Social, mobile, analytics and cloud

By John Halamka
09:19 AM

On Monday at HIMSS, I signed my new book, Life as a Heathcare CIO for 300 folks. During the rest of the day I met with numerous companies, leaders and fellow IT professionals. The theme I heard frequently was the need for care management/population health applications based on data acquisition, normalization, mining and workflow. Common characteristics of such applications included social networking features to gather data from patients/families/providers, a mobile component, a predictive analytics component and cloud hosting.

I had no idea that the major consulting companies and analysts have already coined the SMAC acronym for this nexus of ideas (social, mobile, analytics, cloud)

As I walked the HIMSS floor, some of the care management applications I saw were real, developed in platforms like Salesforce and PegaSystems. Others were “deployed” in Powerpoint, which is a powerful development language used by marketing departments to quickly author software.

As Accountable Care Organizations focus on continuous wellness rather than episodic sickness, the market for new tools will grow exponentially. We have to be careful that social, mobile, analytics, and cloud (SMAC) does not become social, cloud, analytics and mobile (SCAM).

Here are few characteristics to look for in real care management/population health software:

1.  Cohort identification - how can patients be enrolled in disease management and care management programs? A drag and drop interface with concepts such as problems, medications, allergies, labs and demographics should be available to specify cohort selection criteria. Ideally, natural language processing will be available for cohort identification based on free text notes.

2.  Rules authoring - once cohorts are identified, there are likely to be protocols and guidelines that enumerate tasks to be done, gaps in care to be filled, and reminders to be sent to providers, payers and patients. The application should support user definable rules creation.

3.  Workflow - non-physician extenders are likely to use the application to ensure tasks are completed and to monitor patient progress. Dashboards and automated "to do" lists should be available.

4.  Alerts - a change in patient status, based on patient self report or diagnostic data should result in an alert to the care manager, appointment scheduler and other care team members, triggering interventions such as home care visits.

5.  Patient Generated Data - often data about patient health status/outcomes are  best provided by patients and families themselves. Information such as activities of daily living, pain scores, mood and medication compliance are not easily found in provider entered EHR data. Interfaces to home care devices, mobile apps, and patient portals should be part of the care management suite.

Full featured care management software is a foundational strategic requirement for accountable care organizations.

Once we finish Meaningful Use Stage 2, the HIPAA Omnibus Rule, and ICD-10, we’re all likely to turn our attention to care management/population health as part of our Affordable Care Act implementation.