Objective: Arterial blood gas (ABG) analysis is essential in the clinical assessment of potential acutely ill patients. Venous to arterial conversion (v-TAC), a mathematical method, has recently been developed to convert peripheral venous blood gas (VBG) values to arterialized VBG (aVBG) values. The aim of this study was to test the reliability of aVBG compared to ABG in an emergency department (ED) setting.

Method: Twenty ED patients were included in this study. ABG and three aVBG samples were collected from each patient. The aVBG samples were processed in three different ways for comparison: aVBG1 was held steady and analysed within 5 minutes; aVBG2 was tilted in 5 minutes and analysed within 7 minutes; aVBG3 was held steady and analysed after 15 minutes. All VBG samples were arterialized using the v-TAC method. ABG and aVBG samples were compared using Lin's Concordance Correlation Coefficient (CCC) and Bland-Altman's analysis.

Arterial blood gas (ABG) analysis is essential in assessment of respiratory and metabolic status in acutely ill patients. In comparison to peripheral venous blood (PVG) sampling, the ABG sampling procedure is more painful for the patient and technically more challenging for the clinician to perform. Other drawbacks of ABG sampling include adverse events such as subcutaneous hematoma, arterial thrombosis or embolization, and the serious, though rare, complication pseudoaneurysms.

Peripheral venous blood gas (VBG) sampling has been suggested as an alternative to the ABG procedure. This procedure causes less patient discomfort and the sample can be analysed in combination with other venous blood tests. Studies have revealed that pH and bicarbonate have good correlation, whereas venous and arterial blood gasses (pO2 and pCO2) show low agreement. A recent systematic review comparing ABG and VBG in the ED setting found similar results and concluded that venous pCO2 was not an accurate representative of arterial pCO2.

However, a new method has been developed to calculate ABG values mathematically from peripheral venous blood by use of venous to arterial conversion (v-TAC) software (Obimedical, Denmark), supplemented with oxygen saturation measured by pulse oximetry. The principle of the method is a mathematical transformation of VBG values to arterialized values (aVBG) by simulating the transport of blood back through the tissue. To facilitate this simulation the following physiologically relevant assumptions were made: 1) The peripheral extremity was well perfused; 2) change in base excess across the tissue sampling site was approximately zero; 3) the respiratory quotient (rate of CO2 production and O2 utilisation over capillaries) could not vary outside the range 0.7 and 1.0, and 4) the haemoglobin concentration was constant from artery to vein. Initial testing of the method in an emergency department setting showed acceptable clinical congruence between arterial and mathematically arterialized pH and pCO2 with a small difference on 0.001- 0.024 and 0.00 0.46 kPa, respectively. However, inaccurate values of pO2 were seen when oxygen saturation measured by pulse oximetry was above 96%, due to the flat shape of the oxygen dissociation curve (ODC).

The aim of this study was to test appropriate practical handling of venous blood gas samples and evaluate the reliability of the v-TAC method, in an acute medical emergency setting in awake and circulatory stable patients capable of giving consent.

Methods Patient inclusion The study was conducted in the ED at North Denmark Regional Hospital from September through October 2015 in daytime. This hospital is 24-hour hospital with a collective medical and abdominal surgery emergency department, with 7-10.000 annual admissions per year.

Circulatory stable patients needing ABG analysis for respiratory and metabolic assessment were selected randomly for participation in the study. Patients were considered circulatory stable if systolic blood pressure was above 90 mmHg and heart rate was 50 to 110 beats/min in accordance with Danish Emergency Process Triage (DEPT), which were used to triage patients upon admission. A total of 30 patients were included; 10 patients for a pre-study purpose and then 20 patients in the following main study. Allocation to either the pre-study or the main study was performed by simple quasi-random allocation in order of admission. The clinical indication for ABG analysis was decided by the responsible physician in the ED upon patient admission and based on national guidelines and criteria.

Blood collection In the pre-study, venous samples were collected in paired 2 mL ABG syringes and 4.5 mL tubes from each of the 10 patients, to determine which blood collection method was preferred. VBG samples were collected via a butterfly needle with a three-way stopcock in conjunction with routine venous blood sampling upon admission. VBG samples were collected by the biomedical laboratory technician in the same manner as PVB samples in the normal clinical setting. Results from the pre-study were used to determine the preferred blood collection method in the main study. In this study, paired ABG and VBG samples were collected simultaneously from each of the 20 patients. The ABG samples were collected by the responsible physician.

Blood analysis Blood for VBG analysis collected in the syringe and 4.5 mL tube in the pre-study were analysed within five minutes after sampling. In the main study blood for VBG analysis was collected in three 4.5 mL tubes and converted to arterialised VBG (referred to as aVBG). Each aVBG tube was processed differently as follows: aVBG1 was held steady and analysed within five minutes of sample collection, aVBG2 was tilted in five minutes and analysed after seven minutes and aVBG3 was handled as aVBG1, but analysed after fifteen minutes. ABG samples were analysed within five minutes after sampling. All ABG and VBG samples were analysed with ABL800 blood gas analyser (Radiometer, Denmark) an VBG samples were mathematically converted to aVBG using v-TAC software which was integrated into the ABL800 analyser. In our hospital, the reference values for ABG parameters are as follows: pH 7.37-7.45, pCO2 4.30-6.00 kPa, and pO2 9.60-14.4 kPa. All results from analysed ABG and aVBG samples were registered automatically in the hospital database. On the standard of care basis, only the ABG results were used as the usual standard reference in the medical evaluation of the patients, who took part in the study.

Ethics and data protection The Danish Research Ethics Committee in the North Denmark Region was notified. Since the v-TAC method has previously been approved to be used in clinical research and blood sampling was performed as routine practice, ethical approval was not required. This study was approved by the Danish Data Protection Agency.

Inclusion Criteria:

• Admission to the emergency department.
• Need for ABG for respiratory or metabolic assessment.

Exclusion Criteria:

• Circulatory unstable patients (systolic blood pressure < 90mmHg og heart rate <50beats/min or >110beats/min).

Circulatory stable patients needing ABG analysis for respiratory and metabolic assessment were selected randomly for participation in the study. Patients were considered circulatory stable if systolic blood pressure was above 90 mmHg and heart rate was 50 to 110 beats/min in accordance with Danish Emergency Process Triage (DEPT), which were used to triage patients upon admission. A total of 30 patients were included; 10 patients for a pre-study purpose and then 20 patients in the following main study.

Sensitive individual participant data (IPD) will not be available to other researchers. IPD will be kept in a secure personal drive within the regions server. All sensitive IPD will be anonymized before publication.