 mmol/L. The bottom quartile of TCO2 ranged

Table 1

Patient base line characteristics

Age (y), mean (+-SD) 66.1 (+-18.5)

Sex (male), n (%) 125 (44.2%)

Medical history, n (%)

COPD 70 (24.7%)

Congestive heart failure 58 (20.5%)

Pneumonia 27 (9.5%)

Myocardial infarction 14 (4.9%)

Pulmonary embolism 3 (1.1%)

Vital parameters, mean (+-SD)

Oxygen saturation (%) 93.6 (+-5.15)

RR (min-1) 23.5 (+-6.57)

Heart rate (min-1) 93.5 (+-20.9)

Systolic blood pressure (mm Hg) 146 (+-26.2)

Diastolic blood pressure (mm Hg) 80.9 (+-14.6)

Body temperature (?C) 37.2 (+-0.78)

METTS-A category, n (%)

Priority 1 34 (12.0%)

Priority 2 72 (25.4%)

Priority 3 68 (24.0%)

Priority 4 109 (38.5%)

Venous blood gas parameters, mean (+-SD)

BE (mmol/L) 1.31 (+-3.23)

pH (pH scale) 7.40 (+-0.06)

TCO2 (mmol/L) 27.3 (+-3.73)

pCO2 (kPa) 5.80 (+-1.23)

from 4.0 to 24 mmol/L; and the top quartile, from 30 to 40 mmol/L. The bottom quartile of pH ranged from a value of 7.07 to 7.36; and the top quartile, from 7.44 to 7.59. The bottom quartile of pCO2 ranged from 2.9 to 5.0 kPa, and the top quartile from 6.5 to 13.5 kPa.

In Cox proportional hazard models adjusted for age and sex, top quar- tile of BE (hazard ratio [HR], 1.56 [confidence interval {CI}, 1.14-2.15]; P =

.006) and bottom quartile of BE (HR, 1.65 [CI, 1.17-2.32]; P = .004) were

strongly associated with the end point. In the multivariate models adjust- ed for age, sex, METTS-A, saturation, and RR, both top quartile (HR, 1.48 [CI, 1.08-2.04]; P = .016) and bottom quartile (HR, 1.54 [CI 1.08-2.18];

P = .017) of BE were associated with increased risk of readmission or death (Table 2). However, when adjusting for TCO2, there was no associ- ation of top quartile BE with the end point (HR, 1.13 [CI, 0.75-1.71]; P =

.570), and the association of bottom quartile BE was no longer significant (HR, 1.63 [CI, 0.99-2.68]; P = .056).

When adjusting for age and sex, patients in the top quartile of TCO2 had an increased risk of readmission or death (HR, 1.93 [CI, 1.40-2.67]; P = .000). This association was also seen for bottom quartile of TCO2 (HR, 1.46 [CI, 1.02-2.07]; P = .037). After additional adjustment for age, sex, METTS-A, saturation, and RR, top quartile but not bottom quartile of TCO2 remained significant (Table 2). Top quartile of TCO2 was significant in the multivariate model including BE (HR, 1.59 [CI, 1.03-2.45]; P =

.035). Top quartile of TCO2 also remained significant after adjustment for a history of COPD (HR, 1.61 [CI, 1.15-2.26]; P = .005) with a relative risk increase greater than for prevalent COPD (HR, 1.50 [CI, 1.09-2.06]; P = .012). A Kaplan-Meier plot of TCO2 in relation to the end point is shown in Fig. 2. In the patient group with top quartile levels of TCO2, 33.8% (23/68) of the patients were discharged from the ED, 39.7% (27/68) were admitted to a general ward, 22.1% (15/68) were admitted to an emergency ward, and 4.4% (3/68) were admitted to an intensive care unit. Older age (HR, 1. 01 [CI, 1.01-1.02]; P = .002) and higher priority ac- cording to METTS-A (HR, 1.31 [CI, 1.07-1.61]; P = .008) were other factors independently associated with the end point adjusted for BE and TCO2. Neither pH nor pCO2 was associated with increased risk of readmission or death (Table 2). In the subset of patients with both arterial and venous blood gas parameters obtained, there was a good correlation of blood gas parameters. In Spearman correlation analysis, venous pCO2 and pH were strongly correlated with arterial pCO2 (n = 43; rs = 0.763; P = .000) and

pH (n = 42; rs = 0.852; P = .000), respectively.

  1. Discussion

For patients with acute dyspnea at the ED, TCO2 ranging from 30 to 40 mmol/L was a predictor of 1-year readmission and mortality. High and low BE was also related to poor outcome. However, this prognostic information was mediated by TCO2, as BE became nonsignificant in the combined statistical analysis. Total carbon dioxide remained related to the end point with an effect size stronger than for prior COPD.

