| 摘要 |
A computer program, PREdictor of Minimum INcome (PREMIN), has been developed for predicting health care facility future minimum income (FMI). PREMIN was developed using "MICROSOFT" "EXCEL" for "WINDOWS" 4.0. PREMIN allows estimation of anticipated number of office visits with three values, L (least possible), M (most probable), and H (most possible), and speculation of proportion of payments sources, i.e. HMO's Medicare, Medicaid, etc. The mathematical and statistical foundation of PREMIN is based on a method that determines the minimum number of independently distributed random variables required in a linear combination in order that its distribution can be approximated by the normal density function. Number of office visits are treated as a random variable and are approximated by the triangular distribution. Proportion of payment data is used to compute coefficients in linear combinations. If normal approximation is not adequate, PREMIN simulates the FMI distribution based on the input data. Therefore, PREMIN output is, in some cases, a plot of the normal probability density function for a FMI distribution and associated cumulative probabilities. PREMIN output may also be a histogram of simulated FMI values and associated summary statistics such as a frequency distribution by count and percentages. PREMIN provides a useful tool to estimate the distribution of FMI. After several demonstrations of the program to health care professionals, it is concluded that current methods of predicting income are quite immature in comparison with probabilistic calculations utilized by PREMIN. Further, the computer system can provide vital and beneficial financial data to health care facilities.
|
