While not as mature as its use in other areas, health systems across the U.S. are beginning to adopt predictive analytics tools for labor cost forecasting.
Many organizations have automated administrative tasks in revenue cycle and finance and use predictive analytics in backend functions to streamline supply chains. How can these technologies support workforce planning?
Becker’s connected with four leaders to discuss where these tools have the greatest potential — and where they face the most significant barriers — in healthcare labor.
Editor’s note: Responses were lightly edited for clarity and length.
Q: Is your organization adopting or considering predictive analytics for workforce planning or labor cost forecasting?
Beth Alpers, BSN, RN. Chief Human Resources Officer at University of Missouri Health Care (Columbia): We recently implemented a predictive analytics and machine learning technology to increase our predictive abilities as part of our long-term workforce strategy. We recognize its potential to improve labor cost forecasting, optimize staffing, and align resources with patient demand, but adoption requires careful planning, foundational standard work and integration with existing systems. We are also monitoring emerging predictive ecosystem platforms that integrate EHRs, human resources information systems and external epidemiological data for more robust forecasting.
Keisha Malivert. Executive Director of Workforce Strategy at AdventHealth (Altamonte Springs, Fla.): We are actively building workforce forecasting models that help us better understand future clinical talent needs. These tools give early insight into anticipated hiring gaps, shifts in nationally based talent availability and the pacing of other recruitment pathways — including workforce development and international recruitment, which can provide long-term stabilization for key roles. While we are still early in building a fully mature predictive workforce planning model, a significant amount of foundational work is already underway.
Carlos Vargas. Vice President of Human Resources and Technology at Adventist HealthCare (Gaithersburg, Md.): Adventist HealthCare’s use of predictive analytics for workforce planning and labor cost management provides a powerful, forward-looking capability that strengthens financial stability, enhances operational performance, and supports the quality and safety of patient care. Because labor represents the largest share of healthcare operating expenses, the ability to forecast staffing needs with precision enables leaders to better manage costs, reduce dependence on overtime and agency labor, and improve budget accuracy, resulting in significant and sustainable savings. In addition to financial benefits, predictive models help anticipate changes in patient volume, acuity and service demand, allowing clinical leaders to deploy staff more effectively, maintain safe nurse-to-patient ratios and reduce the risk of quality issues associated with understaffing.
Q: Where do you see the greatest potential value for predictive models in labor planning?
Jennifer Weigold, BSN. Associate Vice President of System Clinical Programs for Adventist HealthCare (Gaithersburg, Md.): The greatest value of predictive models in labor planning for a healthcare system lies in their ability to anticipate staffing needs before shortages or surpluses occur, enabling leaders to allocate labor proactively rather than reactively. By forecasting patient volume, acuity, seasonal trends and operational demand, these models help organizations reduce costs tied to overtime, premium pay and reliance on travel or agency staff, which are often the most volatile and expensive components of the workforce budget.
Predictive analytics also identify early warning signs of turnover, burnout and vacancy risk, allowing HR and operational teams to intervene before staffing gaps become disruptive. This strengthens workforce stability, lowers recruitment and onboarding costs and supports safer staffing ratios.
BA: The greatest value lies in aligning staffing with patient volume and acuity trends. Being able to anticipate seasonal fluctuations, census changes, and service line growth through predictive models is key to our continued success. Other high-impact areas include overtime and agency cost reduction by forecasting gaps early, retention and turnover modeling to proactively address workforce risks, and budget accuracy through scenario planning that accounts for demand variability.
KM: The greatest value comes from identifying workforce needs early enough to build intentional and sustainable talent pipelines. Predictive insights, even in their early stages, help us understand where nationally based talent may not keep pace with demand and where we may need to strengthen academic partnerships, workforce development programs, or recruitment pathways. This allows us to plan thoughtfully and reduce the need for reactive or higher-cost labor solutions.
Q: What are the main barriers preventing broader adoption of predictive labor forecasting today?
BA: Key barriers include data quality and integration challenges because workforce, clinical and financial data reside in disparate systems; return-on-investment buy-in from leadership; change management and trust to gain manager confidence in the model outputs; and skill gap analysis for advanced analytics capabilities across HR, finance and operations.
KM: Some of the biggest barriers across the industry include:
- Fragmented data environments that make it difficult to connect long-term projections with real-time operational needs.
- Wide variation across markets, requiring adaptable and locally sensitive forecasting models.
- Economic and regulatory factors that influence talent availability and can shift quickly, reinforcing the need for forecasting approaches rooted in flexibility and scenario planning.
JW: Many health systems struggle with fragmented, incomplete or inconsistent data across multiple platforms, including HR, workforce management, electronic health records and financial systems, making it difficult to produce forecasts that stakeholders trust. Even when data exists, integration challenges prevent real-time insights from being operationalized at the unit or service line level. Many organizations also lack internal analytics expertise to build, validate and maintain predictive models, often relying on external consultants or pilot projects that do not scale.
Cultural resistance further slows adoption, as front-line leaders and clinicians frequently rely on intuition for staffing decisions and may fear that predictive tools could expose inefficiencies or threaten job security. Ownership of workforce forecasting is often unclear, spread across HR, nursing, operations and finance, resulting in diffused accountability and inconsistent implementation.
Financial and IT constraints add another layer of complexity, since investment in predictive platforms competes with other priorities such as EHR upgrades, cybersecurity and compliance initiatives. Concerns about forecast accuracy, coupled with variations in staffing standards, productivity measures and union agreements across facilities, also hinder confidence in model outputs.
