"SummaryData.xlsx" includes the DRG-level summary statistics from the SID data and Community and Academic Hospitals


Section 7.2. Model fitting and validation:
"Coxian_Parameter_Estimation.m" is used to estimate the parameters of a phase-type distribution. 
The input .csv file is "Random_Input_Data_For_Parameter_Estimation.csv" which shows the structure 
of the data for a hypothetical DRG with random length of stay, number of procdures, and procedure dates.
"Likelihood_one_phase.m" and "Coxian_Pdf.m" are two MATLAB functions that are used in "Coxian_Parameter_Estimation.m".

Section 7.3. Mixing it up: optimal hospitalist case-mix:
"SIDDataAnalysis.nb" is the Mathematica code to generate Figure 6 in page 26.

Section 8.1. Case 1: Community Hospital:
"TuningNonHospitalistVisitRateCommunityHospital.nb" is the Mathematica code to tune the non-hospitalist visit rate
"Community_Hospital.nb" is the Mathematica code to solve the case study (Table 6)

Section 8.2. Case 2: Academic Hospital:
"EstimatingServiceRatesInAcademicHospital.nb" is the Mathematica code to estimate service rates
"Academic_Hospital.nb" is the Mathematica code to solve the case study (Table 8)



