A study finds cost savings but no mortality benefit by Edward Doyle
Published in the February 2011 issue of Today's Hospitalist
HOW DO HOSPITALIST COMANAGEMENT SERVICES affect patient care, quality and costs? A new study that asked those questions has uncovered some answers that may be disconcerting to hospitalists.
The study, which was published in the Dec. 13/27 Archives of Internal Medicine, looked at a neurosurgery comanagement service run by hospitalists at the University of California, San Francisco (UCSF). While researchers concluded that the service reduced costs to the tune of $1,400 per patient, they found that the hospitalists made no significant difference in patient mortality, readmission rates or length of stay.
"Mortality in neurosurgery may be driven by things that hospitalists do not fix."
–Andrew D. Auerbach, MD University of California, San Francisco
D. Auerbach, MD, a hospitalist researcher and assistant professor of medicine at UCSF, admits that the data finding no clinical benefit for neurosurgery patients were "disappointing." He points out, however, that the study also revealed some positive results. In addition to cost savings, patients on the hospitalist service reported higher levels of satisfaction, and so did other providers.
Why no benefit in outcomes?
Perhaps one of the biggest questions about the study had to do with its conclusions that hospitalists didn't change patient outcomes like mortality, readmission rates or length of stay. Dr. Auerbach says a likely reason for a lack of any mortality benefit is that mortality isn't very common in neurosurgery patients. Even if hospitalists did produce a benefit, he explains, the total number of patients helped could have been so small that researchers wouldn't detect it.
But there may be an even simpler reason. "Mortality in neurosurgery may be driven by things that hospitalists do not fix," Dr. Auerbach says, "like the severity of intracranial hemorrhage."
Still, the absence of any benefit in terms of mortality, length of stay or readmission rates is particularly puzzling when you consider the types of patients admitted to the service. Neurosurgery patients were seen by hospitalists only if they had a history of CAD, CHF, serious arrhythmias, COPD, chronic kidney disease, ischemic stroke, diabetes requiring insulin therapy or longterm anticoagulation therapy. In other words, the comanagement service was seeing relatively sick patients.
"These patients weren't elective knee replacement patients," Dr. Auerbach says. "They were pretty sick, and many came in with emergency surgery."
He also points out that his isn't the first study on comanagement to conclude that comanagement arrangements don't greatly improve outcomes. Studies from about 10 years ago examining the impact of hospitalist comanagement on orthopedic surgery patients similarly found little difference in patient outcomes.
Cost savings, better satisfaction
Not all of the results were glum. Researchers found that hospitalist comanagement produced savings of $1,439 per admission, a finding that Dr. Auerbach says marks the first time cost savings have become apparent when looking at hospitalist comanagement. According to researchers' calculations, the service saved the hospital about $1.5 million over the year and a half included in the study period. That represents a big re- turn on investment for the service, which cost about $750,000 to run during the same period.
Where did cost savings come from? Dr. Auerbach hypothesizes that costs were lower for hospitalists because they had a better feel for conditions that could be handled in the outpatient environment. "Hospitalists might be more comfortable than a neurologist saying that a patient could have a problem worked up as an outpatient," he says.
The study is also the first in the literature to find that comanagement can boost patient satisfaction. Patients treated by hospitalists showed higher levels of satisfaction in three areas: how well staff responded to concerns, how cheerful the hospital was, and how well staff addressed patients' emotional needs.
Dr. Auerbach suggests that those results might be explained by the fact that hospitalists are used to inserting themselves into situations where they've never met the patient before and have to get up to speed very quickly.
"Hospitalists have been pretty good finding ways to introduce themselves to patients who have their own primary care physician," he says. "A similar approach might go a long way when you're working on comanagement services."
Finally, when nurses, residents and attending surgeons were surveyed, they rated hospitalists particularly well on their treatment of medical issues, which was another big plus.
Changing the model
What do these decidedly mixed results on hospitalist comanagement mean? Dr. Auerbach says that because the data showed little benefit from hospitalist interventions, "Maybe comanagement in its current incarnation just does not work as well as we think it should."
While the conventional wisdom says that surgery patients—particularly those who are older and sicker— will benefit from more medical attention, Dr. Auerbach says, "For some reason, it has not shaken out that way."
In his mind, these data lead to some important questions: Are current comanagement models not intensive enough? Should hospitals be more aggressive to see if ramped-up comanagement models produce more value?
Dr. Auerbach wonders, for instance, if hospitals need to embrace a model that's more like the ICU. Patients would be admitted to a hospital medicine comanagement service that featured intensive group management, with shared rounding schedules and geographically localized services.
While Dr. Auerbach was disappointed by the findings, he's not entirely surprised. He points to other instances in medicine where what seems obvious doesn't always produce the expected results. Rapid response teams are the darling of the quality improvement community, for example, but research has been mixed on the impact such teams have on outcomes.
He is also quick to point out that the study results are somewhat similar to early research on hospitalists. While many of those studies found that hospitalists produced cost savings and reductions in length of stay, they found few improvements in clinical outcomes.
"When you look at the numbers and find that an effort hasn't done anything to save lives, that can be a little disappointing," Dr. Auerbach admits. "But on the other hand, if you implement something that makes everyone feel like they are working in a better, safer environment, that has benefits."