Budget process evolution and maturity in healthcare organizations traditionally has lagged other industries, where leading-edge and data-driven forecasting approaches are currently evolving. Most healthcare institutions today invest a significant amount of time and resources in an annual process, with weeks spent negotiating a budget between department management and administration. This is true whether the approach is bottom up, zero based, volume driven or top down.
This approach has institutionalized an extended budget process. Six months or more of the year is often invested in budget development. The net effect of this extended process commonly produces results that are out of date as soon as the budget is complete. Projections based on assumptions that are six months old often produce erroneous budget targets, primarily due to differences between volume assumptions and actual volumes and unanticipated changes in how services are provided throughout organizations.
Flex budgeting
Flex budgeting was developed to address the forecasting errors of extended budget cycles. Flexible budgeting assists in variance analyses (a process of breaking down budget to actual variances into their relevant components — volume, rate, and efficiency) by removing volume as a cause of the budget variance. Flexible budgeting takes the actual volumes experienced by organizations and flexes variable budget revenues and expenses — calculating variable budgets based on those actual volumes. Flexed budgets are based on what actually happened in hospital departments, as opposed to the original budgets, which are based on stale volume assumptions.
Continuous or rolling budgets
Another approach to improve forecasting accuracy and annual budget processes is to introduce a rolling budget concept into the annual budget cycle. This approach reduces the effort required for the budget process by continuously forecasting the budget throughout the year. Each month, the budget is extended as the year progresses, always maintaining a 12 to 15-month forward view of the forecast. Rolling budgets start after one month of the new fiscal year, by forecasting a new "12th month" to the remaining 11 months of the current budget. This has the effect of generating a continuous 12-month forecast, forecasting a new forward-looking 12th month each fiscal period. The stated annual budget for any fiscal year is essentially derived from this process by a point in time approval of the most recent forecast.
The rolling budget process addresses the challenges of extended traditional budget development processes, by spreading forecasting tasks throughout the year. Including decisionmakers throughout the year also keeps forecasts visible and reduces annual training investments. The rolling budget method also reduces workload by leveraging workflow repetition and efficiency.
Machine learning and AI in budgeting processes
A new approach to budgeting that leverages sophisticated statistical forecasting techniques and machine learning/artificial intelligence capabilities has been developed over the past several decades in manufacturing and other industries outside of healthcare. Statistical forecasts are fiscal management tools that present projected results based on past, current and projected financial conditions. They are used to help identify future revenue and expenditure trends that may have immediate or long-term influences on institutions’ budgets and strategic plans. The statistical techniques used to forecast vary widely, but include the following:
- Extrapolation using historical data to predict future behavior by projecting, trending, moving averages and applying single exponential smoothing
- Regression/econometrics using a statistical procedure based on the relationship between independent variables and a dependent variable
- Hybrid approaches combining knowledge-based forecasting with statistical techniques
The chosen method for one expense type may differ for another expense type. While complex techniques may produce more accurate answers in certain cases, simpler techniques tend to perform just as well or better on average.
The application of AI and machine learning to advanced statistical forecasting techniques leverage recent technological advances in both the hardware and software arenas. As multiple variables are analyzed for statistical validity and forecasting accuracy, an automated feedback loop into the forecasting methodologies can generate improved accuracy over time. Variables that are most effective and accurate in forecasting volumes, revenues and expenses become refined, and new variables are discovered that provide improved forecasting results.
For variable expenses, forecasting using variable relationships leverages the flexible budget model and allows the forecasting entity to focus the forecasting methodologies on the volume forecasts that drive the variable expenses. For expenses with a direct causal link to the chosen volumes, the resultant expense forecast will be as statistically valid as forecasting the expenses directly from history.
In addition, the application of advanced forecasting techniques to volumes allows the forecasting entity to test the statistical validity of the volume indicators chosen to flex the budget. Additional volume indicators can be tested and improve the volume indicators chosen to drive the variable budget.
The accuracy of the financial forecasting approach can be measured and compared against what the traditional budget process projects. Studies are underway at several acute care healthcare facilities nationwide to test the accuracy of advanced financial forecasting techniques. Preliminary results are encouraging and show that the real-world application of these techniques is at least as accurate, and in many cases more accurate, than the traditional budget approaches.
The future of budgeting in healthcare
Traditional healthcare budgeting process trends and directions suggest an evolution toward a hybrid approach that improves future forecasting accuracy. For some healthcare providers, a combined approach using advanced statistical forecasting techniques in conjunction with a rolling budget process provides the optimum balance between traditional budgeting techniques and future forecasting methodologies. Incorporating advanced statistical forecasting techniques into traditional budgeting processes can follow a traditional maturity model approach:
- Background preparation lays the groundwork for forecasting techniques and provides a foundation for its implementation and success.
- A flexible budget model focuses the forecasting on volumes, driving the variable budget and isolating the fixed expenses that are favorable to advanced statistical forecasting techniques.
- Implementation of a rolling update to the budget, adding continuous forward-looking periods updated each month, provides an ongoing framework for the forecasting exercise for the institution.
This framework institutionalizes the workflow throughout the year and smooths the workload for both forecasters and reviewers of the results. It provides the most fertile environment for the successful implementation of advanced statistical forecasting techniques.
Advanced statistical forecasting techniques within this framework will update the new forward-looking periods with improved accuracy and trending. Each month, an updated forecast can be distributed to department managers containing the updated forecast, providing an opportunity to add additional planning variables and projections relating to new services and care delivery evolution. This provides the appropriate feedback loop and introduction point for strategic planning initiatives that are outside of the purview of the statistical forecast bases. It also creates a junction for inclusion of important planning objectives into the rolling forecast workflow.
To maximize decisionmakers’ interest in the forecast, it will be important to emphasize the importance of the forecast as a key factor in the planning and budgeting process. This means incorporating a long-term perspective for the budget, emphasizing financially sustainable decisions, and including long-range forecasting into the process. A 12-month horizon in isolation limits the purview of the forecast, culling longer-term trends and directions from the results. Only by including a longterm horizon into the planning process can trends be validated and understood, and the near-term plans confirmed to drive the organization toward the longer-term objectives and goals.
A hybrid framework can provide the best mix of tools and techniques for providers. It will leverage the recent advancements in forecasting techniques while maintaining flexibility in adapting the results to the specific organizational needs. Advantages of the hybrid approach can be distilled into the following categories.
Improved forecasting accuracy. The traditional approach to budgeting, whether top-down, bottom-up, zero-based or flexible budget driven, is inadequate to deliver accurate forecasts in most cases. Incorporating advanced statistical forecasting techniques will enrich the budget forecast by improving the accuracy of the results.
Time savings. Transforming the budget process from a painful annual exercise into a streamlined forecasting approach, spread throughout the year utilizing a rolling forecast, provides relief from the annual process. It improves the efficiency of the forecasting process and reduces the annual training expenses associated with the traditional budget approach.
Improved buy-in to projections. Distributing the rolling forecasts to departments, while simultaneously gathering the additions from the long-term or strategic plans, is the best approach to gain department manager confidence and ensure that the plan is adhered to and financial objectives achieved.
The continued budgeting evolution for healthcare organizations can build upon and leverage manufacturing and service industry techniques that incorporate advanced statistical forecasting techniques. As healthcare budgeting tools incorporate advanced statistical techniques, we expect the budgeting process to evolve and address the challenges experienced by today’s healthcare organizations.