Healthcare Analytics Adoption Model: A Framework and Roadmap (white paper)

a foundational understanding of analytic technology and organizational use of analytics in step-wise fashion before attempting the more complicated topics of the upper levels. Also, each level of adoption includes progressive expansion of analytic capability in four critical dimensions:

(1) New Data Sources: Data content expands as new sources of data are added to the healthcare ecosystem.
(2) Complexity: Analytic algorithms and data binding become progressively more complex.
(3) Data Literacy: Organizational data literacy increases among employees, leading to an increasing ability to exploit data as an asset to organizational success, including new business and economic models.
(4) Data Timeliness: Timeliness of data content increases (that is, data latency decreases) which leads to a reduction in decision cycles and mean time to improvement.


Data binding grows in complexity with each level

In addition to these trends within the model, organizations frequently operate at various stages of maturity in each level. In that regard, the model is not necessarily linear in its progression, although in an ideal state that would be the case.

Organizations may find themselves operating quite effectively in Level 5 or 6 but ineffectively at Levels 3 and 4. Such was the case at Intermountain Healthcare during the early stages of their EDW development. Consequently, Intermountain adjusted its strategy and reassigned resources to address the laborious inefficiency of report production in Levels 3 and 4. Afterwards, the gains in efficiencies paid dividends in Levels 5 and above, where data architects and analysts were able to spend more time on market-differentiating analytics. Intermountain Healthcare was named the top Integrated Delivery Network in the U.S. market for seven of the eight years following this adjustment.

Level 0 Fragmented Point Solutions

Inefficient, inconsistent versions of the truth. Cumbersome internal and external reporting

Level 0 of the Analytics Adoption Model is characterized by fragmented “point solutions” which have very focused, limited analytics capabilities, typically focused on departmental analytics such as finance, acute care nursing, pharmacy, laboratory or physician productivity. New knowledge generated by these solutions tends to be isolated to one area, which may encourage optimized sub-processes at the expense of enterprise-wide processes. The fragmented applications are neither co-located in a data warehouse nor otherwise architecturally integrated with one another. Overlapping data content leads to multiple versions of the truth. Reports tend to be labor intensive and inconsistent. There is no formal data governance function tasked with maximizing the quality and value of data in the organization.

Point solutions in this level can satisfy the internal and external reporting that is important to Levels 3 and 4, but they are not a market differentiator and cannot scale to the more complicated analytic use cases and business models associated with the upper levels of adoption. Cumulatively, fragmented point solutions at this level also tend to require significantly more labor from data analysts and systems administrators to use and maintain than single, integrated data warehouses. The same inefficiencies of decentralization hold true for the fragmented costs of software licensing and vendor contract management.

Level 1 Enterprise Data Warehouse 

Collecting and integrating the core data content.

Level 1 is satisfied when core transaction source system data is integrated into an Enterprise Data Warehouse. At a minimum, the following data sources are co-located in a single local or hosted data warehouse: (1) HIMSS EMR Stage 3 clinical data, (2) financial data (particularly costing data), (3) materials and supplies data, and (4)  patient experience data. If available, data content should also include insurance claims. A searchable metadata repository is available across the enterprise. The metadata repository provides natural language descriptions of the EDW content, describes known data quality issues and records data lineage. The metadata repository is the single most important tool for the complete democratization of data across the enterprise. The EDW data content is updated within one month of changes in the source systems.

The beginnings of an enterprise data governance function are established with an initial focus upon reducing organizational and cultural barriers to data access, increasing data quality in the source systems and master data identification and management. Data stewardship for the source data content areas in the EDW is forming under clinical and administrative ownership. Organizationally, it is best for the EDW to report to the CIO at this stage, assuming that the CIO can facilitate access to and the extraction of data from the source systems. Later, as the EDW evolves from the construction and early phases of adoption, the organizational alignment can change to another C-level executive who represents the functional use of analytics in the organization, such as the Chief Medical Officer or Chief Quality Officer.

Level 2 Standardized Vocabulary Patient Registries

Relating and organizing the core data content.

At this level, master vocabularies and reference data are defined and available in the EDW. These vocabularies and reference data include local master patient identity, physician identity, procedure codes, diagnosis codes, facility codes, department codes and others. Data stewardship for master data is functioning. Master vocabularies and reference data are identified and standardized across disparate source system data content in the EDW. Naming, definition and data types in the EDW data content areas are standardized according to local master reference data, enabling queries across the disparate source content areas. Patient registries based on billing codes and defined by multidisciplinary teams are available in the EDW to support basic analytics for the most prevalent and costly chronic diseases and acute care procedures in the local environment. Data governance forms around the definition and evolution of patient registries and master data management.

Level 3 Automated Internal Reporting

Efficient, consistent production of reports and widespread availability in the organization.

Level 3 is characterized by automated internal reporting where the analytic motive is focused on consistent, efficient production of reports required for: (1) executive and board level management and operation of the healthcare organization, and (2) self- service analytics for key performance indicators and interactive dashboards at the director and management level. The key criteria for success in this level is efficiency and consistency of reports that are necessary for effective management, but alone are not enough to create differentiating value in the market. Ideally, once developed and deployed, the maintenance of these reports requires little or no labor to support and are nearly entirely self-service. Also, the reports are reliable in their availability when needed, consistent and accurate, thus minimizing wasteful debate and the attractiveness of developing redundant reports that end users and analysts consider more reliable, consistent or accurate.

An analytic services user group exists that facilitates collaboration between corporate and business unit data analysts. Among other synergies, this group is organized to define consistent data definitions and calculation standards. Data governance expands to include data quality assurance and data literacy training and to guide the strategy to acquire mission-critical data elements in subsequent levels of adoption.

