14 July 2009
The notion of business intelligence systems in healthcare is almost an oxymoron. There is no doubt HBI solutions are the Rosetta stone for healthcare reform, process improvement, quality and safety, business development, and more. Business intelligence is a broad umbrella of competencies, technologies, IT solutions, operational and clinical practices. Albeit an array of tools, processes, and systems, the ultimate objective of a business intelligence solution is basically an attempt to understand the context of healthcare within the business sector.
There is numerous business intelligence systems created for healthcare and those modified for the healthcare sector. Organizationally, business intelligence may reside within a finance department, business development division, planning group, or as a standalone function. Recognizing the nuance of business intelligence systems explains why BI success stories in industry have not been replicated (in terms of its outcomes, return on investment, or products) in the healthcare sector.
If the premise of business intelligence solutions is to extract knowledge from the decision-making process, a single BI solution will never exist to support healthcare until the industry has become standardized beyond codification (MS-DRG, CPT, ICD-9) of its services outcome and quality measures defined and imposed on care delivery by disparate components of the healthcare macro system
Nine years ago my colleagues and I began a journey to create a robust, dedicated planning function for an integrated health system with a relatively broad reach and over $IB in patient driven charges. Up to this time the status quo was a small team responding to demography questions and providing standardized reports when asked. The locus of control for the research pair was basically unchanged from twenty years earlier when they joined the system. Hard copy was replaced by computer programs and primary research had shifted to Google searches; the catalyst for work and organizational want were unchanged. The scenario was more akin to a small, stand alone hospital or large group practice. In essence, the research function was equivalent to providing stuff.
The evolutionary journey has a clear end goal: a group providing knowledge versus statistics and an integrated decision support system versus demographic, historical response. The multiyear journey possessed senior champions, clear direction, internal marketing and communication. The analytic team was launched at a point the organization’s wants were broader than existing deliverables. Everything appeared to be right place, right time.
Wrong!
Change management literature includes a plethora of hows and strategies for success. They is a typically a footnote disclaimer that often goes unnoticed. Paraphrased, the footnote could read: “At times leadership will agree to a want or present need. When this want is fulfilled staff may be given answers and solutions they don’t want to hear; therefore the fault is with the information.” This stance can be made because the BI process is delivering knowledge in lieu of data points. In one of those “emperor is wearing no clothes,” moments the reality had taken a new twist. Historic reports and method of sharing those reports was nondescript. Most often the data supported a decision in process and rarely was the product foreign to the predetermined outcome.
The process was fascinating to observe. Leaders asking for data could either accept the information or state that the system was afoul and due to the nature of healthcare was unable to provide the real response. Research staff provides graphs or maps; when challenged the fault was in how the query was posed. The cyclical nature of the process is self-perpetuating – either the product is spot on or it cannot be my issue…it’s theirs. Hence, we will term the traditional process Teflon Thinking.
These soft barriers don’t begin to expose the organizational and systems barriers. These include, but are not limited to:
1. Healthcare as an industry and a system is comprised of various business models glued together by the patient experience. Each business model has its own data process in response to its own need. And these systems do not track outside of themselves or beyond the want it was a solution for.
2. Systems are proprietary and are designed for business units that are proprietary. The care delivery system is not a system of interconnecting parts. Remember when you were a little kid and you had different size legos from different sets. Once they were mixed together you could never build anything because they were of different gauges.
3. Most BI systems are derivatives of finance systems. Their objectives are not the same.
4. Patients lack a common identifier as they move in and out of the system; providers lack a common identifier as services are rendered.
5. Medicine remains an art. Few physicians follow standardized protocols and processes; metrics don’t match.
6. Most integrated systems budget by cost center and mathematically spread indirect costs (sometimes even direct costs). Even accounting for a spread, the average financial system is not set up for cost accounting akin to industry resulting in financial challenges.
7. IF service lines are in place managers often manage cost centers instead of a patient experience.
Any good innovator or strategist would look at these barriers as a challenge to overcome. So I share one last story. For several years the inability to quantify ambulatory and outpatient services has been maddening. Following a yearlong review of systems, one was selected for reporting, codification of services, and a start to creating outpatient or ambulatory knowledge. The absence of reports similar to inpatient market share, estimated market penetration, values in the hundred thousand – the result was disbelief.
Evidence based medicine began as experimentation that eventually became accepted once a sufficient amount of data was available. Evidence based management has a similar process. At some point the silo nature of healthcare will be forced to break down its barriers and assume a system of care approach. The combination of reform and recession may stimulate alignment. This is a situation that first to market will have an advantage; it will be short lived. Business intelligence solutions are too early at this point to be real solutions. Creating business intelligence solutions for service lines (i.e. cardiovascular, neuroscience) or portions of the delivery system (i.e. pulmonary) are useful because it is experimentation. Let’s just be honest, business intelligence for decision making in primary care is a step beyond where we are today – it’s only the landing of the staircase. The integrated system will bop to the top first and win. That’s when a business intelligence solution will be an organizational want.
Leave a Reply