Looking through the sessions at HIMSS 2011, one can see a wide range of topics related to business intelligence and analytics. From different uses for BI to different types of presentation vehicles to the potential of use predictive analytics, one will not be at a loss for finding something analytics related to sink their teeth into. It finally looks like organizations are truly realizing the power that leveraging their data can bring them.
However, in the rush to pursue analytics, organizations must realize that the value it can bring is directly related to the quality, timeliness, completeness and accuracy of the underlying data. Putting analytics on bad data is like putting lipstick on a pig. No matter how much you dress it up, it’s still a pig. In fact, putting analytics on inaccurate, incomplete or stale data can often be worse than not having analytics at all. Consider the situation in which clinicians are using near real-time analytics to help improve outcomes. Bad data will have a negative effect on the analytics presented, but, more importantly, it can have a negative effect on patient safety.
So does this mean that organizations should be hesitant to pursue analytics? Not at all. The benefits to be derived in a competitive healthcare landscape are far too great not to be using analytics to its fullest potential. What it does mean is that just as much care should be taken to put in place a framework to insure the accuracy and completeness of the data. To this end, organizations should consider the following critical aspects of healthcare business intelligence in addition to the reporting and analytics capabilities.
Governance
This is a term that often receives bad press. Many organizations have tried to implement enterprise governance and failed, while others don’t know how to get started. To understand how critical this is, find someone at HIMSS 2011 who works for an organization that has successfully implemented this. I suspect they will tell you the effort was well worth it. Governance provides the enterprise view necessary to truly implement work class analytics.
Master Data
This is another area in which many organizations have either neglected putting the effort into or have done so with a very narrow focus. Master data is the glue that binds organizational data. From common definition for core functional domains (ex. Patient, Physician, etc.) to enterprise hierarchies (roll-up or aggregations of data) to standard reference data (Ex. diagnosis codes, procedure codes, etc), having enterprise master data drives improved analytics.
Data Integration Framework
Having high-quality, timely and accurate data is the foundation for analytics. Data is being derived from more sources today than ever before. From internal clinical, financial and administrative systems to industry organizations to HL-7 messages to manual data, the integration of data in healthcare can be a daunting task. Many organizations are feeling the effects of building so many point solutions in the past. However, the integration of data does not have to be as complex as some make it out to be. Having a strategy in place that clearly defines an integration roadmap and implementation framework, doing the proper analysis and quality checking and taking an incremental approach to implementation reduces the effort and risk.
Phased Roadmap
In the rush to implement analytics, many organizations bite off more than they should. The big-bang approach doesn’t work. It is a much better approach to create a strategy and roadmap that allows for a phased implementation of analytics based on business priorities. This not only allows for the faster realization of critical analytics capabilities, but also provides the flexibility to react to changing priorities and business conditions.
Analytics are an essential component to any world class healthcare organization. However, care should be taken to ensure the proper foundation is put in place to ensure you realize the full benefits.
