A leading healthcare organization operates 16 full fledged, multi-speciality hospitals in many parts of India. The Central office monitors quality of care at all centers, apart from Operational and Financial parameters.

Research into improved treatment modalities is another major focus area for the organization and multiple research programs are run concurrently. The existing set-up tracked Medical and Operational parameters through extensive paper based reporting.

Challenge:
Create a central data warehouse of patient medical records and operations data, manipulate the data (Data Mining) and generate dashboards and detailed reports to meet research and operational requirements.

A comprehensive solution to eliminate the paper based reporting system was required; specifically Patient Medical data was to be complete to enable research oriented reporting.

Important decision influencing criteria while selecting the technology - Cost, Robustness, Ability to handle mixed environments (Linux/ Windows)

Approach Taken:
The process started with the Key Management team identifying the parameters to be tracked and interactions with the enterprise wide HIS to carry out required(minor) customization.

As the organization had planned to make the B.I. system available to a number of users at different levels in the Hospitals and at Head Office, it was very important to be free from ?Per User? licensing costs.

The B.I. solution required the use of multiple products and we had to choose a primary Product Suite. Different product suites were evaluated (8 out of over 20 considered) and finally an open source product suite was selected.

The product suite was found to have a few excellent products and was flexible to allow integration with other products which were identified as superior.

Perceived Benefits:
The paper based reporting system was very time consuming. Staff at every Hospital had to spend many hours weekly, completing the paper reports and they had to be collated again at Head Office.

Accuracy of data was open to question in the old system. A system where report data was ?traceable? to actual patient visit details, made reporting vastly more reliable.
The old system was often a few weeks and sometimes a couple of months late with data. This implied that the data was often obsolete and ?not actionable? by the time it was reviewed. An online system with ?Alerts? meant that Head Office would be able to play a more effective role in providing leadership to the hospitals.

Implementation Challenges:
Overcoming resistance from users who were uncomfortable with the complete transparency that such a system would bring in was something we had spent much time planning for. We went out of the way to create benefits for users at every stage, to make managers at every level, users of the system. This way, there was sharing of information with the creators of the data, and created a sense of joint ownership. This was an involved process and had to be handled sensitively for project success.

When the HIS is being customized, ensuring accuracy of the reports was a challenge as fields mapped to reports could become inactive or be supplemented with other untracked fields. Using a visual report creator was an immense benefit as it allowed the implementation team to sit with users and explain/track exactly where data was obtained.

Things we would do differently: Start the report definition process as the first step of the HIS review. This would help ensure that all parameters we wanted to analyze were effectively tracked through the HIS.

Conduct a process review about 25% into the project. Sometimes we spent a lot of time tracking parameters which were important on paper, but had become redundant when online.

Conclusion
A B.I. implementation can be extremely effective and open source tools provide an eminently functional option