Reducing Length of Stay for Veterans Hospitalized With Pneumonia

This is a pilot study to implement a computer based intervention to safely reduce length of hospital stay and time to conversion to oral antibiotics for patients with pneumonia.

Pneumonia is the 3rd leading medical discharge diagnosis in the VA, and although significant attention has been focused improving pneumonia process of care measures less attention has been paid to resource use, specifically length of hospital stay (LOS). Based on our preliminary analyses LOS is considerably higher in the VA versus other non-federal hospitals. Major causes of this increased length of stay are delayed conversion from intravenous (IV) to oral therapy upon reaching clinical stability, and not discharging on the same day as the conversion to oral antibiotics. Therefore additional interventions are needed to decrease the LOS for veterans hospitalized with pneumonia in the VA system. The long-term goal of this research is to develop sustainable approaches to reduce LOS and thereby improve efficiency of care in veterans admitted with pneumonia to VA Medical Centers through the implementation of computerized clinical stability reminders. Although prior research, including our own, has demonstrated moderate effectiveness of resource intense methods to implement this criteria (e.g., utilization management nurses) new, more effective and less resource intense, strategies are needed to accomplish this long-term goal. The design of the computerized clinical stability reminder, the likely range in magnitude of response to implementation of the intervention, logistics of data collection, data management, and development of study instruments are needed prior to commencement of a more definitive, multi-center study. Hence, we are seeking funding to conduct the necessary pilot work needed to successfully design and implement a large multi-center randomized intervention study to reduce LOS for veterans with pneumonia. Our objectives include: 1) Develop and perform a local pilot implementation of an inpatient clinical reminder within Computerized Patient Record System to identify when veterans hospitalized with pneumonia are clinically stable and ready for conversion form IV to oral antibiotic therapy and hospital discharge. 2) Test measures of the constructs of the Theory of Planned Behavior (TPB) in VA medicine ward physicians and determine if they correlate with intention and discharge behavior. 3) Establish a collaborative of VA Medical Centers for the purpose of developing a larger scale implementation study to use evidence-based criteria to reduce LOS for patients hospitalized with pneumonia. 4) Prepare a VA HSR&D IIR grant proposal for a cluster randomized controlled trial of a clinical reminder intervention to reduce length of hospital stay among veterans hospitalized with pneumonia. The proposed pilot study will develop and implement a computerized-based inpatient clinical reminder to assist physicians with appropriate conversion from IV to oral antibiotics and discharge on the same day for veterans hospitalized with pneumonia. In addition to creating and implementing the clinical reminder at single tertiary care VA teaching hospital, we will examine physician attitudes and potential barriers/facilitators to the introduction of inpatient clinical reminders using the theoretical model of TPB. We will also assemble a cohort of VA medical centers with investigators and chiefs of medical staff willing to participate in a multi-center randomized control trial of this intervention. The results from this pilot study will be used to plan a larger, more definitive multi-center trial of this inpatient clinical reminder with the goal of significantly reducing LOS for veterans hospitalized with pneumonia in the VA health care system.

computer based reminder to change patients from IV to oral antibiotics based on evidence based clinical practice guidelines

Inclusion Criteria: Patient hospitalized with outpatient-acquired pneumonia at South Texas Veterans Health Care System during study period. Exclusion Criteria: none