This article outlines these trends and implications and paradigms for how electronic health records (EHRs) and health IT (HIT) can be applied to the delivery system transformation. Specifically, we'll look at HIT's impact on clinical analytics within accountable care organizations (ACOs), patient centered medical homes (PCMHs) and integrated delivery systems (IDSs).

We will also discuss frameworks for the application of HIT to quality improvement, cost and outcomes management, as well as predictive modeling (PM)/risk adjustment. The premise of this article is that HIT used in this way represents the "virtual glue" that will hold together these delivery systems.

The New HIT-Enabled HealthCare System

The EHR will emerge as the core of a new HIT-enabled integrated healthcare system. While the existing EHR connects the provider with the patient, the personal health record (PHR) connects the patient back to the EHR through a web portal. But a HIT-enabled IDS requires more: an interoperable, community wide EHR backed by clinical decision support (CDS) systems, computerized physician order entry (CPOE) and a PHR populated with home-based biometrics.

As government mandates prevail, IDSs will shift focus to populations as well as individual patients, necessitating expansion of channels for information exchange between providers, public health and vital records, and private and public payers.

A pictorial representation of the system-wide linkages with an HIT supported delivery system is shown in Fig. 1.

New Ways to Measure System Performance

Performance measures derived from HIT will have multiple applications: quality improvement for health delivery organizations, including real-time safety, care management and retrospective evaluation; payer-sponsored pay-for-performance programs; and information reporting via public health agencies. Today we use claims, surveys and review of paper medical records to measure an organization's performance. In the future, most quality measurement within an IDS will be derived from electronic sources, called electronic quality measures or "e-QMs."

Click to view larger graphic.

Paradigm for the five categories of HIT-based e-QM measures:

• Translated: (Level 1) Traditional (e.g., paper records and claims) measures translated for use on HIT platforms such as the EHR version of existing HEDIS measures.

• HIT-Facilitated: (Level 2) Measures that while not conceptually limited to HIT would not otherwise be feasible, such as the percentage of a population that attains a blood pressure below a certain threshold.

• HIT-Enabled: (Level 3) Measures that generally wouldn't be possible outside the EHR context, such as the percentage of primary care physicians (PCPs) who read key sections of specialists' referral notes.

• HIT System Management/Continuous Quality Improvement (CQI): Measures needed to implement, manage and evaluate HIT systems, such as attainment of EHR interoperability targets.

• E-Iatrogenesis/HIT Safety: Iatrogenesis is a term describing healthcare-induced harm, a new type of harm caused at least in part by sub-optimal application of HIT systems. An example of this measure would be the percentage of e-prescriptions that result in the wrong drug.

For ACOs and other IDSs to effectively use HIT as a source of e-QMs, they may need support from IT, quality measurement experts and strong provider education programs. As these new electronic systems of performance measurement are put in place, key priorities will include:

• Make quality measurement part of workflow and HIT system design.

• Learn how new data sources affect reliability and validity.

• Make HIT systems management indicators a central component of the HIT CQI process.

• Look for population and person-based measures within EHR/HIT systems.

• Figure out the management of e-iatrogenesis.

HIT: Supporting ACO & PCMH Transformation

While PCMHs are likely to be the first to implement HIT-based e-QMs, ACOs will also reap the rewards.

ACO infrastructure must include accountability for cost, quality and capacity, shared savings with the payer, and performance measurement. At the core of accountable care is the PCMH, made possible by the EHR, HIT supported patient care plans and care management.

A transformed healthcare system will require HIT support and paradigm-level shifts on multiple dimensions. Over the next few years ACOs, PCMHs and other IDSs will need to support the following transitions:

Predictive modeling (PM) and risk adjustment will be essential. ACOs and IDSs will apply HIT data for multiple functions, including:

• Financing, payment and planning (e.g., morbidity-adjusted capitation, actuarial rate and premium setting, allocation of budgets and service targets);

• Provider performance assessment (profiling and pay-for-performance); and

• Care management quality improvement, monitoring, research and evaluation.

For most of these tasks, they must account for the varying morbidity burden ("risk") across various groupings of patients/members using sophisticated risk adjustment/PM tools. As ACOs and IDSs are called on to improve care and bend the cost curve, they will be especially needed to use PM tools to identify, stratify and respond to the medical needs of the populations they care for.

HIT-Supported Care Management Along the Risk Continuum

Click to view larger graphic.

Using data in their HIT systems, ACOs will be called on to apply PM tools to stratify their members to support care management initiatives across the risk pyramid. Once the risk level of all members is identified and documented, appropriate levels of care management can be planned (see Fig. 2).

In addition to using computerized data to stratify members, HIT can also be used to help support care management, most of which will focus on high impact chronic disease and multiple chronic conditions. Fig. 2 illustrates digitally supported approaches at each level of member risk.

An example of this process at an advanced IDS was the application of the ACG PM methodology developed at Johns Hopkins, by the Minnetonka, MN-based Medical Health Plans, which launched an HIT-supported health coaching program to help members improve self-management and achieve healthcare goals. The program has reduced healthcare expenditures by $139 per member per month while slashing inpatient utilization and emergency department use.

As ACOs, PCMHs and other IDSs apply HIT for the purposes discussed here, they will face multiple challenges. Challenges will include the need to integrate advanced analytics with changing clinical and fiscal operations, the growing requirements for transparency and accountability, achievement of HIT interoperability and standardization, weak structural cost containment present within the ACA, and the rollout of new funding mechanisms, such as bundling and pay-for-performance.

For the first time in history, HIT will be widespread and sophisticated enough to provide the virtual glue needed to support the transition to high-performance, accountable care. The challenges may be numerous, but so are the opportunities as we move toward a more effective and efficient care system that will cover more Americans than ever before.