Published in the March 2008 issue of Today’s Hospitalist
WHEN IT COMES TO finding heart failure in hospitalized patients, hospitalists have their work cut out for them. There’s no single lab test that can definitively pinpoint the condition, and the symptoms of heart failure mirror many other illnesses.
“Heart failure can be difficult to pin down,” says Lakshmi Halasyamani, MD, a hospitalist at St. Joseph Mercy Medical Center in Ann Arbor, Mich. “A typical symptom like shortness of breath could be attributed to pneumonia or COPD, or it could be a side effect of medications.”
The delay in a heart failure diagnosis raises concerns on several levels. For one, Dr. Halasyamani explains, the time that it takes physicians to reach diagnostic certainty could be better spent helping patients recover and educating them about the disease.
But identifying patients with heart failure early is also important for hospital performance. Although heart failure is one of Medicare’s costliest DRGs, says Dr. Halasyamani, studies show the rate of conformity with all Joint Commission heart failure measures in hospitals is far from 100%.
After reviewing data on heart failure diagnosis, Dr. Halasyamani and her colleagues at St. Joseph’s decided to take action. “We needed to do something to improve the care of our patients with heart failure,” she says, “and we had to figure out how to find them in real time.”
Dr. Halasyamani led the creation of a two-step process to identify heat failure patients at her 487-bed community hospital. The details of that process, along with the results, were published in the November/December 2007 issue of Journal of Hospital Medicine.
The first step entailed using the receipt of a loop diuretic as a possible surrogate for a heart failure diagnosis; adding additional predictors improved the sensitivity and specificity slightly. (Echocardiograms were not included because their results were not available electronically.)
The second step was to embed an intervention to improve care once a patient was identified as having heart failure.
Dr. Halasyamani’s team found that receipt of an IV loop diuretic predicted a discharge diagnosis of heart failure with a sensitivity of 0.89 and a specificity of 0.87. Screening additional lab and other data boosted sensitivity to 0.91 and specificity to 0.89.
Using the screening strategy, physicians were able to shave an entire day off the amount of time it took to diagnose inpatients with heart failure.
That extra day allows a multidisciplinary team “which consists of nurses, pharmacists and nutritionists “to begin educating patients about required lifestyle changes, including the need to weigh themselves daily, watch what they eat and learn to use a pillbox.
In addition, the extra day gives patients more time to absorb information and ask questions. “When you’re in the hospital and listening to all the different things doctors and nurses are telling you,” Dr. Halasyamani says, “it can be overwhelming. Patients need time to understand how they can optimally manage their heart failure.”
A one-hour-per-week intervention
The screening strategy has been so successful that it continues to be used at the hospital today.
Because all orders at St. Joseph’s are entered in a computerized system, they can be tracked on a daily basis. Each morning at 5:30 a.m., after filtering out pediatric and obstetric patients, a list of inpatients treated with an IV diuretic is generated. A clinical pharmacist or case manager reviews those patients’ medical records to determine if they needed the diuretic for heart failure or another condition.
“This review takes one minute per patient,” says Dr. Halasyamani. While total review time needed depends on the number of patients listed that day, the range is from five to 15 minutes per day, she says, or the equivalent of one work hour per week.
The downside of screening patients? “You screen a lot of patients who do not have heart failure,” she admits. Still, given how little time it takes to screen patients and the benefits of early identification, “we felt it was worth the effort.”
If the loop diuretic order is indeed linked to heart failure, the screening nurse alerts team members to begin educating patients. (Each member of the team has a role in the patient’s care, including education and medication review.) The nurse also flags and puts a sticker on the chart to notify other clinicians.
Obstacles and benefits
When Dr. Halasyamani and colleagues began their two-year study, they tried to integrate the process into workflow. “We didn’t hire anyone,” she points out. “We embedded screening and education in the usual workflow of the pharmacists and nurses, and automated as much as we could.”
Her team did have the advantage of access to the hospital’s computer physician-order entry (CPOE) system, which can identify medication orders electronically. But even at hospitals without CPOE, she notes, pharmacies have some process for dispensing medications. By tapping into that process, a pharmacy could still lead the identification of these patients by alerting the team when a patient receives IV loop diuretics.
Still, she says, the process isn’t foolproof. Occasionally, for example, the indicators turn up false positives. When that happens, Dr. Halasyamani explains, “approximately 80% of the time, it’s clear from the chart that the patient doesn’t have heart failure. For times it isn’t clear, we’ve built in a conversation with the patient. False positives are addressed at the physician level.” She estimates that a hospitalist gets one call per month regarding a false positive.
The innovations have also reduced the possibility of dropping information during verbal handoffs.
When patients are admitted on a Saturday, for example, they are cared for over the weekend by one group of doctors. When the next team of hospitalists comes in on Monday, patients’ heart failure information is flagged on their chart “even when the condition is not their primary illness.
Before and after
The screening strategies have also help physicians significantly improve the hospital’s performance on core heart failure measures, specifically rates of giving discharge instructions, Dr. Halasyamani points out.
Still, she is planning improvements. “We are hoping to put the alert at the time the medication is ordered,” she says. “That way, we can eliminate the intermediate step of looking at the list.”
Even without that next innovation, screening has brought big payoffs. “Before, we would monitor our data in terms of how well we were doing after the fact,” Dr. Halasyamani explains. “Now, instead of finding out that we missed things after the patient was discharged, we can find out while the patient is in our midst.”
Ingrid Palmer is a freelance health care writer based in West Chester, Ohio.