Thursday, September 27, 2012

Cultural Principles and Patterns in Healthcare Analytics

On Sept. 13, HealthLeaders Magazine published an excellent article with interviews from several healthcare visionaries, entitled, “Metrics that Matter”.
The comments in the article subtly underscored some very important principles and patterns affecting healthcare analytics and data warehousing, especially the manner in which analytics evolves within the culture of an organization. It is critically important for CIOs and other C-levels to understand these patterns and principles, so that they can choreograph the execution of their strategy accordingly; know when to move quickly when they see the patterns emerging; and alternatively know when to induce the pattern when it’s not emerging.
Stated otherwise, I’ve made many mistakes in my own organizations — and witnessed the same in organizations that I was advising — when I failed to anticipate these cultural patterns and/or failed to follow the principles.
Highlighted below are specific quotes from the HealthLeaders article, then a summary of the general, related pattern.
HealthLeaders:  ‘When Cleveland Clinic deployed its first dashboards eight years ago, they were “relatively basic, showing overall volume information in the organization,” he says. “It’s evolved drastically, to where now on our dashboards, we have information that’s updated as frequently as every 30 minutes.”’
  • Pattern #1:  Analytics in the organization always trends towards greater complexity.
  • Pattern #2:  Analytics in the organization always trends towards more timely updates and analysis of data; and towards more real-time decision making.
In the early days of a data warehouse, simple static reports issued on a monthly or weekly basis are sufficient, but as the culture becomes more data driven — more agile in its response to data — the demand for more timely data and more complex analysis of that data increases. What does this imply for the CIO?
Your data warehouse technology must be capable of growing and adapting accordingly, towards more complicated analysis and more timely updates. However, deploying your analytics and data warehouse technology to deliver real-time, complex data analysis before the need coincides with the culture of the organization can be a costly mistake, technically and organizationally.
Real-time data warehouses are more expensive to deploy and maintain, but can easily be designed with no additional cost. Design for real-time capability, but don’t implement it until you need it. If you provide that real-time capability before the culture can adopt it, you face a negative ROI from the technology investment. Even worse, by providing real-time data to an organization when the tempo of decision making and cultural change is still measured in weeks and months, you stand a very high likelihood of alienating data analysts, managers, and executives who are wondering why you are providing them with real-time data when they prefer slow-time data.
Every culture has its own “Mean Time To Improvement” (MTTI). Gauging that MTTI and aligning the delivery of data so it stimulates a paced reduction in the MTTI without accelerating past the cultural capabilities of the organization is critical.
Finally, as the data analysis becomes more complex, so do the tools required for that analysis. Early in the analytic maturity of an organization, simple bar and pie charts with static content are acceptable. Over time, especially in healthcare, the analysis will mature into very complex statistical process control, predictive analytics, and pattern recognition. No single data analysis and visualization tool can meet every need, nor should it for the sake of cost control.
CIOs should pace their advocacy for data analysis tools to match the analytic maturity of the organization, likewise they should avoid the common fault of assuming that one tool can meet all analytic needs in the organization.
HealthLeaders:  ‘But the real key to today’s business intelligence at an institution such as Cleveland Clinic is distributing the information gathered by the digital nervous system not just to top leadership, but to all those in the organization with a need to know.’
  • Pattern #3:  The most successful analytically-driven organizations invest at least as much time and money in creating a data-literate culture as what they invest in data warehousing technology.
Data literacy is measured by: (1) The ability to interact with data through analytic tools, e.g., Excel, and (2) The ability to turn the knowledge that is revealed through that data interaction into actionable results, within the role of the person interacting with the data.
Quite often, I see organizations invest significantly in the technology of analytics, e.g., an enterprise data warehouse, but they do nothing to address the data literacy of their organization in such a way that the culture can fully exploit the value of the technology. Early in my career, I would frequently consult and advise on the design and development of data warehouse technology, but I ignored the data literacy of the organization. It was quite common for me to revisit these organizations 2-3 years later to find that my prized technology was collecting dust, like a public library in a community that couldn’t read.
In addition to leading the technology strategy, CIOs must lead the charge in increasing the data literacy of the organization by hiring data literate employees across the organization — this includes details such as changes in job descriptions and training existing employees to make the transition into healthcare’s “Age of Analytics.” As we have seen with EMRs, it is not sufficient to simply install the technology, but it must be Meaningfully Used.
HealthLeaders: ‘The goal was to populate an information repository to let service line workers go after whatever they need, Schooler says. “We’ve been heads down at this for maybe a year and three or four months, [and] we believe there is a three- to five-year initial deployment to get to where there is, what we would consider, a critical mass of information from across the organization.”’
  • Pattern #4:  The number of source systems whose content is represented in an enterprise data warehouse grows in proportion to the data literacy of the organization.
As the data literacy of the organization increases, so does the demand for new data content in the data warehouse. Again, we see metaphors to public libraries, where the demand for new and more challenging reading material increases as the surrounding community becomes more literate. In some healthcare organizations, where there is a high literacy rate and pent-up demand for data, the number of source systems feeding the data warehouse will grow exponentially at first, to as many as 50 source systems within five years.
CIOs must adopt data warehouse designs that can adapt to this demand for new source content and keep pace with demand, otherwise they will lose the faith and support of their analytic customers and, worse, will be blamed for suppressing the analytic maturity of the organization. Mapping new sources of data content into a data warehouse can be a very laborious and time-consuming process, especially for data warehouse designs that are Inmon-style, enterprise data models.
CIOs must anticipate this demand for new data content and plot their staffing strategy accordingly. The most valuable staff members in these source-system mapping projects are the source-system data stewards and application experts who are supporting production-system activities. They typically can’t be removed from their production system duties at a moment’s notice to focus for weeks at a time on the data-warehouse data-mapping project. The CIO can either hire this expertise for permanent assignment to the data warehouse or plan far enough ahead to ensure that the source system data stewards and application experts are available to the data warehouse team when needed. Again, timing is critical.
These are just a few of the many patterns and resultant principles that emerge in the analytic culture of an organization. Understanding and anticipating them is fundamentally important to the success of a data warehouse and analytic strategy. By observing their subtle appearance in the early-adopter organizations that are highlighted in this article, and looking for the same in their organizations, CIOs can plan and execute with even greater efficiency.

No comments:

Nuclear and Healthcare Decision Making

Nuclear warfare operations was my data-driven decision making environment before the healthcare phase of my career. It was all about recogni...