Data Analytics

How Hospitals Use Analytics to Staff Up Before a Rush

Healthcare providers are using data analytics to forecast staffing requirements, which improves care and reduces costs.

Twitter

Wylie Wong is a freelance journalist who specializes in business, technology and sports. He is a regular contributor to the CDW family of technology magazines.

Frequent midday surges in patient traffic recently left emergency department teams at Bergen New Bridge Medical Center hustling to handle the influx.

They also prompted the 1,070-bed hospital in Paramus, N.J., to use data analytics to better measure and predict staffing levels with the hope of reducing the need for overtime or calling in employees at the last minute due to an unexpected rush.

A high-level review of patient data helped the organization do just that. As a result, it added a shift at 11 a.m., when extra patients were most anticipated.

Bergen, which is in the early stages of its data analytics implementation, plans to increase its use of predictive analytics across the system and with community partners. A recent IT infrastructure upgrade — featuring new Cisco Unified Computing System servers, Dell EMC storage and Cisco networking gear — powers the effort, says CIO Jennifer D’Angelo.

“We now have the foundation in place to track and trend data and be more predictive with it,” she says.

An increasing number of healthcare providers are investing in data analytics, and some early adopters are using predictive analytics to optimize staffing levels. The ability to take recent historical data and run algorithms allows management to forecast patient volume on any given day in advance of an expected rush or slowdown.

The result is improved patient care and cost savings — plus boosted employee morale thanks to a more dependable work schedule, hospital leaders say.

“Any improvement in knowing the right number of people and the right number of skills to have at a hospital at any given time is important,” says Cynthia Burghard, research director at IDC Health Insights. “It improves operational efficiency and effectiveness and improves the bottom line for hospitals.”

Vanderbilt Uses Analytics to Estimate Demand for Surgery

Predictive analytics has helped Vanderbilt University Medical Center improve operating room staffing levels over the past six years, says Vikram Tiwari, the organization’s director of surgical business analytics.

His tools: a Dell OptiPlex desktop and a Dell Precision 3530 notebook, each with an 8th Generation Intel Core processor. Tiwari uses SQL queries to extract data from VUMC’s data warehouse for analysis. He also relies on IBM SPSS statistical software and open-source R analytics software.

Although Vanderbilt’s finance department ably predicts expected surgery case volumes in its annual operating room staff budget, monthly estimates can be off by as many as 30 surgeries, Tiwari says — which equates to about 10 to 12 operating rooms’ worth of staffing.

“Every week, there are days that we could have excess staffing, or not enough and have to call in nurses, or the existing staff could stay later for overtime pay,” says Tiwari, also an associate professor of anesthesiology and biomedical informatics at Vanderbilt. “Managing these swings is the objective.”

Jennifer D’Angelo, CIO, Bergen New Bridge Medical Center

Jennifer D’Angelo, CIO of Bergen New Bridge Medical Center, says the center is becoming a more data-driven healthcare provider. Photography by Colin Lenton

It’s why Tiwari has built a model that analyzes up to a year’s worth of historical surgery data and compares it to scheduled elective surgeries for each of the next 30 days. It also accounts for potential last-minute emergency surgeries because VUMC, located in Nashville, Tenn., is a Level I trauma center.

Through predictive analytics, Tiwari is able to estimate daily demand for surgery to within seven cases of the actual number 80 percent of the time. Operating room managers automatically receive reports each morning that show the latest surgical caseload projections for the next 14 to 30 days, which enables them to make staffing adjustments as needed.

If the model predicts that the case- load is below expectations for a few days during the next two weeks, managers can reduce staffing levels on those days, Tiwari says.

Data Analytics Predicts Number of Daily ER Visits

Envision Physician Services, a Nashville, Tenn.-based multispecialty medical group providing care in more than 900 facilities nationwide, has built a web-based application that can accurately predict the number of expected daily emergency room visits with an accuracy rate of 80 percent to 85 percent.

The tool can also forecast what time patients are expected to arrive, what ailments they will have and what services they will need, says Dr. Kirk Jensen, Envision’s chief innovation officer.

Housed in the public cloud, the app automatically grabs patient data from the company’s billing system and physicians’ schedules from the company’s scheduling software and stores the data in a Microsoft SQL Server database. The app uses Microsoft’s Power BI software to run reports.

“Physician and operational leaders have the ability to make sure our capacity is matched against patients’ needs,” Jensen says.

80-85%

Accuracy rate of web application built to predict the number of expected daily emergency room visits

Source: Dr. Kirk Jensen, Chief Innovation Officer, Envision Physician Services

Executives can review the reports to validate their current schedules and shift budgets if necessary. During the year, if something changes, such as significant population growth, a demand spike from flu season or a desire to change hospital processes, they can run the model again to see how capacity and workflow are affected.

Envision released the app in early 2019, and about 50 to 100 emergency department partners now use it to forecast physician staffing needs. In the coming years, Envision intends to share the app with the several hundred other emergency departments it supports.

For now, the app makes a big difference. “We can schedule the right type of clinician and the right number of clinicians to match incoming workflow needs,” Jensen says. “It ensures that clinical quality remains high. It improves flow, and we avoid long wait times.”

Bergen New Bridge Invests in IT Modernization

In New Jersey, a $20 million IT modernization effort has allowed Bergen New Bridge Medical Center to become a more data-driven healthcare provider, D’Angelo says.

In 2017, Bergen deployed Dell EMC’s VxBlock converged infrastructure, which features Dell EMC Unity all-flash storage, Cisco UCS B-Series servers and switches, and VMware software. The hospital also upgraded its network with Cisco Nexus switches, wireless controllers and access points, and Palo Alto Networks firewalls.

The facility, which used to have spotty wireless access, now has full Wi-Fi coverage throughout. The company also invested in new Lenovo desktops and notebook computers and new applications, including a major electronic health record upgrade, to support the hospital’s data analytics goals.

“Stabilizing and modernizing our network created the foundation we needed to move forward, to go beyond looking at data and move toward creating effective ways of analyzing it,” D’Angelo says. That progress, she adds, is critical for advancing interoperability.

During the past year, Bergen has integrated its EHR with the New Jersey Health Information Network, which provides access to patient health records from other healthcare providers. Now, when patients are admitted to Bergen’s emergency room, clinicians can quickly see patients’ medical histories.

The hospital will invest in a data warehouse in 2020 to pool data to bolster its predictive analytics efforts, including the ability to predict staffing needs in advance.

“We’re not quite there yet with the data automation,” D’Angelo says. “But access to data and real-time decision-making and having that predictive analytics component is in our future.

gpointstudio/Getty Images