Surprisingly, little is known about TCO2 in patients with dyspnea at the ED, its prognostic value, and its impact on patient outcome. The mechanisms causing dyspnea are still incompletely understood [1], and the knowledge of which factors contribute to the blood concentra- tion of TCO2 is largely based on experimental animal studies and theo- retical conclusions [29-33].

In the steady state at the dissociation equilibrium, TCO2 is used as a surrogate marker of bicarbonate. Both BE and TCO2 increase as a result of metabolic alkalosis. The underlying causes to these shifts are many. Possible explanations to metabolic alkalosis in the acutely dyspneic pa- tient may be diuretic treatment, hypokalemia, or posthypercapnia [28]. Total carbon dioxide additionally increases in compensation to respira- tory acidosis [28]. In this study, elevated BE did not associate with the end point when adjusting for TCO2, although it may have correlated to the end point with a larger study size. This suggests that elevated TCO2 in the dyspneic patient is not caused by metabolic alkalosis. It also suggests that the raised levels of TCO2 are not only a consequence of increased bicarbonate levels.

Table 2

Cox proportional hazard models adjusted for age, sex, METTS-A, respiratory rate, and oxygen saturation

Model 3

Quartile 1

Quartiles 2 and 3

Quartile 4

No. events/N

HR (95% CI)


No. of events/N

HR (95% CI) P

No. of events/N

HR (95% CI)




1.535 (1.080-2.181)





1.482 (1.075-2.043)




1.348 (0.942-1.930)





1.682 (1.205-2.349)




0.918 (0.507-1.663)





1.139 (0.633-2.051)




0.903 (0.638-1.277)





1.196 (0.850-1.683)


Regardless of the causal mechanism, patients with acute dyspnea and poorer outcome in terms of readmissions and mortality tended to have both elevated TCO2 and negative BE values. In the multivariate analysis, elevated TCO2 remained significant, but Base deficit became borderline significant. This finding can partly be explained by the pres- ence of hypoventilated and metabolically compensated patients with COPD. As BE is reduced by respiratory acidosis, this can also explain the trend of base deficit with increased rates of readmission and mortal- ity. In comparison to BE that only quantifies metabolic acid base disor- ders, TCO2 is a marker of both respiratory and metabolic acid base balance disorders [34,35] that, in this study, correlated to poorer out- come. Therefore, we speculate that TCO2 in patients with acute dyspnea is an indicator of uncompensated respiratory acidosis. Given this, TCO2 may be more suitable than BE for assessing acute dyspnea in the chron- ically ill patients, as it involves both metabolic and respiratory compo- nents. Total carbon dioxide may also be more suitable than venous pCO2 in predicting outcome in the critically ill with history of COPD. It is less transient in nature than the gas parameters and seems to be a bet- ter marker for long-term outcome than venous pCO2 and venous pH that showed no correlation to the end point.

In common conditions causing acute dyspnea such as COPD, pneu-

monia, ischemic heart disease, congestive heart failure, and pulmonary embolism, deranged TCO2 levels can represent a transient or permanent systemic impairment. As high TCO2 predicted poor outcome during as much as 1 year after the acute episode of illness, it can stand for an unmasked underlying cardiopulmonary fragility, which, in the long term, associates with increased readmissions and mortality rates. In the heterogeneous patient group with dyspnea, the diagnostics is chal- lenging, and a correct early management is important for prognosis. There are some biomarkers used for diagnostic and prognostic purposes in dyspnea patients. These have been included in scores often together with vital signs to predict poorer outcome [9-11], but at present, there

is no separate blood biomarker for evaluation of prognosis and risk stratification of dyspnea patients. Traditionally, arterial blood gas analy- sis has predominantly been used to evaluate blood gas and acid-base disorders. In fact, in this and recent studies, there is a good correlation of venous to arterial blood gas parameters (with the exception of pO2) [19,22-24,36]. Easily accessible venous blood gas may thus be used in the ED to add clinically important information in a broad patient group with acute dyspnea without restricting the use of arterial blood gas when indicated.

The causes of dyspnea, death, and readmission were not systemati- cally registered, and the underlying conditions leading to the events re- main unclear. However, the intention of the study was to evaluate long- term prognostic factors for all-cause readmission or death in unselected patients with acute dyspnea. The knowledge of the actual underlying conditions seems to be less important for this purpose. The diseases causing dyspnea often coexist and share the same risk factors. A distinc- tion between them is difficult to accurately assess in the clinical setting. Identification of the prognostic factors for specific diagnoses in patients with dyspnea was, therefore, not in the scope of this study. Apart from having dyspnea as the main complaint, we intentionally applied no se- lection criteria for the study population to maintain the heterogeneity and aiming at making results applicable on a random dyspneic patient seeking medical care. We do acknowledge that many patients were ex- cluded due to missing blood gas parameters. A selection bias would de- spite this probably have lead to enrichment of even more severely ill patients. Most of the patients excluded were either relatively young pa- tients not hospitalized or terminally ill patients where the ED physician judged that prognostic and diagnostic efforts were unnecessary.

