Many healthcare organizations still struggle to plan long-term operations according to the highly variable and seemingly chaotic nature of patient demand.
Bryan Kennedy, RN, director of nursing at AnMed Health in Anderson, S.C., discussed how data analytics improved forecasting and predictive utilization processes at AnMed during a webinar sponsored by McKesson Nov. 4.
Economic, environmental challenges in capacity management
Increasing economic, labor and scheduling pressures have made accurate forecasting a strategic priority for hospital systems in the transition to value-based care. But the inherently variable nature of patients' need for medical services still presents a significant challenge to maintaining optimal operating efficiency.
I. Highly variable census. A hospital's geographic location can influence patient volume throughout the year. For instance, a hospital in Vail, Colo., typically sees increased volumes in the winter, when tourists flock to the mountains for winter sports. If not managed properly, census variability can negatively impact employee satisfaction and financial performance.
II. High labor costs due to agency staff and overtime. Labor is typically hospitals' greatest operating expense. Inefficient alignment between patient demand and hospital resources can drive increased labor costs, cause sluggish patient flow and artificially inflate a hospitals' average length of stay. Staffing employees in proportion to patient volume, therefore, is crucial to reducing overhead.
III. Average length of stay. About 68 percent of webinar registrants said their hospital's average length of stay was greater than they liked. Reducing patients' days admitted can improve customer experience and operational performance.
Case Study
AnMed Health, a five-hospital, 700-bed system in upstate South Carolina, implemented an integrated capacity management model to better manage staffing, seasonal and economic variability.
Integrated capacity management refers to a resource management process where staff incorporate data analytics into strategic planning and scheduling practices. Traditionally, hospitals relied on historical data and anecdotal experience to predict trends in census and staffing. Integrated capacity management combines historical information with predictive analytics to significantly increase the accuracy of volume forecasts, such as patient flow and throughput, to improve hospital staffing.
AnMed partnered with McKesson to develop and implement a predictive analytics system modeled on the platform McKesson offers to New Zealand hospitals. AnMed was the first hospital to adopt and use McKesson Capacity Planner and McKesson Performance Visibility in the U.S. healthcare market. Capacity Planner uses analytics to predict patient demand and Performance Visibility uses a visual dashboard to represent patient care and resource allocation.
"Through this [implementation] process, we learned that just because something is variable does not mean it is unpredictable," said Mr. Kennedy. In fact, "we found a 98 percent or greater accuracy in predictions."
Integrated forecasting positively impacted AnMed's operational and financial performance, as well as physician engagement and patient satisfaction. Using data analytics, AnMed was able to measure and predict patient admissions and discharges to each particular unit at its hospitals. This helped nurses and staff stay on track with patient discharges and bed turnovers, causing average length of hospital stay to decrease by a half-day in nine months.
Looking at capacity data also enabled administrators to be more proactive in identifying and addressing staffing challenges, said Mr. Kennedy. For instance, the data revealed certain times when particular units were under- or over-staffed and caused friction among employees and physicians. Pinpointing and devising strategic initiatives to support clinicians generates a huge return in clinical performance, patient care and employee job satisfaction, Mr. Kennedy said.
Barriers to overcome when implementing data-driven processes
Mr. Kennedy identified the following issues administrators may encounter when implementing an integrated capacity management model.
I. Gain executive buy in. Engaged and effective leadership plays a fundamental role in guiding healthcare staff through times of instability and change. Oftentimes, successful change management largely depends upon a leader's ability to show staff why change is necessary and positive, Mr. Kennedy said.
II. Transform hospital culture. Administrators may encounter resistance to implementation by displacing traditional practices in strategic planning. At the beginning, employees may find it difficult to incorporate data analytics into capacity management planning.
III. Help frontline staff incorporate data. Executives and administrators are more accustomed to using projections and data to plan long-term strategy. Clinicians and frontline staff, on the other hand, are often more consumed by the day-to-day needs. Mr. Kennedy recommended administrators provide frontline staff with support to see the value of data analytics.
IV. Build employee trust in the data. Mr. Kennedy found data analytics refuted some long-held assumptions about patient flow trends and staffing at AnMed. This caused frontline staff to mistrust the data and made process transformation more difficult than anticipated. "It was a challenge to win clinician trust, but they eventually recognized the data was more accurate and bought into the concept," Mr. Kennedy said.