Unplanned 30-day hospital readmission after PCI related to comorbidities, discharge location
Unplanned readmission to the hospital within 30 days of PCI, which occurred in 9.3% of more than 800,000 patients, is associated with a number of factors, including place of discharge and baseline comorbidities, according to recently published data.
Researchers found that 30-day readmissions were more common among older patients (67.3 vs. 64.5 years; P < .001), women (39.4% vs. 31.3%; P < .001) and those admitted for the index procedure during the weekend (22.8% vs. 22.3%; P = .016).
Patients readmitted at 30 days also had more frequent in-hospital complete heart block (1.4% vs. 1.1%; P < .001), transient ischemic attack or stroke (3.7% vs. 3.1%; P < .001), cardiogenic shock (4.3% vs. 3.1%; P < .001), cardiac arrest (2.4% vs. 2.2%; P = .016), acute kidney injury (1% vs. 0.6%; P < .001), major bleeding (1.1% vs. 0.7%; P < .001), need for blood transfusion (0.11% vs. 0.05%; P < .001) and vascular complications (1% vs. 0.7%; P < .001). Additionally, length of hospital stay for the index PCI was longer among patients with unplanned readmission at 30 days (4.7 vs. 3.9 days).
Predictors of readmission
Independent predictors of unplanned 30-day readmission included index hospitalization discharge against medical advice (OR = 1.91; 95% CI, 1.65-2.22), transfer to short-term hospital for inpatient care (OR = 1.62; 95% CI, 1.38-1.9), discharge to a care home (OR = 1.57; 95% CI, 1.51-1.64) and chronic kidney disease (OR = 1.5; 95% CI, 1.44-1.55).
Independent associations were also found between increased likelihood for unplanned readmission and greater number of comorbidities (OR = 1.18; 95% CI, 1.17-1.18) and higher Charlson Comorbidity Index score (OR = 1.28; 95% CI, 1.27-1.29).
However, unplanned 30-day readmission was less likely among patients with private insurance (OR = 0.67; 95% CI, 0.64-0.7), no insurance (OR = 0.69; 95% CI, 0.64-0.74), elective index admission (OR = 0.7; 95% CI, 0.66-0.74) and drug-eluting stents (OR = 0.82; 95% CI, 0.8-0.85).
The mean total cost of the index hospital stay for PCI and first readmission was $37,524 compared with $23,211 for the index hospital stay alone.
More than half of hospital readmissions (56.1%) were for noncardiac causes, including nonspecific chest pain, infection, gastrointestinal disease, respiratory disease and major bleeding complications. The most common cardiac causes for readmission were CAD, including angina, HF, acute MI, arrhythmias and pericarditis.
The study included 833,344 patients undergoing PCI between 2013 and 2014 in the U.S. Nationwide Readmission Database, which uses discharge-level data of hospitalizations from 21 participating states, accounting for slightly less than half of the U.S. population and hospitalizations.
The completeness of the study’s data, its inclusion of patients with various types of insurance and differentiating between planned and unplanned readmissions strengthen the study’s findings, but information on inappropriate and appropriate hospital readmissions is lacking, Ankur Kalra, MD; Mehdi H. Shishehbor, DO, MPH, PhD; and Daniel I. Simon, MD, all from the Harrington Heart and Vascular Institute, University Hospitals Cleveland Medical Center, wrote in an accompanying editorial. Shishehbor is also a member of the Cardiology Today’s Intervention Editorial Board.
“Although we should strive to reduce all admissions, developing risk models that identify patients at risk for inappropriate readmissions is a priority,” they wrote. “In addition, a comprehensive approach that extends patient care from the hospital facilities to patients’ homes in a continuum will be indispensable to curb inappropriate readmissions.”
They also highlighted the growing percentage of PCIs that are performed as same-day discharge or in observation status that are not included in readmission data.
“Because the percentage of these patients is increasing in most programs, quality efforts should also focus on understanding the factors that lead to rehospitalization and admit status for this group of patients,” they wrote. – by Melissa Foster
Disclosure: This study was conducted as part of one author’s PhD research, which was supported by Biosensors International. One author reports he has received unrestricted research grant support from Boston Scientific, Haemonetics, Heartflow and Philips Volcano; has received speaker fees from Boston Scientific and Heartflow; has received travel sponsorship from Edwards and Eli Lilly/Daiichi Sankyo; and has served as a consultant for Haemonetics. All other authors report no relevant financial disclosures. Kalra reports he is a consultant for Medtronic. Shishehbor reports he is a consultant and advisor to Abbott, Boston Scientific, Medtronic and Philips. Simon reports he has received honoraria for educational activities from Medtronic.