Future Directions for Clinical Business Intelligence (Module 6)

  • Expanding Scope of Clinical Data

  • Improvements in BI Technologies

  • Exploiting Really Big Data

  • Personalization of Care

Clinical Analytics is rapidly expanding, and several technologies that are in the hands of innovators and early adopters will become mainstream over the next several years. Some trends, such as expanding settings of care in the data warehouse, can proceed incrementally while building on current capabilities. Other trends, such as the personalization of patient care, will require the adoption of new management philosophies and clinical processes and implementation of new technologies.

 Expanding Scope of Clinical Data

The scope of clinical data that health systems need to acquire, integrate and analyze is rapidly expanding along several dimensions:

  • Settings of care included in enterprise data warehouses will expand to physician services, home care, long-term care, call centers and care management services. Data models will need to grow to accurately represent them. These and other types of health services and systems will need to take advantage of more advanced integration techniques (e.g., web services messaging protocols) as they build interfaces to new source systems.
  • Structured clinical data elements will multiply as electronic medical records (EMRs) are implemented and interfaces to data warehouses are established.  As health systems use EMR data for performance measurement and other analytics, they are recognizing the need to capture more elements in a structured format, rather than free text notes.  The standardization of more detailed clinical measures will be an important focus as they compare performance among providers and facilities. 
  • Patient flow measures will be more widely captured and used in clinical analytics. Information on wait times, turnaround times and bottlenecks in clinical workflows will be integrated into analytics frameworks.  They will improve the patient experience as well as improve the efficiency of clinical processes.
  • Population health measures will expand in provider data warehouses as business models evolve to take on more responsibility for wellness, prevention and chronic disease management.  Accountable Care and other initiatives shift risk from payers to providers.

Back to Top

 

Improvements in BI Technologies

New software technologies and architectures will be needed to help health systems effectively deal with dramatic increases in the scope and scale of clinical data and analytics.

  • Rules engines will provide a robust mechanism for implementing the clinical logic needed for clinical decision support and performance measurement.
  • Terminology services will help health systems manage and translate among diverse medical vocabularies such as SNOMED CT, LOINC, CPT, ICD-9 and ICD-10.
  • Natural language processing will bridge the gap between free text medical narratives and structured data with controlled vocabularies to facilitate analytics.
  • Messaging protocolswill mature and become implemented more widely, facilitating near real-time interfaces from source systems and reducing reliance on conventional periodic batch updates.
  • Measures catalogs are envisioned as centralized, shared services that fully define all details of performance measures (numerators, denominators, etc.) so they can be consistently implemented in different systems and environments.

Standards such as HL7’s Health Quality Measure Format should help advance the industry towards enterprise services that encompass all measure types. 

  • Plug-and-play standardswill evolve to provide mechanisms for easily embedding clinical logic for decision support into clinical workflows.  Currently, extensive hand-coding and customization is often required to implement the clinical rules for best practices within an EMR.  The Health eDecisions initiative promises to make it easier to share and implement clinical decision support protocols.

Back to Top

 

Exploiting Really Big Data

“Big data” has been used to describe almost any type of analytics initiative. Here we consider “big data” projects that incorporate clinical details about patient populations on a regional or national scale, that have the potential to impact health measures for the US population as whole.

Large-scale aggregation of clinical data for analysis could dramatically transform several aspects of healthcare.

Examples include:

  • Coordinated care communities can be enabled by sharing patient medical information among providers linked through a health information exchange (HIE). While the public discussion of HIE benefits has focused on improving the handling of individual cases with transactions (e.g., sharing previous test results, alerting a primary care physician to a patient’s emergency room visit), value may also be created by combining information from all community providers into a data warehouse. 

A bold experiment in aggregating data is underway in Maine. HealthInfoNet plans to combine the state’s all-payer claims database with clinical data from the HIE.  This would enable more meaningful analysis of health trends across the whole population.

  • Public health surveillance and response could be improved by aggregating comprehensive data about nationwide patient treatment.  More extensive data acquisition, aggregation and analysis could speed detection and response to disease outbreaks and provide better analytic support for other public health initiatives.

Though a major data collection effort focused on surveillance was not included in the PPACA health reform legislation, many provisions are foster the creation an infrastructure of electronic data capture, data standardization and data exchange. This could be leveraged in the future for more comprehensive public health surveillance.

  • Comparative effectiveness research (CER) will become urgent in the years ahead as concerns about cost and quality grows and new medical treatments are rapidly introduced.  In addition to existing institutions pursing CER, the health reform act created the new Patient-Centered Outcomes Research Institute (PCORI), which provides grants. 

An explicit priority of PCORI is to encourage broader use of electronic data in CER by funding research that “promotes a more comprehensive, complete, longitudinal data infrastructure” and improves analytic methods. While the degree of future political support and funding for PCORI may be uncertain, the general trend towards broader and deeper clinical databases for CER seems clear. 

