IBM’s Patient Care powers preventative patient care with analytics

By October 25, 2012
doctor gear in a computer screen

IBM’s new Patient Care and Insights software expands the application of electronic health records to preventative care. The actionable information practitioners can receive from harnessing patient data could improve all levels of healthcare.


The role that IT has played in improving patient care has been expanding with the spread of electronic health records and the US government’s incentives to continue the thread. IBM hopes to encourage a transition to predictive healthcare with new software for practitioners and hospitals that leverages data analytics. This analytics software, Patient Care and Insights, creates predictive models of various health conditions, identifies early intervention opportunities, and coordinates patient care. These analytics are based on capabilities and algorithms that examine thousands of characteristics at once - demographic, social, clinical and financial factors combined with physicians’ notes.

Harnessing big data with software

Data analytics are helping IBM power its health care focus on case-based, patient-centric electronic care plans and population analysis. This tool enables much more clinical data to be used effectively than can be properly processed by the human brain. As IBM’s research shows (PDF), acute decisions by clinical phenotype are based on 5 facts, which represent the Human Cognitive Capacity. But other decisions use many more facts - those based on structural genetics, multiple chronic conditions, etc. all exceeding the maximum facts that can be best juggled by practitioners alone. With healthcare’s shift to EHR support, and creating analytics systems to be able to best use this data, the patient-centric model provides several levels of value to the health system as assessed by an econometric model.

Predictive care results in lower costs and increased ROI

By seeing how other patients with similar characteristics fared with specific treatments, as well as pairing patients with optimal physicians based on condition, caregivers tap into “the collective memory” of the care delivery system. Practitioners can process the text-based information that is output by PCI, which enables them to identify at-risk patients and anticipate outcomes and intervention effects. These outputs appear to be optimized for higher-risk categories like residents of Patient Centered Medical Homes. At a hypothetical sample hospital, IBM projected that the econometrics of combined care quality, operational cost and revenue for an 800-bed Academic Medical Center would be $81.3 million.

Legal mentions © L’Atelier BNP Paribas