Proper Use of the National Inpatient Sample Database to Analyze Safety of Cardiovascular Procedures

Analysis of the National Inpatient Sample database for the safety of procedures that are primarily outpatient in nature can be biased due to the selective hospitalization of patients who have complications during these procedures.

This is a retrospective cross-sectional analysis of hospitalized patients in the National Inpatient Sample data set who underwent an atrial fibrillation catheter ablation between 2016 and 2022. Patients were divided into 2 groups based on the day of hospitalization on which the ablation was performed: day 1 of admission or before (group 1), versus day 2 of admission or later (group 2). Patient baseline characteristics and procedural complication rates were compared.

The analysis included 47 925 weighted hospitalizations of patients who underwent atrial fibrillation catheter ablation. Group 1 had significantly fewer octagenerians (6.9% versus 10.4%), lower rates of renal disease (12.6% versus 20.0%), heart failure (28.3% versus 45.8%), and ischemic heart disease (16.0% versus 20.5%), P<0.001 for all, compared with group 2. The rate of complication was higher in group 1 (9.5% versus 6.6%; P<0.001). Multivariate analysis showed study group 1 to be an independent predictor of complications (odds ratio, 1.54 [95% CI, 1.30-1.83]; P<0.001).

Patients undergoing procedures on the day of admission or before were younger and had fewer comorbidities. Despite this, these patients had a significantly higher rate of procedural complications. The likely explanation for these results is selection bias, where patients undergoing outpatient procedures who are at higher risk or develop complications are admitted as inpatients, thus falsely inflating the complication rate of patients in the National Inpatient Sample data set.