Issue link: https://viewer.e-digitaledition.com/i/155901
speciAl feAture 'Analytical' Care organizations By Jennifer Webster to successfully manage population health, accountable care organiZations (acos) need sound business and clinical intelligence strategies. W HAT FUNCTIONS ARE critical for an ACO? The U.S. Department of Health and Human Services, laying out the shared savings proposal, describes an ACO as more than just a pool of patients or a network of health providers: "The ACO would have an infrastructure, such as information technology, [enabling it] to collect and evaluate data and provide feedback to the ACO providers across the entire organization." Such data would focus on the people enrolled in the ACO: the patients. Their progress could be monitored over time, and groups of patients might be singled out for additional care or best-practice interventions (e.g., follow-up with fall risk patients for a home safety survey or medication reconciliation). Medicare ACOs receiving reimbursement (or penalties) from the Medicare Shared Savings Program have a very specific set of quality measures to meet, some involving documentation of behaviors on the part of physician practices and others 8 | Central New York MD News ■ MDNEwS.COM related to the health status of patients. In other words, at some level, providers, patients and circumstances all factor into an ACO's outcomes. Of the 33 quality indicators measured for Medicare shared savings reimbursements, some require patient efforts (such as non-use of tobacco among patients with diabetes); some require physician action (such as qualifying for electronic health record [EHR] incentives); some involve patient/provider interactions (such as screenings, which require the physician to provide the screening and the patient to show up for it); and many also involve outside circumstances (such as hypertension management, which can be affected by variables such as stress). Rewards for meeting these quality measures can be substantial. But providers need good data — first of all to document they are meeting measures within their control (such as EHR meaningful use), and, more importantly, to understand the needs of their patient base and determine which interventions will best help these patients maintain their health. Ongoing collection and analysis of data will also contribute to sound after-action reviews, helping physicians see which interventional strategies have been most effective and which need improvement. Documents into Data: Semantic Analysis Information may be collected from a variety of sources, including claims, EHRs and paper records as well as unstructured references, such as images and physician narrative. Because unstructured documents are not amenable to analytical processes such as adding and averaging, finding like items, identifying patterns, and predicting trends, these documents need to be transformed into structured data — the kind that can be used to populate a database. Physicians may dread having to pay for hours of data-entry from paper records or outdated EHR to achieve this transformation. However, tools are on the way to make the process less painful. Some EHRs use natural language