Given the easy accessibility, fast response time, low costs, and negli- gible risks of a venous blood gas analysis, we find our results encourag- ing and potentially clinically applicable. Total carbon dioxide may help deciding the level of care upon admission and determining the follow-

Fig. 2. Cumulative end point curves for quartiles of total carbon dioxide during the 1-year follow-up period.

up strategies for out-of-hospital care in patients with acute dyspnea to prevent readmissions and death. It should, however, be emphasized that the sample size is moderate and replications of our results in similar patient cohorts are essential for clinical implementation.

In conclusion, a high level of TCO2 in patients with acute dyspnea is a marker for worse outcome. This easily accessible blood gas parameter may prove useful in the ED for risk stratification of dyspnea patients.

Author contributions

Conception and design: NL, AR, KG, OM Analysis and interpretation: NL, AR, PS, KG, OM

Drafting the manuscript for important intellectual content: NL, AR, PS, SE, TW, KG, OM


  1. Burki NK, Lee LY. Mechanisms of dyspnea. Chest 2010;138(5):1196-201.
  2. Palla A, Ribas C, Rossi G, Pepe P, Marconi L, Prandoni P. The clinical course of pulmo- nary embolism patients anticoagulated for 1 year: results of a prospective, observa- tional, cohort study. J Thromb Haemost 2010;8(1):68-74.
  3. Terzano C, Conti V, Di Stefano F, Petroianni A, Ceccarelli D, Graziani E, et al. Comor- bidity, hospitalization, and mortality in COPD: results from a longitudinal study. Lung 2010;188(4):321-9.
  4. El-Menyar A, Zubaid M, Sulaiman K, AlMahmeed W, Singh R, Alsheikh-Ali AA, et al. Atypical presentation of acute coronary syndrome: a significant independent predic- tor of in-hospital mortality. J Cardiol 2011;57(2):165-71.
  5. Almagro P, Calbo E, Ochoa de Echaguen A, Barreiro B, Quintana S, Heredia JL, et al. Mortality after hospitalization for COPD. Chest 2002;121(5):1441-8.
  6. Delerme S, Ray P. Acute respiratory failure in the elderly: diagnosis and prognosis. Age Ageing 2008;37(3):251-7.
  7. Hurd S. The impact of COPD on lung health worldwide: epidemiology and incidence.

Chest 2000;117(2 Suppl.):1S-4S.

