Employing the HOSPITAL score
In another effort to reduce hospital readmissions, Jacques Donzé, MD, MSc, associate physician, Bern University Hospital, Switzerland, and research associate, Brigham and Women’s Hospital, Boston, and his colleagues used the HOSPITAL score to identify patients at high risk of 30-day potentially avoidable readmission.
To most efficiently reduce hospital readmissions, hospitals need to target complex and intensive discharge interventions for patients at high risk of potentially avoidable readmission who are more likely to benefit.2 “However, prior research indicates that clinical health care providers are not able to accurately identify which patients are at high risk for readmission,” Dr. Donzé said.
Dr. Jacques Donze
In their large international multicenter external validation study, Dr. Donzé and his colleagues found that the HOSPITAL score accurately predicted the risk of 30-day potentially avoidable readmissions. The HOSPITAL score is easy to use and can be calculated before discharge, which makes it a practical tool for identifying patients at high risk for preventable readmission and the timely administration of high-intensity interventions designed to improve transitions of care.2
Dr. Donzé believes that several factors may influence the performance of a prediction model, such as the initial selection of the potential predictors, the quality of the derivation method, including readily available predictors commonly available, and including reliable factors that aren’t subject to subjective evaluation. “All of these factors can play a role in the performance and generalizability of the HOSPITAL score,” he said.
When a patient is identified as high risk to be readmitted, hospitalists can take certain actions to prevent readmission. “Interventions are more likely to be effective when they include several components,” Dr. Donzé said. “These include follow-up phone calls and/or home visits, review of the patient’s medication list, patient education, and sending a discharge summary to the patient’s primary care physician in a timely manner. For now, enough evidence for a specific effective multimodal intervention to be generalizable to the majority of patients is lacking.”
Currently, the HOSPITAL score has been validated in approximately 180,000 patients in 14 hospitals across five countries and three continents – always showing good performance and generalizability. The HOSPITAL score includes seven variables readily available before hospital discharge, is easy to use, and is the most widely validated prediction model for readmission, Dr. Donzé said.
Before being implemented into practice, a score should ideally reach the highest level of validation, that is, show its clinical impact. “We expect that the score will not only be able to accurately predict high-risk patients, but using the score will also impact patient care by reducing readmissions when coupled with an appropriate intervention,” Dr. Donzé said.
In summary, research has shown that a variety of methods can be used to reduce hospital readmissions, including studying inpatient and outpatient physicians’ perspectives regarding factors contributing to readmission; interviewing patients regarding readmissions; and identifying patients at high risk of readmission using the HOSPITAL score.
Many researchers are continuing their studies in these areas.
Karen Appold is a medical writer in Pennsylvania.
Using hospitalist reflections as a means to reduce readmissions
Readmission studies and the development of readmission scoring systems and prediction tools rely on data from a large number of patients, typically extracted from administrative databases.
To complement this data, Deanne Kashiwagi, MD, consultant, Hospital Internal Medicine, Mayo Clinic, Rochester, Minn., and her colleagues asked hospitalists to reflect upon the readmissions of patients for whom they cared to add insight into the culture of patient care transitions within the health system.
“We felt there was some value in considering these nuances of the local care environment, which may not be represented in studies drawing from large databases, as potential targets for readmission efforts,” she said.
Dr. Deanne Kashiwagi
Dr. Kashiwagi and her colleagues developed a chart review tool to guide hospitalists through reflection about their patients’ admissions and readmissions. “We included factors frequently cited in the literature as contributors to readmissions and added factors that our study group, after a chart review of 40 patients’ readmissions, identified as variables contributing to our own patients’ readmissions,” Dr. Kashiwagi said. “Some of these variables reflected our local care system, such as our staffing model, which led to some patients being cared for by more than two hospitalists during their admission. The study group considered such variables as potential contributors to our own group’s readmissions, but they were not necessarily common readmission risk factors identified in large-scale studies.”
Dr. Kashiwagi believes that including elements of local practice and culture was the strength of their work. “Groups interested in replicating this reflective process should consider including factors specific to their practices that may contribute to readmission,” she said.
Asking hospitalists to perform reviews has led to implementing changes. Physicians were prompted to schedule earlier follow-up appointments and nurse practitioners and physician assistants have worked to improve the quality of their discharge summaries. The exercise also engaged hospitalists to suggest system changes that might contribute to decreased readmissions, such as a geriatrician-run service (which was recently begun) to provide multidisciplinary acute geriatric care for hospitalized older adults.
“Although large-scale studies are clearly important, readmission review at a more granular level may have merit as well,” Dr. Kashiwagi said, noting that such reviews identify local practice factors that groups may quickly act upon to help decrease readmissions. “Hospitalists readily engaged in this reflective exercise, which yielded actionable information to decrease readmissions.”
In commenting on why a different similar study7 didn’t mimic the results of Mayo Clinic’s study, Dr. Kashiwagi said there were some differences in methodology that may explain the difference in readmission rates. “First, this group excluded patients on dialysis, which in our study was a common comorbidity of our readmitted patients,” she said. “It is also notable that the chart review tool was different. Perhaps there is less representation of local factors, unique to that hospitalist group and their practice culture, than on our review form. These investigators also discussed their readmissions at routine intervals. Additionally, their preintervention readmission rate was lower than Mayo Clinic’s group, and although the readmission rate trended downward postintervention, it did not reach statistical significance.”