How hospitals prepare for health IT in an ACO world

A survey by consulting firm Beacon Partners last month found that even though most hospital executives (92 percent) were planning an accountable care organization (ACO), almost half (48 percent) of them didn't know how an ACO would affect their own organization and whether it would improve care.

Amid the uncertainty around ACOs, FierceHealthcare caught up with Jonathan P. Weiner, director of the Johns Hopkins Center for Population Health IT and Johns Hopkins University professor of health policy and management and of health informatics, about IT preparations needed for ACOs. Hear what he has to say about care coordination, risk adjustment, and health information exchanges, and why electronic health records (EHR) are key to it all.

FierceHealthcare: What must healthcare organizations do to prepare for ACOs regarding health IT?

Jonathan Weiner: Accountable care organizations are virtually integrated delivery systems. It will be impossible for them to achieve this integration unless key participants are able to actively share patient information. To that end, the proposed federal ACO regulations strongly encourage the adoption of standardized electronic health records (EHR) and the sharing of these records via community-wide health information exchanges (HIE). These regulations require that at least 50 percent of an ACO's primary care physicians achieve the fed's EHR Meaningful Use criteria by year two of the ACO agreement. So due to both the functional need and regulatory requirement, ACOs will give EHR implementation a big shot in the arm.

FH: Does an ACO need special IT expertise to be successful?

JW: EHR interoperability across the ACO will be key, but the community HIEs will not likely be able to achieve this in the near term. Therefore, the ACO's support infrastructure will need to include a sort of internal HIE for, at least, prime referral partners. Beyond the exchange function, ACOs will need to integrate medical, financial, and administrative data; apply advanced analytics to support the care delivery process; and provide a wide range of business and clinical intelligence needed to effectively coordinate and manage the care of the entire patient "denominator" population.

FH: You work heavily in predictive modeling and risk adjustment as the codeveloper and CEO of the research and development team of the Johns Hopkins ACG Predictive Modeling/Risk Adjustment System. What role does predictive modeling and risk adjustment play in the ACO environment?

JW: ACOs will be responsible for the care of populations and will need tools similar to those used for many years by private and public health plans. Predictive models will be essential as a means of identifying persons at risk for high impact events, such as hospitalization and or doctor shopping. Risk adjustment will be essential for ACOs to ensure that physician pay-for-performance and gains-sharing schemes all fairly consider the case mix of their patients. To be most useful to ACOs, predictive modeling tools will need to be available on close to real-time basis to influence care, which means they ideally should be integrated into EHR systems as automated "population health decision support" systems. The types of risk targets, which currently center on utilization, should be expanded to include a wider range of patient outcomes.

FH: Finally, many healthcare execs are feeling pretty uncertain about health IT even though one of its touted benefits is that it will foster care coordination. What advice do you have to them who may be limited to their existing systems?

JW: It is possible to assess the level of patient care coordination using existing health IT systems, such as claims records. For example, the ACG predictive modeling/clinical analytic system we've developed at Johns Hopkins identifies a series of markers of coordination within physician networks. These include the presence of a majority source of care, the number of separate providers in the mix, and the presence or absence of a PCP (primary care provider). We and others are also working on new approaches to measure achievement of coordination and care handoffs by tracking information flow within EHRs. For example, was appropriate information sent and received via the EHR and did the provider review and act on it on a timely basis? Whatever the source of data, ACOs will need to act on these metrics in real time so that case managers and PCPs can identify and correct instances of poor coordination as they occur.

This interview has been edited and condensed for clarity. Karen M. Cheung contributed to this article.