Department of General Medicine, Krishna Institute of Medical Sciences [Deemed to be University], Karad, Maharashtra, India - 415110.
Corresponding Author E-mail: kimssubmission1@gmail.com
DOI : //dx.doi.org/10.13005/bpj/2381
Abstract
Background Diabetic ketoacidosis [DKA] is one of the most severe complications of diabetes mellitus [DM]. Arterial blood gas analysis [ABGA] has been used as a conclusive diagnostic test for DKA. However, ABGA sampling is technically challenging, painful and may cause multiple complications. Venous blood gas analysis [VBGA] is a minimally invasive alternative for ABGA; however, the correlation between ABGA and VBGA has been controversial. Thus, the correlation between arterial and venous pH, partial pressure of CO2 [PCO2], and bicarbonate [HCO3] was studied. Objective To determine whether VBGA can replace ABGA in the evaluation of patients presenting with DKA. Material and Methods The present observational study was carried out in 76 patients with DKA. Two samples for ABGA and VBGA were obtained from each patient as close to each other as possible and were immediately sent to the laboratory. Data analysis was done using Pearson’s correlation coefficient [R] and Bland and Altman plots. Results The Bland and Altman plots and Pearson’s correlation coefficient depicted excellent agreement between arterial and venous pH [R = 0.69] and acceptably good agreement between arterial and venous PCO2 [R = 0.93] and HCO3 [R = 0.82]. Conclusion Thus, VBGA can be used in the initial diagnosis and evaluation of DKA, allowing the utilisation of a minimally invasive, safer option to ABGA.
Keywords
Blood Gas Analysis; Diabetic Ketoacidosis; Diagnostic Tests; Hydrogen-Ion Concentration
Download this article as:Immadisetty S. K, Patange A. P. Comparison of Blood Gas and Acid–Base Measurement in Arterial and Venous Blood Samples in Patients with Diabetic Ketoacidosis. Biomed Pharmacol J 2022;15[1]
Copy the following to cite this URL:Immadisetty S. K, Patange A. P. Comparison of Blood Gas and Acid–Base Measurement in Arterial and Venous Blood Samples in Patients with Diabetic Ketoacidosis. Biomed Pharmacol J 2022;15[1]. Available from: //bit.ly/3KuChnd
Introduction
Diabetes mellitus [DM] is a group of common metabolic disorders characterized by hyperglycaemia. It is due to a defect in insulin action, insulin secretion or both. A host of genetic and environmental factors are responsible for this. Diabetes as well as the complications arising from it are major causes of death in many countries [Trachtenbarg, 2005].
Diabetic Ketoacidosis [DKA] is one of the most severe complications of DM characterized by hyperglycaemia, ketonemia, ketonuria, and metabolic acidosis [Trachtenbarg, 2005; Jameson et al., 2018]. The annual incidence in the western countries has been reported as 4.6 cases per 100,000 of patients with DM and 0.14 cases per 100,000 of the general population [Ganie et al., 2012]. Although inpatient mortality rates in western countries are low [< 1%], 30% of hospitalised DKA cases result in inpatient death in India [Farsani et al., 2017]. This metric proves the severity of this complication in India.
DKA is biochemically characterised by hyperglycaemia [blood glucose > 250 mg/dL], a blood pH of < 7.30, and a bicarbonate [HCO3] level of ≤ 18 mmol/L and can be categorized as mild [pH 7.25–7.30, HCO3 15–18 mmol/L], moderate [pH 7–7.24, HCO3 < 10 mmol/L] or severe [pH of < 7, HCO3 < 10 mmol/L] [Trachtenbarg et al., 2005; Cashen and Petersen, 2019]. Arterial blood gas analysis [ABGA], which measures the blood pH, partial pressure of CO2 [pCO2], and bicarbonate [HCO3] levels, is an essential diagnostic test in patients with suspected DKA as metabolic acidosis is a prominent feature of DKA [Kitabchi and Wall, 1995; Gokel et al., 2000].
However, arterial blood sampling is a technically challenging and painful procedure associated with risks such as haemorrhage, pain, artery damage or thrombosis, infection, aneurysm formation, and even loss of limb function. Additionally, it increases the risk of needlestick injury to the healthcare personnel [Roberts et al. 2017; Mortensen, 1967]. Therefore, many studies have tried to search for an alternative to ABGA and have compared arterial and venous blood gas in patients with DKA in an attempt to eliminate ABGA in the initial diagnosis and evaluation of DKA [Gokel et al., 2000; Brandenburg and Dire, 1998; Kelly, 2006] Venous blood gas analysis [VBGA] is a minimally invasive procedure that bypasses all the complications of ABGA. Although the correlation between arterial and venous blood parameters [especially pH] is well established, several studies have conflicting opinions on the same [Kelly, 2006; Brashear et al., 1979].
Therefore, this study was designed to determine whether VBGA could replace ABGA as the initial diagnostic measure in patients with DKA by correlating the values.
