Implementation of an arterial blood gas indication algorithm in cardiac surgery

Arterial blood gasses (ABGs) account for an estimated 10-20% of all costs during an ICU stay. Non-clinically indicated ABGs increased costs of care, lengths of stay, ventilator days, and line days, increasing the risk of adverse outcomes in already vulnerable critically ill patients. A cardiac surgery intensive care unit (CSICU) within a large urban mid-Atlantic academic medical center accounted for 31% of the entire institution's ABG analyses between 2018-2019, was identified as a top utilizer due to inappropriate ordering practices compared to current guidelines.

The purpose of this quality improvement project was to implement an algorithm using evidence-based guidelines that identified appropriate standardized clinical indications for ABGs, with the intention of reducing non-clinically indicated blood gas analyses orders within the CSICU. Anticipated outcomes of this practice change included decreasing the total volume of ABGs sent, resulting in reduced costs of care, lengths of stay, and improved morbidity and mortality rates.

An evidence-based ABG indication algorithm was created focusing on acute changes in oxygenation, ventilation, acid base balance; changes in hemodynamics, post-operative baseline, and for patient ABGs to correlate with extra-corporeal membranous oxygenation values. Routine ABGs for monitoring were eliminated. Implementation occurred over fourteen-weeks in the fall of 2020 following staff and provider education. Training emphasized the use of non-invasive monitoring such as pulse-oximetry and capnography. Compliance and gross laboratory totals and indications were obtained from weekly auditing.

There was an 8.8% reduction in ABGs obtained and 32% decrease in ABGs per patient day. The most common indications were extra-corporeal membranous oxygenation (ECMO)-correlated ABGs, post-operative, and changes in oxygenation and/or ventilation; 7.8% were non-indicated.

Implementation of an ABG indication algorithm resulted in fewer ABGs sent, mostly due to a reduction in routine monitoring, and ABGs were more likely to be clinically indicated in response to an acute concern. Implementing an ABG indication algorithm is safe, feasible, and can lead to significant cost reductions for the institution.