A Lean Startup Approach to Predictive Clinical Analytics in Healthcare, Part II

In the previous post, we explored why the time is right for healthcare organizations to implement an advanced analytics program as a core capability. To execute on this, I proposed that organizations consider the StartupCommons.org startup framework which outlines three stages of maturity. Using a startup framework to implement this capability encourages thinking about the program more holistically, with the intended outcome of improving the odds of long-term relevance and impact. In this post, we'll go deeper on the first stage, Formation.

The Formation stage of implementing an analytics capability should be to comprehend the business, organizational, and technical alignment and resources for long-term success.  There are many steps that clearly define value proposition and success criteria:

  1. Map to high-level organizational strategy and capabilities
  2. Establish a cross-functional team with the right stakeholders and executive sponsorship
  3. Plan for proof of concept to production
  4. Data strategy including data governance  and analytics maturity
  5. Business process strategy

The spirit of this last step is summed up well by a recent quote from Chad. W. Konchak, Director of Clinical Analytics as NorthShore University Health System "Predictive Analytics is one of the least complicated aspects of forecasting clinical outcomes." (Source: Healthcare Executive Magazine, Volume 31, Number 5, September/October 2016). Using the results of predictive models requires changing business and clinical workflows. This can be a significantly heavier lift than the technical aspects of building the predictive model. We'll explore this step in more details as part of the Predicting stage, but it's important that the organization consider business process changes from the beginning of the project. It is recommended for one of the team members (either internal or external consultant) be an expert in change management. Many organizations will outsource this role to an objective 3rd party who can operate through this process without the encumbrance of political agendas and other outside influence. This agent should be involved in developing an implementation plan as part of the initial strategy, and they should observe the work done at all stages to capture best practices and observe obstacles. All of these learnings should be fed back into the implementation plan as the program evolves.

Note that four of these five steps are focused more on business issues than technology. It is not uncommon for the excitement of trying a new technology to entice teams to skip over some of these steps. Exercising discipline at this early stage and committing time up front for proper planning can improve the odds of success (can we find an online quote on the success of projects that are properly planned?). Make sure your team has a solid Understanding of the reasons why an analytics capability will contribute to the organization's success, and what resources are needed to make it happen.

In the next blog, we'll look into the steps commonly observed as an organization starts the work of proving the value of predictive analytics.

Read Part I of this blog series

Published on Categories ArchiveTags , , ,
Andrew Bartley

About Andrew Bartley

Senior Solutions Architect for the Health & Life Sciences Group at Intel Corporation. Work with providers, payers, life science organizations, and government agencies around the world as a trusted adviser on the development and implementation of leading-edge collaborative care and distributed care solutions. Leverage the latest mobile business client (2-in-1's, tablets, smartphones), Internet of Things, and wearable technologies to deliver superior patient experiences that achieve critical cost, quality, and access goals. Collaborate closely with Intel business and product development teams along with industry partners to define and evangelize standardized architectures that incorporate security best practices and enable the latest data analytics techniques. Regular speaker on the topics of innovation in healthcare and entrepreneurship. Contribute to thought leadership on these topics through the Intel Health & Life Sciences online community. Specialties: Healthcare solution architecture, connected care, medical devices, IoT, wearables, predictive analytics, product and project management, mobile application development, customer journey design, business development, strategic finance, agile software development, lean, SOA, UX, web services, big data solutions, system architecture