Helmut Herwartz, Nadja Klein, and Christoph Strumann, "Modelling
Hospital Admission and Length of Stay by Means of Generalised Count
Data Models", Journal of Applied Econometrics, Vol. 31, No. 6, 2016,
pp. 1159-1182.

The data used in the article are drawn from the Heritage Health prize
data set (https://www.heritagehealthprize.com), an online competition
organised by 

  Heritage Provider Network
  8510 Balboa Boulevard Suite 285
  Northridge, CA 91325
  inquiries@heritagemed.com

The data set of the competition includes information on claims of
113,000 patients and the number of days in hospital for three
subsequent years. The data and definitions of variables are available
from the competition homepage for registered users under
https://www.heritagehealthprize.com/c/hhp/data.

The variables used in our study are obtained by aggregating the claim
level data to the patient level and are described in the file
variable_def.txt, which is contained in the zip file hks-files.zip.

The file hks-files.zip also contains a folder data and several MATLAB 
programs and subprograms (m-files, which can be opened with any text 
editor) to generate the dataset. To obtain the dataset you need to

1) stack the 7 original csv data files into the folder data

2) set the MATLAB-Working-Path to the folder with the folder
containing the m-files 

3) run the script makedata

Please address any questions to:

Christoph Strumann
Institute for Innovation Research
Christian-Albrechts-University Kiel
Olshausenstrasse 40
24098 Kiel
cstrumann [AT] bwl.uni-kiel.de