Forecasting Nursing Staffing Requirements by Intensity-of-Care Level

One of the more difficult jobs confronting the hospital administrator is estimating the level of nurse staffing needed to meet patient requirements. The further in advance these estimates can be made, the easier it is to plan levels of staffing, vacation time, and training programs. These factors are vital inputs into the control of nursing salary costs, which usually represent at least 50% of the operating costs of the short-term acute-care hospital. In determining staffing requirements, such factors as total census, intensity-of-care levels, and type of ward must be estimated for appropriate planning to be accomplished. To aid in this process, the authors in this study developed a series of regression models for a short-term 220-bed hospital to determine nursing staff levels. Utilizing data from one year of manually prepared daily reports by shift, ward, and intensity-of-care levels, statistically significant models were developed which provided the following: 1. Using ward, month, day, shift, and time as independent variables, the number of patients in each level of care can be predicted. The number of patients by these levels can then be used to predict nursing man-hour requirements. 2. Or in a similar manner using only time, beds, ward, and month as the independent variables, the model can predict total hospital census for a specific month and by ward. Using this census prediction, the model can compute the levels of care for each ward based on the predictive model above, where time, ward, month, and shift are the independent variables. Using these levels of care and standard nursing hours, the model can determine the staffing requirements for the ward.