ICD’s Incongruity: Same-day Discharge Safe But May Not Cost Less

Results from a randomized, controlled clinical trial that evaluated the safety and cost effectiveness of same-day discharge for elective implantable cardioverter-defibrillator (ICD) procedures may reassure centers that already provide the option as well as those considering it. But whether the approach meets the promise of lowering costs is anyone’s guess.

Practice meets proof

The standard of care in the United States for ICD implantation for primary prevention of sudden cardiac death generally calls for overnight observation to ensure that the patient isn’t at risk of implant-related complications. But technological advancements and the adoption of techniques that reduce the risk of complications should allow physicians to shorten that timeline to a same-day procedure in selected low-risk patients. So reasoned Ranjit Suri, MD, president of Heart Rhythm Associates and a cardiologist at the Mount Sinai School of Medicine in New York City.

At the Heart Rhythm Society’s (HRS) 2016 scientific sessions in San Francisco, Suri unveiled preliminary findings from a randomized, controlled clinical trial he co-led that compared the safety and cost effectiveness of same-day discharge with an overnight stay in low-risk ICD patients who did not require bridging anticoagulation using heparin. The multicenter trial had been given a green light after Suri and colleagues proved its feasibility in a single-center pilot study (J Cardiovasc Electrophysiol 2012;23:1123-1129).

The trial showed that same-day discharge was as safe as next-day discharge, with a procedural complication rate of 3.2 percent vs. 1.5 percent (Heart Rhythm 2016;13[6]:1371-1372). Counter to expectations, though, they found no statistically significant difference in overall costs per patient between the groups.   

“The trial largely confirmed what has been a growing practice in the United States,” says HRS President Michael R. Gold, MD, PhD, director of the cardiology division and a professor at the Medical University of South Carolina in Charleston. “It is an important piece of information that adds to our growing comfort. The safety of these devices continues to improve and the safety of the procedure continues to improve as the technology improves and the monitoring capability improves.”

With advancements in remote monitoring, the technology leapfrogged forward between the pilot’s enrollment period of 2007 to 2008 and the clinical trial’s window of 2014 and 2016. “Remote monitoring was in its infancy during the pilot,” Suri observes, and it was hobbled by transmission limitations and patients’ technical skills. The clinical trial, on the other hand, required implantation of an ICD that was compatible with St. Jude Medical’s Merlin@home transmitter. “We eliminated that by having remote monitoring using a wireless system independent of the patient’s home phone or technological capabilities.” Suri points out all the major vendors offer similar attributes.

Short of the goal

The trial originally set a target enrollment of 600 participants, but it was stopped early in January with an initial sample size of 313. Of those, only 265 patients were randomized, leaving the study underpowered. Among other challenges, temporal trends in ICD utilization led to a smaller pool of potential candidates. “The overall ICD volumes have gone down,” Suri says.

The rate for primary ICD implantations in the United States skyrocketed before the pilot study, climbing from 6.1 per 100,000 person population in 1993 to 46.2 in 2006, according to an analysis of Nationwide Inpatient Sample data (Pacing Clin Electrophysiol 2010;33[6]:705-711). Another Nationwide Inpatient Sample analysis found a similar trend, with ICD implantations growing from 95,062 in 2003 to a peak of 143,262 in 2006 (Clin Cardiol 2016;39[2]:63-71). That total steadily declined after 2006, dropping to 86,457 by 2011.

[[{"fid":"22822","view_mode":"media_original","type":"media","attributes":{"height":318,"width":577,"alt":" - icd-utilizations-millions","class":"media-element file-media-original"}}]]

Other physician and hospital factors also may have helped stack the deck against the trial. The two-midnight rule, which the Centers for Medicare

& Medicaid Services established in fiscal year 2014, may have changed practices. In its June 2015 report to Congress, the Medicare Payment Advisory Commission noted that concerns over denial of payments flagged by Medicare recovery audit contractors (RACs) for short inpatient stays prompted some hospitals to increase their use of outpatient observations.

The report also noted that the diagnosis-related group (DRG) code for cardiac arrhythmias and conduction disorders was one of six DRGs that “also generated the most RAC overpayment revenue.” In other words, this is where RACs, which earn contingency fees based on denials, seemed to make money.

