From the 2019 HVPAA National Conference
Dr. Katie Chakrabarti (University of Michigan), Dr. Apu Chakrabarti (University of Michigan), Dr. Alex Miller (University of Michigan), Dr. Mark Ziats (University of Michigan), Ms. Liz Spranger (University of Michigan), Ms. Linda Bashaw (University of Michigan), Dr. Lauren Heidemann (University of Michigan)
Background
The anion gap (AG) is a well-documented calculation based on sodium, chloride, and bicarbonate values to identify, diagnose, and manage acid/base derangements. The generally accepted normal range for AG is 4-12. Unnoticed elevations of the AG may represent missed opportunities to intervene upon pathological acid/base abnormalities and may result in increased adverse events.
Objective
Our goal was to reduce the number of patients discharged with elevated anion gap by implementing automated reporting of the AG within the electronic medical record (EMR) in our healthcare system.
Methods
Our study population included adult patients discharged from the emergency department (ED) or inpatient setting at our healthcare system between February 1, 2017 and July 31, 2017. The primary outcome was percentage of total discharges with elevated AG. We utilized a higher cutoff for AG of ≥15 to ensure true abnormal values were captured. For current state analysis we examined data from February 1, 2017 through July 31, 2017 which revealed 0.479% of patients are discharged with elevated AG. Root cause analysis revealed that AG elevation may go unnoticed by providers if they do not manually calculate AG. We hypothesized that automated calculation of AG in the “Results Review” tab of the electronic medical record (EMR) would serve as an effective countermeasure to reduce the primary outcome. This intervention was implemented in January of 2018. We then compared the primary outcome pre- and post- intervention for 6 month periods using a chi-squared test for the ED and inpatient setting (Stata, College Station, TX).
Results
There were 88,285 total discharges (32.8% ED) and 92,030 total discharges (32.43% ED) in 2017 and 2018 respectively. From 2017 to 2018 there was a statistically significant decrease in the number of discharges with elevated anion gap (0.479% vs 0.401%, p=0.0121). On further analysis, this decrease was most significant in the discharges from inpatient setting (0.189% vs 0.121%, p=0.0024).
Conclusions
The automatic calculation and report of the AG within the EMR was associated with a decrease in the number of adult patients discharged with an elevated anion gap, particularly from the inpatient setting.
Clinical Implications
The addition of the AG to results review was a simple intervention that helped to avoid potentially unsafe discharges within our hospital system. More work is needed to better characterize the clinical characteristics of patients discharged with elevated AG.