  • Clinical trials of new medications and devices are essential to science and patient safety but are also time-consuming and expensive. Analytics applied to large clinical databases have the potential to improve several aspects of clinical trials. Recruitment of patients for participation could be accelerated by mining clinical databases to locate eligible patients.  Design of clinical trials can be improved by using clinical databases to ensure optimal coverage of different patient populations within the trial. Larger data sets collected during clinical trials may help identify additional indications for medication use and help flag adverse events before they become public safety issues
  • Transparency of provider quality for consumers is another major opportunity for ‘big data’.  Initial efforts by the Centers for Medicaid and Medicare Services (CMS) and others to create transparency around quality have validated the need for quality data to guide consumer choices – even if they have fallen short so far of providing readily actionable information on a broad scale.  Both government and private companies will continue to pursue better data and analytics to support consumers “shopping” for healthcare.
  • Challenges and opportunities for “big data”.Major opportunities exist for using ‘big data’ to improve costs and quality, though there are challenges to exploiting these opportunities. These go well beyond technology and include issues such as the lack of alignment in financial incentives among patients, providers, payers, and suppliers and the legal barriers to the aggregation of medical data for the purpose of analyzing cost and quality. 

Healthcare reform regulations have made initial steps towards addressing these challenges, but legal and economic barriers to broad collaboration on care improvement remain in place. 

Back to Top

 

Personalization of Care

Health systems are recognizing the opportunity to improve care by tailoring it to the needs of individual patients. Implementing new modes of personalized care delivery will be a major nexus of innovation in the years ahead.

Personalization will advance on at least two distinct tracks:

1.      Personalized Medicine 

Scientific advances such as genomic medicine involves tailoring drug regimens and other treatments based on the genetic characteristics and family medical history of individual patients. 

Genomic medicine is likely to expand widely in the years ahead:

  • Medical scientists are starting to understand the relationship between an individual patient’s genetic profile and that patient’s risk for developing disease – as well as their likely responses to different types of drug treatments.
  • Lower costs for creating genetic profiles of individuals will make such data much more widely available.
  • Clinical analytics software at the population and patient levels will need to fully utilize this new category of medical information. 

At the population level, analyses of genetic data may create more accurate predictive measures of disease incidence and healthcare costs. It can also identify individuals who would benefit from additional screening or preventive care (e.g., early detection of breast cancer may be enhanced with genetic data).

For clinical decision support in treatment of individual patients, genetic data will expand the clinical information available for consideration in rule-based guidelines for decisions by physicians.  For example, effective dosages of cancer drugs may be prescribed by taking genetic variations into account.

2.      Patient Experience

The patient experience may lead to dramatic changes in customer service and patient engagement.

Consumer analytics were initially developed to market and sell commercial products. Techniques for optimized consumers communication and motivation are becoming relevant for health systems as they recognize the importance of patient engagement in achieving positive health outcomes. 

As patient perceptions captured in HCAHPS surveys increasingly impact Medicare payments, health systems will need to use analytics to understand how to improve the patient experience.

Patient-centered care will mean tailoring the experience to match the needs and expectations of the individual patient.  Techniques like consumer segmentation may be borrowed from marketing to aid this effort.

Example:  An active retiree who plays tennis every day will have different expectations about what makes for a good visit to his orthopedist than a nursing home resident coming in for follow-up on her hip replacement.

Advertisers have long studied data to understand the modes and styles of communication that are most effective in attracting different types of customers.  Health systems are now recognizing the need to communicate effectively with different types of patients to ensure that they follow discharge instructions and avoid readmission, etc. 

Consumer analytics will be very important to the improvement of patient portals. Meaningful Use criteria require a basic ability to electronically provide health information to a patient. It will be essential to use web analytics to understand how patients like to interact with their hospital’s or physician’s website in order to make portals into effective tools for ongoing patient engagement, while positively impacting outcomes.

Online digital health coaching tools and remote monitoring systems are being deployed to implement wellness and disease management programs beyond the simple provision of health information. These systems capture valuable data from patients and can feed outcome and health status back into clinical databases, where they can be analyzed to help improve care.

The National eHealth Collaborative is one organization that is active in the area of patient engagement. In 2012, NeHC published a Patient Engagement Framework on its website.

Back to Top

 

HIMSS Clinical & Business Intelligence Primer

Module 1: What is Clinical Business Intelligence (CBI)?

Module 2: The Urgent Need for Clinical Business Intelligence

Module 3: Types of Clinical Analytics

Module 4: Essential Technologies for Clinical Analytics

Module 5: Organizing for Success in Clinical Analytics

Module 7: Opportunityies for Payers and Big Data