Workforce Metrics Forecasting & Predictive Modelling – Sickness Absence and Workforce Planning
Summary
Northumbria Healthcare NHS Foundation Trust introduced workforce forecasting to strengthen proactive planning, improve operational visibility, and support more informed workforce decision-making across the organisation. Initially developed to forecast flu-related sickness absence during winter pressures, the approach combined workforce, vaccination, performance, and patient activity data to provide forward-looking insight for key stakeholders.
Following its success, forecasting capability expanded across Trust and Business Unit workforce reporting, including sickness absence, headcount, FTE analysis, and student nursing projections. Eighteen months on, forecasting is now fully embedded within workforce reporting and planning processes, helping teams anticipate workforce pressures earlier, improve resource planning, and strengthen overall workforce intelligence across the Trust.
The Challenge
The primary challenge was the absence of future-dated workforce data, which limited strategic planning to historic and current metrics. This restricted the Trust’s ability to anticipate and respond proactively to workforce pressures, resulting in reactive planning approaches across Business Units.
To address this, a forecasting model was developed using 10 years of sickness absence data, enabling the prediction of workforce trends and supporting more informed planning decisions. This allowed Business Units, including Medicine and Surgery, to incorporate forecasted sickness absence and FTE changes into operational and performance planning, strengthening safe and effective patient care delivery.
Historically, workforce planning relied heavily on retrospective data and professional judgement, particularly for headcount and role-specific requirements. The introduction of forecasting provided structured projections across key workforce variables, including expected ranges with a 95% confidence level, significantly improving planning accuracy and consistency.
Given the risks associated with forecasting accuracy, extensive validation, testing, and ongoing variance monitoring were undertaken prior to deployment. This ensured reliability and continuous improvement of the model. Overall, the work addressed a critical gap in workforce intelligence by enabling forward-looking, data-driven planning and improving decision-making for staffing supply and demand.
The Solution
The solution involved developing a forecasting model in Microsoft Excel, using existing sickness absence data to validate accuracy and ensure reliability. A collaborative approach was established between Workforce Reporting and Information Services colleagues to determine the most appropriate method for producing forecasted workforce data.
Although automated solutions using SQL and Python were considered, analysis demonstrated that Microsoft Excel’s Forecast Sheet tool produced the most accurate and consistent results. Over time, the model has proven highly reliable, with sickness absence rates forecasting within a very small margin of actual outcomes.
The approach was then expanded to provide more granular insights, including forecasting by sickness reason (such as mental health absence), at Business Unit level, and across additional workforce metrics including turnover and broader workforce planning indicators. This allowed for deeper and more actionable workforce intelligence.
Initially led by two members of the Reporting Team, the workstream has now transitioned into business-as-usual within Workforce Reporting and is embedded within the People Services data strategy, forming a core component of ongoing workforce analytics and planning capability.
Results & Next Steps
The introduction of sickness absence forecasting has supported more proactive workforce planning during a period in which overall sickness absence has significantly reduced across Northumbria. The organisation has recorded a year-on-year decrease, achieving its lowest levels since 2019 and performing well below the regional average. This improvement has contributed to increased productivity across Business Units and delivered substantial cost savings in sickness absence, amounting to hundreds of thousands of pounds.
Forecasting has also enabled more proactive workforce and service planning. It supported the flu vaccination workstream by informing a more targeted and strategic rollout, resulting in lower-than-forecast flu-related absence during the winter period. In addition, nursing workforce projections have provided valuable insight into future supply over a three-year horizon, particularly in relation to student nurse intake and pipeline planning.
The model has further strengthened workforce planning by identifying trends in headcount, FTE, age profiles, and staff group movements, enabling more accurate recruitment and resourcing decisions. Overall, the forecasting approach has improved strategic visibility and supported safer, more efficient service delivery.
The approach has been widely adopted across Business Units, embedding a forward-looking culture in operational planning. There are also plans to extend forecasting to Occupational Health and Staff Psychology referrals to further enhance proactive service planning and demand management.