“In many places, the horse is already out of the barn, so to speak,” Gold says. “Many people have accepted this approach and, based on their own anecdotal experience, are maybe less willing to randomize patients—not because they think it is dangerous, but because they have trouble keeping patients in the hospital or the hospital may be pushing back, saying, ‘Why don’t you just send them home anyway?’”

In his presentation at HRS.16, Suri also noted the different payment scenarios and said there was an “opportunity to save hospitals some money if they adopt this [same-day discharge] approach.” 

[[{"fid":"22824","view_mode":"media_original","type":"media","attributes":{"height":512,"width":600,"style":"font-size: 13.008px; width: 180px; height: 154px; margin: 5px; float: left;","alt":" - m-gold","class":"media-element file-media-original"}}]]

In addition, same-day discharge was already practiced at some centers, and has been for at least a decade. An analysis that linked the National Cardiovascular Data Registry’s ICD Registry data to Medicare claims from 2006 to 2009 found that 5.3 percent of elective cases had been discharged on the same day as admission.

Rates varied, with 64 percent of hospitals reporting no same-day discharges and 2 percent with more than half their cases clocking same-day procedures. The rate increased modestly over the three-year span, from 4.1 percent to 6.6 percent. Paul L. Hess, MD, MHS, then a fellow at the Duke Clinical Research Institute in Durham, N.C., and now a cardiologist at the VA Eastern Colorado Health Care System and an assistant professor at the University of Colorado School of Medicine in Aurora, and colleagues also found no difference between the groups in rates of death, all-cause readmissions or device-related readmissions.

“The long-term perspective is to think about ICD technology advancement over time,” Hess says, calling less invasive implantation techniques “a big advancement. Now that we have remote monitoring, it looks like the stage is set even better for same-day discharge.”

The clinical trial’s findings, even with their statistical limitations, provide reassurance for practices like his that offer shorter stays, Gold says. His practice developed a protocol for selecting patients, preparing them for the possibility of an early discharge and educating them before and at discharge. Hess and Suri have suggested that the societies develop guidelines or a consensus statement outlining protocols for institutions.

“In general, when you start to go against what has been standard of care, even if it is very logical, it is always a good idea to come up with protocols so you can both standardize practices” and perform quality assessments, Gold says.

Counterintuitive costs

The cost analysis casts some doubt on the argument that a shorter stay lowers costs. The study found that per-patient costs were actually lower by $1,334 for next-day discharge, although the findings were not statistically significant. Costs were all over the map, with some institutions charging more than $59,000 for implantations, same day or next day, while others kept mean costs to about $12,347 for same-day and $10,251 for next-day discharges.

[[{"fid":"22825","view_mode":"media_original","type":"media","attributes":{"height":353,"width":577,"alt":" - same-day-next-day-discharge-icd-implantation","class":"media-element file-media-original"}}]]

Suri blamed “dodgy accounting” for the variation. “The whole country has a great deal of difficulty understanding the true cost of doing anything in a hospital,” he says. “We were struck by the wide variation in the reported cost and reported charges by institutions, even in the same city. … We would need thousands of patients to dilute this dodgy accounting.” 

Suri’s group is not the first to report this variation. An issue brief published by the Berkeley Center for Health Innovation found that total costs for ICD implantations in 45 hospitals in California ranged from $24,078 to $57,347 in 2008, for instance.

Hess, whose research focuses on value, points out that the value calculation takes into account outcomes and costs. “In thinking about healthcare value broadly, the numerator would be outcomes,” he says. “I think the data are emerging that those are pretty neutral or pointing in that direction. The denominator in this case is costs. Heterogeneity in that can impact looking at this at a broader scale.”

Hospitals may use different approaches for determining costs, including how they calculate the use of an electrophysiology lab, personnel time and other resources, Gold says. “It reflects the challenges in healthcare institutions,” he says. “Many institutions have trouble understanding their costs and how they prorate their costs.”

Given these stumbling blocks, same-day discharge advocates may have a difficult time proving the value of a shorter stay. But for cardiologists who implant devices, the trial was “an important step forward” with data that will help them improve patient care, Gold says.

“When I sat and watched [Suri’s presentation], it made me feel good, that what we are doing is the right thing,” he says. Colleagues at the conference concurred with him. “For us, it was very reassuring.”