Three years ago, Andrew Masica, MD, MSCI, joined the MedProvider Inpatient Care Unit hospitalist group at Baylor University Medical Center (BUMC) in Dallas just as the national debate on Medicare recidivism rates was focusing on high-risk populations.
Dr. Masica’s master’s degree in clinical investigation, combined with the roughly 35 hospitalists operating at the 900-bed BUMC, suggested it made sense to see what Baylor’s doctors could add to the conversation. And a study was born: “Reduction of 30-Day Post-Discharge Hospital Readmission or ED Visit Rates in High-Risk Elderly Medical Patients Through Delivery of a Targeted Care Bundle.” The single-center study will be published in this month’s Journal of Hospital Medicine.
The study found readmission/ED visit rates were lower after 30 days for those given an individualized care bundle of educational information compared with those who received the center’s standard treatment (10% individualized care bundle compared with 38.1% for standard treatment, P=0.04). Analysis also showed that for those patients who had a readmission or post-discharge ED visit, the time interval to the second event was longer in the intervention group compared with the control group (36.2 days to 15.7 days, P=0.05). At 60 days, however, readmission/ED visit rates were not affected positively for the intervention group versus the control group (42.9% vs. 30%, P=0.52).
The study team emphasizes that its small sample size—20 in the intervention group, 21 in the control—make it nearly impossible to extrapolate the results to large population sets; however, the results fuel the debate. “We don’t want to overstate our conclusions,” says Dr. Masica, the principal study investigator. “Important questions need to be asked. Is it the specific characteristics of the care coordinators? Can you reproduce this at other facilities? Is it the care bundle or the personnel? …We view this as early-phase work that people can build upon.”
Still, Dr. Masica believes hospitalist-centric conclusions can be reached. Since the study used in-house personnel only, other HM groups could easily reproduce the bundle without added expense. Additionally, because the coordinated-care approach involves a checklist of patient interaction activities, not medical procedures, the barrier to replication is further reduced. However, hospitalists will need the cooperation of more than their own medical directors.
In BUMC’s case, that meant the assistance of patient-care support services and the pharmacy department. Liz Youngblood, RN, MBA, supervised the care coordination in her role as vice president of patient-care support services for the Baylor Health Care System. Brian Cohen, PharmD, MS, was the pharmacy lead. Dr. Masica notes the confluence between departments was one of the keys to the reduction in recidivism over the first 30 days post-discharge.
“If you pick the high-risk patients and deliver the care in a bundle, you would be able to improve outcomes,” Dr. Masica says. “When you deliver just pieces of the care—just the coordinated care or just the pharmacist—you get inconsistencies.”
The first struggle BUMC researchers encountered—once they secured funding from Baylor’s Institute for Health Care Research and Improvement—was enrolling enough patients who met the criteria set for the study. The high-risk patient thresholds were:
- At least 70 years old;
- Regular use of at least five medications;
- At least three chronic, comorbid conditions;
- Assistance with at least one activity of daily living; and
- Preadmission residence at home or at an assisted-living facility with a reasonable expectation of disposition back to that residence.
Researchers also wanted patients with common DRGs admitted, and set exclusion criteria as well: lack of fluency in English; admission primarily for a surgical procedure; terminal diagnosis with life expectancy of less than six months; and residency in a long-term care facility. Patients who could not be enrolled within 72 hours of admission were excluded.
Dr. Masica notes hospitalists interested in replicating the research should pay attention to the consent forms they used. When the Baylor team conducted its research from March to September 2007, they used a long-form consent waiver. Baylor’s consent form for similar studies has since been shortened, and Dr. Masica says a less complicated form would have helped encourage more patients to enroll. In the end, 60 patients declined to enroll in the Baylor study and 56 were unable to give their consent due to impairment.
Once enrolled, the patients were delivered the care bundle in stages (see Table 1). Care coordinators (CCs) saw patients daily, instructing them on specific health conditions with an eye toward teaching home care, should post-discharge problems arise. Clinical pharmacists (CPs) visited patients to focus on medication reconciliation and education. CCs and CPs would follow up with post-discharge phone calls to confirm receipt of medical equipment and medications, use and affects of those medications, home-health arrangements, and to schedule follow-up appointments. If patients indicated any issues, the coordinators recommended action plans.
“It would be surprising to find out how little patients really understand about why they’re in the hospital and what they’re being treated for,” Youngblood says. “To have the reinforcement is really valuable.”
One topic the study skirts is the ever-contentious realm of post-discharge care and who takes over responsibility for patient care. While the Baylor study examined readmission/ED visit rates through 60 days, Dr. Masica says a transitional-care program is the best way to manage that care continuum.
“We did see a difference at 30 days,” Dr. Masica says. “At 60 days, that effectively washed out. That makes sense. You can only control things so much from the hospital side. After 30 days, you need transitional care, good primary care.”
Baylor’s research team is working on a follow-up study that would apply the coordinated-care bundle to specific disease management. Youngblood notes that directing specific services at a targeted population—for example, congestive heart failure patients—should show an even more concentrated reduction of 30-day recidivism. “The key is to identify the high-risk groups,” he says. “You can’t apply this to every single patient. That would be low-yield. Your yield is going to come in on the very high-risk folks.” TH
Richard Quinn is a freelance writer based in New Jersey.