ABSTRACT
“Information System Success in Todays’ Health Care:
An Evaluation of the Dependent Variable”
Kirsti Askedal a, Elisabeth H. Rabbersvik b, Lillian N. Q. Solberg c
a Consultant, Sykehuspartner IKT, Oslo Universitetssykehus
b Milieu Therapeutist, Kristiansand kommune
c Consultant, SSHF Klinikk for rehabilitering, Kongsgård


Introduction
Evaluating information systems (IS) in health care has increasingly importance. Statements on how IS will improve quality and effectiveness in patient care, empower patients and help reduce medical errors, can only be validated through evaluation of these systems. DeLone and McLean (1992, 2003) developed a model for assessing IS success. Great developments in IS in health care made us wonder if D&M’s IS Success Model (2003, D&M 2) is current to evaluate IS in today’s health care. To assess success, a definition of success - the dependent variable - is necessary.
Method
To approach the problem “Does D&M IS Success Models categories have validity in today’s health care, and would it be suitable to adjust these categories to health informatics’ conditions?”, five research questions were set. To assess whether the categories defined in D&M 2 are valid to evaluate IS success in today’s health care, we conducted a quantitative content analysis of scientific papers. The papers used are recommendations about criteria for IS success in health care, and papers on evaluations of IS in health care.
Results
Chi square are used for testing differences in significance and the validity of models, between D&M 2’s categories and expert recommendations, and between D&M’s categories and evaluations of success criteria. Absence of agreement led to composition of a Health Information System (HIS) Success Model.
Results showed that D&M 2’s categories were significantly different from experts’ recommendations. The HIS Success Models’ categories was not significantly different from experts’ recommendations. The HIS Success Models’ categories was also more adapted to the categories used in evaluations, than D&M 2’s categories.
Discussion
Our results show that D&M 2’s categories do not accommodate IS in health care. The health care systems are complex organizations, with many different stakeholders, and success in one organization may not be a success in another. At the same time, success is dynamic, and what was considered success in 2003 is not necessarily a success today. Data used in this study are papers published from 2003 – 2010. Possibly, the technological and organizational changes in health care have altered the criteria for evaluating success.
Conclusion
From the basis of the research questions, our conclusion is: The categories in D&M 2’s IS Success Model does not have validity in todays’ health care. It would be suitable to adjust the categories into health informatics’ conditions. On this basis, we have composed a new model: Health Information System (HIS) Success Model, which has greater validity in health informatics’ conditions, than D&M 2s categories.

Acknowledgements
The authors would like to thank our teaching supervisor Rolf E. Nikula, Bernhard Mogens Ege and Olav Birger Holen for mathematical advice and input to our analysis. We also would like to thank Hilde Moen Solberg for reading the proofs of our Master Thesis