Average Change
This method of forecasting considers both negative and positive month-to-month changes. The first step entails the computation of monthly changes in the number of clinic visits, followed by the calculation of the average change in these figures. The next step involves the determination of the midpoint of data using this formula: (n + 1)/2, where n is the number of data points (Lewis, McGrath, & Seidel, 2011). Ultimately, the calculation of the monthly mean of data provides the basis for forecasting clinic visits. Therefore, the following equation allows the determination of forecasting based on the average change:
- November Forecast = Monthly Mean + (Midpoint ⨉ Average change).
Confidence Interval
The confidence interval is a forecasting method that provides a range of values based on varied alpha levels of standard deviation, such as 90%, 95%, and 99%. Lawrence and Klimberg (2017) explain that the accuracy of forecasting has an inverse relationship with the confidence interval. This forecasting method uses mean value, standard deviation, and confidence interval. In the process of forecasting using this method, the first step involves the determination of the mean number of clinic visits.
The second step entails the calculation of the standard deviation of the number of clinic visits. Next, the evaluation of the confidence interval associated with the standard deviation, for example, 95% and 1.96 Z score (Lewis et al., 2011). Finally, the calculation of the lower and upper levels of confidence interval using the mean, the standard deviation, and Z score allows the forecasting of clinic visits in November.
- November Forecast = Mean ± (Z score * Standard Deviation).
Average Percent Change
Forecasting, in this method, entails the conversion of the month-to-month change into the percent change using the previous data point as the baseline. The forecast commences by the calculation of the monthly changes and then the determination of the average change. The next step includes the calculation of the percent change using the change in the number of visits in a given month dividing by the number of visits in the previous month and multiplying by 100% (Lewis et al., 2011). Successively, the assessment of the average percent change provides important information for forecasting. Like the use of average change in forecasting, the average percent change applies the following formula:
- November Forecast = October Visits + (Average Percent Change ⨉ October Visits).
Moving Average
Moving average is a forecasting method that flattens variations in values due to seasonality and gives a more accurate prediction. Depending on the number of data points, the length of the moving average can vary between three months to even a year. The first step of forecasting is the identification of an appropriate length of moving average in which six months seems reasonable. The next step involves the calculation of the average of the number of clinic visits every six months until November 2008. Hence, the following formula applies to the six-month moving average
- November Forecast = (May + Jun + July + Aug + Sept + Oct)/6.
Exponential Smoothing
This forecasting method assigns decreasing weights to data points and provides a smooth trend of projection. The application of this forecast method needs the selection of the smoothing factor, which varies from zero to one (Lawrence & Klimberg, 2017). Hence, the forecast method will use a smoothing factor of 0.2. The next procedure is the application of the formula in forecasting, as shown below.
- November Forecast = 0.3*October Visits + 0.7*September Forecast.
Health Services Organizations
Health services organizations in the United States, such as John Hopkins Cleveland Clinic, Mayo Clinic, and Massachusetts General Hospital, employ various forecasting methods in estimating the flow of patients, admissions, drug utilization, expenditure, and profits. Forecast methods enable healthcare organizations to plan and provide timely services to patients and improve efficiency in their operations (Lawrence & Klimberg, 2017). These organizations hire experts who assess their records and provide accurate predictions using appropriate methods of forecasting. Moreover, due to the shortage of healthcare providers, mainly nurses, healthcare organizations forecast trends of shortage so that they can make suitable adjustments in their human resource management.
Forecasts for November 2008
Average Change Forecast
November Forecast = 29.46 + (23.5 ⨉ 0.4) = 29.46 + 9.4 = 38.86 = 39 visits.
Confidence Interval Forecast
November Forecast = 29.46 ± (10.01*1.96) = 29.46±19.62 (9.84 – 49.08), which is between 10 and 49 visits.
Average Percent Change
November Forecast = 39 + (5.31% ⨉ 39) = 39 + 2.07 = 41.07 = 41 visits.
Moving Average
November Forecast = (38 + 19 + 28 + 29 + 37 + 39)/6 = 31.67 = 32 visits.
Exponential Smoothing
November Forecast = 0.3*39 + 0.7*32.52 = 11.7 + 22.76 = 34.47 = 35 visits.
The Best Forecast Method
Comparison of the methods used in forecasting the number of clinic visits shows that exponential smoothing is the best forecast method. The assessment of the trend of data shows that it has significant variations. Thus, the smoothing factor considers variations and provides an accurate prediction of the number of clinic visits. Principally, exponential smoothing does not overestimate forecast because it is a lagging indicator based on a weighted moving average (Lewis et al., 2011).
In this case study, it is evident that exponential smoothing predicts that the number of clinic visits is 35, which is slightly more than the average number of visits (30) and the forecast of the six-month moving average (32), but less than average forecast (39) and the average percent change (41).
References
Lawrence, K. D., & Klimberg, R. K. (2017). Advances in business and management forecasting. New York, NY: Emerald Group Publishing.
Lewis, J. B., McGrath, R. J., & Seidel, L. F. (2011). Essentials of applied quantitative methods for health services managers. Sudbury, MA: Jones and Bartlett Publishers.