What Data Was Used?
The data was obtained from interview transcripts of semistructured interviews as well as one-on-one interviews with experts. Furthermore, issues that were identified from the workshops were also synthesized and used.
What Were the Models and Concepts Used in This Study?
Visualizing Data Governance in Health Care Practices
In the context of Australian primary health care providers, Andronis and Moysey (2013) conducted an investigation based on a structured framework (IAIDQ 2014) to assess the indicative level of maturity of data governance. The framework posed questions derived from the content of performance domains and focusing on data governance as illustrated in Figure 9.2.
The key pattern indicates that overall maturity for data governance was quite low across all performance areas. Where it was relatively mature, it focused on aspects of privacy or clinical records, and there was action required. None of these warranted treating data as a crucial asset. This does not mean that data governance itself is uncontrolled or patient outcomes are at risk. Rather, the effect is a less obvious combination of multiple versions of data, poor quality data, loss of efficiency, and relinquished opportunities.
In Australia, there are significant changes in health and hospital administration and funding that have major implications for the use of data in reporting and management. Today, individually managed and operated sites that were funded based on size have been grouped into local health networks that operate as single, multicampus entities funded based on health care outputs; their reliance on data to function thus
Figure 9.2 Structured framework for assessing data governance maturity.
has magnified. There is also additional government overlay in the form of 60+ local administration areas, known as Medicare locals, in order to better integrate primary health services on a localized basis. This generates a requirement for detailed data covering many services that were previously relatively scantily documented. The government administrators, on the other hand, want to make evidence-based decisions, which increases the need for complete granular data. Many data-intensive situations require the ability to combine data from multiple sources and, therefore, align definitions and compliant data recording and encoding (Andronis and Moysey 2013).
In the health environment, the key interest is to reuse once-entered data many times over. However, this is rather challenging in an environment in which nonintegrated systems coexist. Andronis and Moysey (2013) presented a high level model for data governance as presented in Figure 9.3.
Subsequently, the authors applied this model to the context of a large health provider as in Figure 9.4.
In Figure 9.4, the application of the proposed data governance model to a health care provider is depicted and explained as follows.
CXX-executive sponsor: The executive sponsor has overall responsibility and authority for (in this case) data governance and exercises this through control of a group of line managers responsible for different business and clinical areas.
Figure 9.3 A high level data governance model. (Adapted from K. Andronis and K. Moysey, Data Governance for Health Care Providers, Health Information Governance in a Digital Environment, Studies in Health Technology and Informatics, Volume 193, IOS Press, Australia, 2013.)
Figure 9.4 Data governance model applied to a health care provider. (Adapted from K. Andronis and K. Moysey, Data Governance for Health Care Providers, Health Information Governance in a Digital Environment, Studies in Health Technology and Informatics, Volume 193, IOS Press, Australia, 2013.)
Data owners: A data owner typically should be someone who is responsible for an area of the business that is a key user of the data domain. However, an important consideration for this role is the management of a data domain on behalf of the organization, not just for the area of direct responsibility.
Governance working teams: Governance working teams align with data domains, recognizing that knowledge of a data domain is usually spread across multiple individuals.
Business data stewards: Explicit inclusion of business data stewardship recognizes that day-to-day responsibility for creation of data rests with the people at the “coal face” of business and clinical processes.