Aggregate planning in health care foodservice systems with varying technologies

The health care delivery system has been criticized widely for ineffective cost control, excessive pricing, and poor productivity. It is clear that a segment of the United States economy has continued to be inflationary the past few years. The health care foodservice system is expected to meet the daily food and nutritional needs of patients, employees, and visitors throughout the year. Some of the difficult-and potentially costly-management decisions involve establishing staffing levels and food production rates to meet this changing demand. The purpose of this investigation was to evaluate two heuristic aggregate planning models in three types of health care food production systems to determine if the models were more effective in cost reduction than the management decisions which were made. The Search Decision Rule (SDR) and Management Coefficients Model (MCM) were selected based upon previous testing in industrial applications. In each of three different hospital food systems-conventional, cook-chill, and minimal cooking-the aggregate plan decision rules (from the models) were compared with the hospital’s master labor schedule over a one year test period. Cost of implementing the decision rules and the existing master labor schedule was determined using an objective cost function derived from cost records in each foodservice system. The SDR model performed statistically better than the MCM and the existing master labor schedule provided by management. SDR model performance depended upon the technology in which it was applied, providing the greatest savings in the conventional foodservice technology. Potential annual dollar savings between the hospital’s existing master labor schedule and the SDR for each foodservice system varied from $74329 to $182487. Implementation of the SDR model could help the foodservice administrator be more effective in scheduling expensive labor resources by determining in advance decision rules that will simultaneously meet production demand and significantly reduce cost. The major resistances to implementation are expected to be lack of model understanding and the resistance to trusting a computerized model where the potential for some food shortage will still exist, even though it exists now with manual scheduling systems.