Case Write-Up #1: Tahoe Healthcare Systems

As you read in the case, Tahoe Healthcare Systems is very concerned with the readmission level of patients who have been discharged from their hospitals. Not only does hospital readmission indicate potential quality issues with the care being provided, but it may also impact their financial performance due to penalties assessed on Medicare reimbursements to hospitals with high readmission rates.

In response to these concerns, THS has developed the CareTracker program to provide greater education to patients prior to discharge and follow-up with patients after discharge. Results of a pilot test of the CareTracker program suggest that it reduced readmission rates by 40%, but it comes at a significant cost of $1,200 per patient.

The spreadsheet Tahoe Healthcare Data.xlsx contains three tabs: Training Data, Validation Data, and Test Data. The Training Data contains information (as described in Table 1 of the case) on 2,382 patients including whether they were readmitted within 30 days of discharge this data can be used to develop various models that may help you decide which patients should be enrolled in the CareTracker.program. The Validation Data includes similar information for an additional 1,000 patients this data should be used to estimate the impact of any models you develop from the Training Data. The Test Data contains information on an additional 1,000 patients, but does not disclose whether they were readmitted within 30 days you should apply your final decision model to these patients and indicate (in the column labeled CareTracker?) whether you are choosing to enroll that patient in CareTracker (1) or not enrol that patient (0). Save the spreadsheet with your indicated decisions, and submit this spreadsheet in addition to the memo described below.

Please form a team of 3 or 4 students to prepare a write-up for this case assignment. Your write-up should be in the form of a memo addressed to the CEO of THS that describes your recommendations regarding implementation of CareTracker. In this memo, please be sure to address (with economic/financial justification) the following:

• In the absence of any modelling capabilities, are you better off (base this on the 1,000 patients in the Validation Data) rolling out CareTracker to all patients or no patients?
• What is the Expected Value of Perfect Information? (again base this on the 1,000 patients in the Validation Data)
• How did you approach using the data to come up with a model to help determine which patients should be enrolled in CareTracker? What did you attempt and what did you learn during the modelling process?
• Summarize the final plan you recommend to determine which patients to enroll in CareTracker, including an estimate (based on the 1,000 patients in the Test Data) of the percent of patients you would enroll in CareTracker.