Risk Score IDs Patients Safe for Same-Day Discharge After Low-Risk TAVI

The model performed well, but its strength lays largely in predicting those at low risk for readmissions, says Grant Reed.

Risk Score IDs Patients Safe for Same-Day Discharge After Low-Risk TAVI

NEW YORK, NY—A clinical risk score designed for practicality can accurately predict low-risk patients who might be eligible for same-day discharge after TAVI, according to a new study.

Asa Phichaphop, MD (Minneapolis Heart Institute, MN), who presented the data today at New York Valves 2024, previously known as TVT, said the risk score, which includes nine variables easily documented during the preprocedural workup, can help physicians plan procedures and prepare patients for going home the same day TAVI is performed.

“Same-day discharge started during the COVID-19 era where we tried to minimize hospital admissions,” Phichaphop told TCTMD. “At Minneapolis Heart, where I work with Paul Sorajja, we tried to capture the patients who are low risk for an in-hospital event. That means, we can know from the beginning which patient is low risk and who has the potential for same-day discharge. If the procedure goes well for that patient, they stay in hospital for a short postoperative period and, if they’re still doing well, go home the same day.”

The risk score was derived from 1,035 patients undergoing TAVI between 2016 and 2021 at their institution. Patients treated with general anesthesia were excluded, as were those with a prior cardiac implantable electrical device, leaving 730 patients eligible for the analysis. In-hospital events occurred in 12.5%, largely driven by the need for permanent pacemaker because of advanced AV block (9.0%).

Researchers identified multiple patient characteristics from prior studies associated with adverse in-hospital outcomes and inserted these clinical and imaging variables into a random forest machine-learning model, which is used for classification and regression analyses. The final risk model included nine variables—age, body mass index, history of MI, presence of significant CAD, previous right bundle branch block, left ventricular ejection fraction, calcium at the left ventricular outflow tract (LVOT), diameter of the femoral artery, and valve type (percentage of balloon-expandable valves).  

In terms of diagnostic performance of the risk model, the area under the curve was 0.76, which is deemed fair for predicting risk. For patients at low risk for complications, the model was more accurate because most of the TAVI patients included in the model’s development had a low risk of in-hospital adverse events, said Phichaphop. Low-risk patients were defined as those with a predicted risk of in-hospital complications of less than 3%, while high-risk patients were those with a risk 6% or higher.

Works Well at Low End of Risk

Ashraf Hamdan, MD (Rabin Medical Center, Petah Tikva, Israel), who moderated the session, questioned why the researchers did not include more CT-derived imaging variables in the model, such as annular calcification, which he believes is a more important predictor of annular rupture than LVOT calcification. Additionally, the researchers opted not to include septal length measurements even though these can predict AV block, he said.

Phichaphop said they considered those parameters but wanted to focus on a risk score that was easy to use with variables that are easily collected. “We wanted to make it feasible for more utilization,” he said, adding that there is evidence showing that LVOT calcification is linked with annular rupture, too.

Grant Reed, MD (Cleveland Clinic, OH), one of the panel discussants, said their center also discharges patients home the same day after TAVI, as long as they meet certain criteria. These include transfemoral TAVI performed under conscious sedation, 6-hour post-TAVI bedrest with rhythm monitoring, and no major complications or need for further observation. Patients must also be stable (ECG and hemodynamics), comfortably ambulated, and have in-place social support to help with recovery.  

Patients with preexisting right bundle branch block, incident conduction disease, or a change in the PR interval during the procedure are not eligible for same-day discharge, said Reed.  

“Most of what drives readmission and major complications in the period immediately post-TAVR is [the need for a permanent] pacemaker and stroke,” said Reed. “You’re very likely to notice bleeding right away, but stroke can be delayed. Pacemaker can be delayed. These are the features that I think are really important.”

As Phichaphop noted, Reed said the model performs well for patients at low risk for readmission, but he stressed that physicians don’t “want to be burned” discharging a patient home the same day who later returns with a complication. Where the risk model isn’t quite up to par is in the medium- and high-risk patients, said Reed.

To TCTMD, Phichaphop pointed out that two-thirds of the cohort used to derive the risk score were treated with a balloon-expandable valve. “What we found is that most of the patients develop AV block at the time of the procedure, or within an hour of the procedure,” he said. After 6 hours recovery without any ECG changes, they believe the patient can be safely discharged, particularly in the context of balloon-expandable valves.

Next steps for the group include attempts to validate the risk score in other populations as well as a prospective study to see how well it performs in the real world, said Phichaphop.

Michael O’Riordan is the Managing Editor for TCTMD. He completed his undergraduate degrees at Queen’s University in Kingston, ON, and…

Read Full Bio
  • Phichaphop A, Okada A, Fukui M, et al. Clinical risk score prediction for same-day discharge after TAVR. Presented at: New York Valves 2024. June 6, 2024. New York, NY.

  • Phichaphop and Hamdan report no conflicts of interest.
  • Reed reports consulting fees/honoraria/speakers’ bureau payments from Boston Scientific and Edwards Lifesciences.