  1. Seneff MG, Wagner DP, Wagner RP, Zimmerman JE, Knaus WA. Hospital and 1-year survival of patients admitted to intensive care units with acute exacerbation of chronic obstructive pulmonary disease. JAMA 1995;274(23):1852-7.
  2. Eurlings LW, Sanders-van Wijk S, van Kimmenade R, Osinski A, van Helmond L, Vallinga M, et al. Multimarker strategy for Short-term risk assessment in patients with dyspnea in the emergency department: the MARKED (Multi mARKer Emer- gency Dyspnea)-risk score. J Am Coll Cardiol 2012;60(17):1668-77.
  3. Rehman SU, Martinez-Rumayor A, Mueller T, Januzzi Jr JL. Independent and incre- mental prognostic value of multimarker testing in acute dyspnea: results from the ProBNP Investigation of Dyspnea in the Emergency Department (PRIDE) study. Clin Chim Acta 2008;392(1-2):41-5.
  4. Maisel A, Mueller C, Nowak R, Peacock WF, Landsberg JW, Ponikowski P, et al. Mid- region pro-hormone markers for diagnosis and prognosis in acute dyspnea: results from the BACH (Biomarkers in Acute Heart Failure) trial. J Am Coll Cardiol 2010; 55(19):2062-76.
  5. Ray P, Birolleau S, Lefort Y, Becquemin MH, Beigelman C, Isnard R, et al. Acute respi- ratory failure in the elderly: etiology, emergency diagnosis and prognosis. Crit Care 2006;10(3):R82.
  6. Ray P, Delerme S, Jourdain P, Chenevier-Gobeaux C. Differential diagnosis of acute dyspnea: the value of B Natriuretic peptides in the emergency department. QJM 2008;101(11):831-43.
  7. Kociol RD, McNulty SE, Hernandez AF, Lee KL, Redfield MM, Tracy RP, et al. Markers of decongestion, dyspnea relief, and clinical outcomes among patients hospitalized with acute heart failure. Circ Heart Fail 2013;6(2):240-5.
  8. Latini R, Masson S, Anand IS, Missov E, Carlson M, Vago T, et al. Prognostic value of very low plasma concentrations of troponin T in patients with stable chronic heart failure. Circulation 2007;116(11):1242-9.
  9. Januzzi Jr JL, Peacock WF, Maisel AS, Chae CU, Jesse RL, Baggish AL, et al. Measure- ment of the interleukin family member ST2 in patients with acute dyspnea: results from the PRIDE (Pro-Brain Natriuretic Peptide Investigation of Dyspnea in the Emer- gency Department) study. J Am Coll Cardiol 2007;50(7):607-13.
  10. Christ M, Thuerlimann A, Laule K, Klima T, Hochholzer W, Perruchoud AP, et al. Long-term prognostic value of B-type natriuretic peptide in cardiac and non- cardiac causes of acute dyspnoea. Eur J Clin Investig 2007;37(11):834-41.
  11. Burri E, Potocki M, Drexler B, Schuetz P, Mebazaa A, Ahlfeld U, et al. Value of arterial blood gas analysis in patients with acute dyspnea: an observational study. Crit Care 2011;15(3):R145.
  12. Lim BL, Kelly AM. A meta-analysis on the utility of peripheral venous blood gas anal- yses in exacerbations of chronic obstructive pulmonary disease in the emergency department. Eur J Emerg Med 2010;17(5):246-8.
  13. Widgren BR, Jourak M. Medical Emergency Triage and Treatment System (METTS): a new protocol in primary triage and secondary priority decision in emergency med- icine. J Emerg Med 2011;40(6):623-8.
  14. Slenter RH, Sprooten RT, Kotz D, Wesseling G, Wouters EF, Rohde GG. Predictors of 1-year mortality at hospital admission for acute exacerbations of chronic obstructive pulmonary disease. Respiration 2013;85(1):15-26.
  15. Malatesha G, Singh NK, Bharija A, Rehani B, Goel A. Comparison of arterial and ve- nous pH, bicarbonate, PCO2 and PO2 in initial emergency department assessment. Emerg Med J 2007;24(8):569-71.
  16. Middleton P, Kelly AM, Brown J, Robertson M. Agreement between arterial and cen- tral venous values for pH, bicarbonate, base excess, and lactate. Emerg Med J 2006; 23(8):622-4.
  17. Kelly AM, McAlpine R, Kyle E. Agreement between bicarbonate measured on arterial and venous blood gases. Emerg Med Australas 2004;16(5-6):407-9.
  18. ABL800 FLEX analyzer Specifications 2011. Available from: http://www. radiometeramerica.com/~/media/Files/RadiometerComCloneset/RAME/Brochure s/ Products/ABL800%20specs.pdf.
  19. Morgan TJ. Invited commentary: putting standard base excess to the test. J Crit Care 2009;24(4):492-3.
  20. Kofstad J. Base excess: a historical review–has the calculation of base excess been more standardised the last 20 years? Clin Chim Acta 2001;307(1-2):193-5.
  21. Centor RM. Serum total carbon dioxide. In: Walker HK, Hall WD, Hurst JW, editors. clinical methods: the history, physical, and laboratory examinations3rd ed. ; 1990 [Boston].
  22. Winaver J, Walker KA, Kunau Jr RT. Effect of acute hypercapnia on renal and proxi- mal tubular total carbon dioxide reabsorption in the acetazolamide-treated rat. J Clin Invest 1986;77(2):465-73.
  23. Cohen ND, Stanley SD, Arthur RM, Wang N. Factors influencing pre-race serum con- centration of total carbon dioxide in Thoroughbred horses racing in California. Equine Vet J 2006;38(6):543-8.
  24. Astrup P, Jorgensen K, Andersen OS, Engel K. The acid-base metabolism. A new ap- proach. Lancet 1960;1(7133):1035-9.
  25. Atkins EL. Assessment of acid-base disorders. A practical approach and review. Can Med Assoc J 1969;100(21):992-8.
  26. O’Leary TD, Langton SR. Calculated bicarbonate or total carbon dioxide? Clin Chem 1989;35(8):1697-700.
  27. Berend K, de Vries AP, Gans RO. Physiological approach to assessment of acid-base disturbances. N Engl J Med 2014;371(15):1434-45.
  28. Berend K. Acid-base pathophysiology after 130 years: confusing, irrational and con- troversial. J Nephrol 2013;26(2):254-65.
  29. Rang LC, Murray HE, Wells GA, Macgougan CK. Can peripheral venous blood gases re- place arterial blood gases in emergency department patients? CJEM 2002;4(1):7-15.