Hospital Charges Case Study

1419 Words6 Pages
According to Case 42, Hospital Charges, physicians have control over hospital costs, length of stay, and which drugs and other supplies are needed for patient treatment. However, physicians are not employees of the hospital and they usually have their own billing service. Furthermore, hospital’s revenues are determined by the patient’s insurance coverage, which pay according to DRG codes (Diagnosis Related Group). The 289 patient records in the data are for normal deliveries of babies. The variables used are: DAYS (the number of days the patient spent in the hospital), CHARGES (the total expenses charged to that patient), PHYSICIAN (is a code identifying the physician), and PAYOR (indicates the type of insurance the patient carries: 1 for managed…show more content…
The F-model equals 170.48. The higher the F-model, the more significant our data is. The p-value equals 2.73E-63, which means that is below our alpha of .01 showing that there is a high significance between hospital charges and (DAYS, PHYS, and PAYOR). Moving on to our regression output, one could see that our most significant values in relation to CHARGES, is DAYS and PHYS. Our regression equation equals – y=847.4293 +889.7494x1+8.2620x2+04499x3. DAYS – For one day added to patient stay, CHARGES went up $889 dollars. For one change to PHYS code, the charges went up $8.26, and for one change in PAYOR, the charges go up $0.44. In rationalizing this data, I can understand the change in CHARGES due to an added stay in the hospital. The p-value of 3.66E-64 proves that there is strong significance between those two. However, PHYS and PAYOR with p-values are not significant to…show more content…
Looking at the differences between the first regression analysis and this one, I have concluded that our Adjusted R² value increase but not much if anything, even deleting the least significant variable, our correlation relationship is only 64% and weak. The p-value is still low, less than our alpha of .10, and the F-model is still a large number (256.62). The coefficients stayed the same, however, out p-values changed. The p-value for our Intercept changed from 6.25E-09 to 2.03E-09, the p-value for DAYS changed from 3.66E-64 to 3.02E-65, and our p-value for PHYS changed from .4065 to .4044. Even though the p-value for our Intercept and DAYS increased, it didn’t affect the significance of our model. Getting rid of PAYOR variable didn’t have much of a significant change on our PHYS p-value as it did for our Intercept and DAYS. The overall p-value changed from 2.73E-63 to 1.51E-64, however, the number is still low and there is still a low chance of inaccuracies in this model right now. In order to make this model more accurate, I am going to take PHYS out of the

More about Hospital Charges Case Study

Open Document