Level 4 External Reporting

Efficient, consistent production of reports and adaptability to changing requirements.

The analytic motive at this level is focused on consistent, efficient and agile production of reports required for external needs, such as: (1) regulatory, accreditation, compliance and other external bodies (e.g. tumor and communicable disease registries); (2) funding and payer requirements (e.g. commercial financial incentives and federal Meaningful Use payments); and (3) specialty society databases (e.g. national cardiovascular data registry). Master data management at this level requires data content in the EDW that has been conformed to current versions of industry-standard vocabularies such ICD, CPT, SNOMED, RxNorm, LOINC and others. ln addition to the low-labor, low-maintenance requirement for producing reliable, accurate and consistent reports at this level, the EDW must be engineered for agility in this context, due to the constantly changing nature of external reporting requirements.

Data governance and stewardship is centralized for external reporting. Stewardship processes exist to maintain compliance with external reporting requirements and govern the process for approving and releasing the organization’s data to external bodies.

EDW data content at this level has been expanded to include text data from patient- record clinical notes and reports. EDW-based text query tools are available to support simple keyword searches within and across patient records.

Level 5 Waste & Care Variability Reduction

Reducing variability in care processes. Focusing on internal optimization and waste reduction.

At Level 5, organizations are moving away from utilitarian internal and external reporting. They have a significant opportunity to differentiate themselves in the market based on quality and cost and enabled by analytics. Data at this level is used explicitly to inform healthcare strategy and policy formulation. The analytic motive is focused on measuring adherence to clinical best practices, minimizing waste and reducing variability, using variability as an inverse proxy for quality. Data governance expands to support multidisciplinary care management teams that are focused on improving the health of patient populations. Population-based analytics are used to suggest improvements to individual patient care. Permanent multidisciplinary teams are in place to continuously monitor opportunities that will improve quality and reduce risk and cost across acute care processes, chronic diseases, patient safety scenarios and internal workflows.

The precision of registries is improved by including data from lab, pharmacy and clinical observations in the definition of the patient cohorts. The EDW content is organized into evidence-based, standardized data marts that combine clinical and cost data associated with patient registries. The data content expands to include insurance claims (if not already included) and HIE data feeds. On average, the EDW is updated within one week of source system changes.

Level 6 Population Health Management Suggestive Analytics™

Tailoring patient care based upon population metrics. Fee-for-quality includes bundled per case payment.

Level 6 is characterized by organizations that have achieved a sustainable data driven culture and established a firm analytic environment for understanding clinical outcomes. The “accountable care organization” shares in the financial risk and reward that is tied to clinical outcomes. At least 50 percent of acute care cases are managed under bundled payments. Analytics are available at the point of care to support the Triple Aim of maximizing the quality of individual patient care, population management and the economics of care. EDW data content expands to include bedside devices, home monitoring data, external pharmacy data and detailed activity- based costing.

Data governance plays a major role in the accuracy of metrics supporting quality- based compensation plans for clinicians and executives. On average, the EDW is updated within one day of source-system changes. The EDW reports organizationally to a C-level executive who is accountable for balancing cost of care and quality of care.

Level 7 Clinical Risk Intervention & Predictive Analytics

Organizational processes for intervention are supported with predictive risk models. Fee- for-quality includes fixed per capita payment.

Level 7 organizations are able to move into the arena of predictive analytics by expanding on their optimization of the cost per capita populations and capitated payments. Their focus expands from the management of cases to collaboration with clinician and payer partners to manage episodes of care, including predictive modeling, forecasting and risk stratification.

The analytic motive at this level expands to address diagnosis-based, fixed-fee- per-capita reimbursement models. Focus expands from management of cases to collaboration with clinician and payer partners to manage episodes of care using predictive modeling, forecasting and risk stratification to support outreach, triage, escalation and referrals. Physicians, hospitals, employers, payers and members/ patients collaborate to share risk and reward (e.g., financial reward to patients for healthy behavior).

Patients who are unable or unwilling to participate in care protocols due to constraints such as cognitive disability, economic inability, geographic limitations to care access, religious restrictions and voluntary non-participation are lagged in registries. Data content expands to include home monitoring data, long-term care facility data and protocol-specific patient reported outcomes. On average, the EDW is updated in one hour or less of source system changes.

Level 8 Personalized Medicine & Prescriptive Analytics

Tailoring patient care based on population outcomes and genetic data. Fee-for-quality rewards health maintenance.

At Level 8, the analytic motive expands to wellness management, physical and behavioral-functional health and mass customization of precise, patient tailored care. Healthcare-delivery organizations are transformed into health-optimization organizations under direct contracts with patients and employers. Fixed-fee, per capita payment from patients and employers for health optimization is preferred over reimbursement for treatment and care delivery. Analytics expands to include natural language processing (NLP) of text, prescriptive analytics and interventional decision support. Prescriptive analytics are available at the point of care to improve patient specific outcomes based upon population outcomes.3 Data content expands to include 7×24 biometrics data, genomic data and familial data. The EDW is updated within minutes of changes in the source systems.

At this level, healthcare organizations are completely engaged as a data-driven culture and shift from a fixation with care delivery to an obsession with risk intervention,  health  improvement  and  preventive  medicine. New data content in the enterprise data warehouse is combined  with  not-yet-discovered  algorithms  that  can identify relationships between genomics,  family  history  and  patient  environment. Eric Topol’s book, The Creative Destruction of Medicine: How the Digital Revolution Will Create Better Health Care,