Material and Methods
Study design
The present cross-sectional, observational, comparative study was conducted on 76 patients above 18 years of age with DKA admitted to the medical emergency ward of a tertiary care centre in Karad, Maharashtra from November 2017 to May 2019 after institutional ethics committee clearance was obtained. Farsani et al., [2017] conducted a systematic review on the incidence and prevalence of DKA, and reported an overall prevalence of 50–100 events per 1000 adult patients of DM. Hence, the error of the study was set at 5% and the power of the study was set at 95%. Using the formula for cross-sectional studies,
N=4pq/d2
the sample size came to 76 patients. Sample allocation was done using convenient sampling technique. Patients > 18 years with random blood sugar > 250 mg/dl, urinary ketone bodies, HCO3 level < 18 mmol/L, and pH < 7.35 with known or newly detected DM [Type 1/Type 2] were included in the study. Patients not fulfilling the abovementioned criteria were excluded from the study. Data were recorded in the study proforma of consenting individuals.
Data collection
A pre-tested validated proforma was developed to collect data for the research purpose. A detailed case history of the patient was taken including chief complaint, history of present illness, family and medical history; a thorough clinical examination was performed; and the necessary investigations [blood and urine tests] were done.
A sample of arterial blood [0.5–1.0 mL] for ABGA was obtained from the radial artery of the patient using a 2.5-ml syringe [Dispovan, India] [Lwanga and Lemeshow, 1991] whereas a sample of venous blood [0.5–1.0mL] for VBGA was obtained from a peripheral vein at the time of venepuncture for other laboratory reports. The two blood gas samples were obtained as temporally close to each other as possible before the initiation of treatment and were immediately sent to the laboratory.
Statistical Analysis
Data were analysed using the statistical software R version 3.6.3 and MS Excel. Categorical variables such as age group, type of DM, severity of DKA, and clinical symptoms were represented by frequency tables and continuous variables were represented by mean ± SD form. The strength of association between arterial and venous pH, PCO2, and HCO3 was measured using Pearson’s correlation coefficient [R]. A P-value less than or equal to 0.05 was considered statistical significance. The degree of agreement between the arterial and venous measurements were evaluated on Bland and Altman plots.
Results
Table 1 presents the demographic variation and characteristics of the patients with DKA. Of the total 76 patients, 53 [69.74%] were males and 23 [30.26%] were females, with a mean age of 55.99 ± 16.19 years. Out of 76 patients, 9 [11.84%] had type 1 DM and 67 [88.16%] had type 2 DM. Also, 59 [77.63%] patients were known cases of DM and 17 [22.37%] were newly diagnosed cases of Diabetes Mellitus on admission.
Out of 76 patients, 35 [46.05%] patients had mild DKA and 9 [11.84%] patients had severe DKA, whereas 15 [19.74%] patients had mild hyperglycaemia [RBS 250-350 mg/dl] and 26 [34.21%] patients had severe hyperglycaemia [RBS >450 mg/dl]. The mean RBS was 415.55 ± 79.37 mg/dl.
Table 1: Demographic Variation and Characteristics of Patients with Diabetic Ketoacidosis
Distribution of patients based on Age and GenderAge GroupNumber of PatientsMaleFemaleTotal6022 [28.95%]9 [40.79%]31 [40.79%]Total53 [69.74%]23 [30.26%]76 [100%]Distribution of patients based on History of Diabetes MellitusType of DMNumber of PatientsKnown case of DMNewly diagnosed DMTotalType15 [6.58%]4 [5.26%]9 [11.84%]Type254 [71.05%]13 [17.11%]67 [88.16%]Total59 [77.63%]17 [22.37%]76 [100%]Distribution of patients based on clinical symptomsClinical symptomsNumber of PatientsAbdominal pain35 [46.05%]Breathlessness/Dyspnea19 [25%]Nausea/Vomiting34 [44.74%]Distribution of patients based on Severity of DKA and Random Blood Sugar Sub-CategoryNumber of Patients
Severity of Diabetic Ketoacidosis
Mild [pH 7.25-7.35 and HCO3 15-18]35 [46.05%]Moderate [pH 7.00-7.25 and HCO3 10-15]32 [42.11%]Severe [pH 15 in both arterial and venous blood [i.e. 56.58% and 64.47%, respectively].Table 2: pH range, PCO2 range and HCO3 range in Arterial and Venous Blood.
Sub-CategoryNumber of PatientsArterialVenouspH 15.0043 [56.58%]49 [64.47%]Figure 1 depicts the Bland-Altman plots for arterial and venous pH, PCO2, and HCO3. Figure 1a illustrates that majority of the points lie within the interval and are close to the line representing mean difference. Hence, there was excellent agreement between Arterial pH and venous pH. Figure 1b illustrates that majority of the points lie within the interval but are scattered. Hence, there was acceptably good agreement between Arterial PCO2 and Venous PCO2. Figure 1c illustrates that, majority of the points lie within the interval but are scattered. Hence, there was acceptably good agreement between Arterial HCO3 and Venous HCO3.
Click here to view figure
Pearson’s Correlation test proves that there was significantly high positive correlation between arterial parameters and venous parameters [Table 3]. The scatter plots in Figure 2 depict the same.
Table 3: Correlation of arterial and venous pH, PCO2, and HCO3.
ParametersCorrelationp-valuepH0.83< 0.001*PCO20.96< 0.001*HCO30.91