Forecasting
Part 4 · Chapter 2

Forecasting

The forecasting process, barriers to effective forecasting, and strategies for data-driven workforce management.

9 min read

Forecasting

Forecasting is key to workforce management, yet many organizations still rely on outdated models like hours per patient day (HPPD), which ignore indirect costs (sickness, non-clinical work) and often underweight patient acuity.

Over 50 years of research has produced forecasting solutions — optimization methods, predictive models, algorithms, and decision-support systems — that deliver significant cost savings and quality improvements when put into practice.

Forecasting Process

Forecasting is often categorized by one of three phases:

Budget and Planning Phase

This is often how staffing standards are set for an inpatient unit. The monthly average midnight census is forecasted annually by the finance team. This gives an inpatient unit their HPPD — calculated by dividing the total number of nursing hours by the total number of patients forecasted in a 24-hour period.

Scheduling Phase

Often described as the "nurse scheduling problem" — a complex combinational optimization problem that can be modeled mathematically. Various forecasting approaches include simple averaging, exponential smoothing, regression, and ARIMA models.

Staffing Phase

The key is to use real-time census data feeds compared to the number of scheduled nurses to ensure the right number of staff are available at the right time. Technology solutions can support this effort by modeling real-time data to identify staffing gaps.

Forecasting Barriers

Research-Practice Gap

A substantial gap exists between research findings and implementation. Reasons include:

  • Limited collaboration between researchers and practitioners
  • Research too narrowly focused
  • Limited partnership between scheduling vendors and health systems
  • Reluctance to embrace new technology and workflow processes
  • Research published in disciplines other than nursing not being translated

Lack of Technology

Without technology, meaningful workforce data will be minimal. Resistance to changing manual processes and discomfort with systems means nurses miss opportunities to capture meaningful data.

Best Practice
Investment in scheduling and staffing technology systems yields significant return on investment by improving efficiencies for better labor management.

Breaking Barriers

Strategies for breaking through:

  1. Encourage publication of academic research with nurses as co-authors in nurse-management journals
  2. Keep current with research and evidence-based projects; advocate for change
  3. Invest in technology systems for operational efficiency and transparency
  4. Be willing to innovate — challenge the status quo in scheduling and staffing processes
Success Story: Forecasting Cost Savings

A study developed and implemented a nurse scheduling forecasting model for a 33-bed inpatient unit. The predictive model used historical datasets from 2015–2017 to predict daily nurse needs. The forecasting identified weekly variation trends — different from scheduling the same number of nurses every day. Results: schedule accuracy improved by 271% and achieved $12,458 in projected overtime cost savings.

Into the Future

Innovative strategies can tackle age-old workforce challenges. By critically evaluating existing programs and committing to adjusting, redesigning, or building entirely new ones, organizations can break the continuous cycle of symptom management and solve the nurse staffing problem for good.