Cavanaugh Consulting Group

Chief Data Officer – a New Requirement or Unnecessary?

Two articles got me thinking about this, the first in Information Management  states that sixty-one percent of CIOs want to hire chief data officers (CDOs) within the next 12 months. The article goes on to list 15 requirements and in doing so provides some insight into why a CDO may be useful and what they could accomplish. At the same time, another article postulates that the CDO addresses a temporary need.  Information Management  Neither of these articles were written with health care in mind. However, our industry is one of the most data intense and can benefit from the experience of those that are further along.

I will limit my comments to the provider segment of health care (hospitals, clinics and physicians) where I have spent most of my career. There is a lack of understanding leading to a lack of direction about what to do with all of the data healthcare organizations are collecting. Everyone knows it’s powerful and will eventually be essential, but few have the knowhow to address the opportunity. This maybe surprising to some since data has always been critical to health care and hospitals, clinics and larger physician practices have a department dedicated to managing health information (HIM). HIM is a bulwark dedicated to managing patient treatment data but rarely has the authority to dictate IT data standards.  Our data like our business model has its roots in segmentation. Developed at different times by different groups for different purposes. This was a reaction to both the complexity of the data model and business reality. For example, you could develop and implement an effective automated tool for cardiac care services even if it had little linkage outside that department.  Even today in the era of “super vendors” (Epic, Cerner, McKesson, Allscripts, Meditech) hospital providers typically have over 100 different information technology systems.

We have many data sets, typically each for a special purpose. Some are for patient care, others for teaching, some are for research, others for business management. Few were established with an expectation of integration with one another and sometimes not even between related datasets. Standards exist within the data set but usually not between them. Historically most were departmentally developed and managed and paper driven. What I mean by this is that the automated data tools were often created from paper based thinking.  Now most are electronic, want to access and could share data with other parts of the health care organization. Understanding how the data can be usefully, allowably, accurately and securely shared is a major undertaking. This has been somewhat mitigated by organization wide systems that manage large segments of data such as electronic medical record applications. However, those application tools are largely focused on the operational needs of users not the management, planning and evaluation needs of management, medical science or public health.  Additionally, as the health care industry is restructured by ourselves and government mandate, we need to access the full range of data appropriate to our purpose; i.e. patient care, research, teaching or business to make informed decisions.

Describing the specifics of what needs to be done is beyond the scope of my remarks but I will comment on the need for the CDO as a thought leader and coordinator.

Soon after the Meaningful Use regulations were enacted most providers recognized the need for an individual to become the point of coordination. Data management and integration in a health care setting is daunting, even more so than meeting the Meaningful Use regulations. I was involved in the early days of the HL-7 data standards organization over 25 years ago and while progress has been made in integrating and sharing data we still have a long way to go within our provider organizations let alone between them.  Although some of the below listed activities are by necessity accomplished today by HIM leaders, business planners, medical directors, decision support specialists and others, there is a lack of overall coordination. The issue is that we need to do more because of the importance of accurate, timely and comprehensive data.  I see the need for the CDO in health care provider organizations and I see the role as critical for years to come.

Here are the points I found most useful in considering the role of the CDO, modified somewhat for health care:

  • Big Picture Responsibility – Vision and strategy, establish data policies, standards, organization and enforcement.
  • Manage Data Governance – Organize and implement policies, procedures, structures, roles and responsibilities
  • Study and Maximize Data Quality – Determine and improve the organizations data quality and maturity level
  • Explore Data Warehousing – A single massive database or a collection of data marts that are integrated. The data warehouse strategy should provide consistent, clean and integrated data.
  • Embrace Enterprise Data Modeling – The EDM model may never be completed, but it still delivers value by allowing CDOs to discover and resolve data discrepancies among different views and data sets.
  • Explore Metadata Management – An example of metadata is the information embedded in a digital picture – such as the brand and model of the camera, the data and time the picture was taken, etc.
  • Navigate Unstructured and Big Data – Unstructured data includes social media, emails, pictures, videos and sensor data like RFID.
  • Weigh Data in the Cloud – Weigh variables like privacy, security, compliance, ownership and performance questions tied to cloud storage.
  • Establish Performance Metrics – build performance dashboards providing timely information and insights that enable clinicians, researchers, employees and management to improve decisions, optimize processes and plans, and work proactively.
  • Maintain Security and Privacy – An especially important area in health care considering the almost daily protected information breaches. Many health care organizations have Privacy and Security Officers, if so, this may be a coordination function.
  • Develop Intellectual Capital – Defining the protocols, processes or techniques that identify what is unique about your organization. Predictive analysis tools that help the organization more quickly adapt to opportunities and manage the transition. Eventually, IT will become a commodity, the organizations uniqueness will be how you manage and use your resources, especially your data, knowledge and skills. CCG