Concepts and Methodologies of Data Validation
Explain the concepts and methodologies of data validation.
Think in terms of data feeds from an EMR/EHR that ultimately ends up in an
enterprise-wide data warehouse. How can data validation ensure data quality and
integrity? What are recommendations on how to ensure only validated data ends
up in consumable form?
As healthcare organizations (HCO’s) begin to leverage data to
improve quality initiatives and overall performance. Disparities in how data is
created, shared, and stored create barriers to analyze data. Organization must
focus on a data management to improve the way data is entered and stored
(Strome, 2013). Kerr, Norris, and Stockdale (2008) suggest that the reactive
approach in healthcare today is due to poor data management. Primarily users
and outside data sources (e.g. electronic records from a referring provider)
are a major cause of these disparities.
Total data quality management (TDQM) is a methodology that
takes a proactive rather than a reactive approach to improving data quality.
Marsh (2004) provides a four step approach that can be leveraged to improve
current data disparities in healthcare. The steps include: audit, clean, error
prevention, and compliance. Auditing the data entails reviewing current data
for any discrepancies such as missing values. Clean, refers to removing errors.
Error prevention is the step that includes staff education, and compliance is
continuous monitoring of data consistencies. Leveraging different methodologies
to improve data is the first step to improving and maintaining consistency in
data quality.
One of the most important factors in ensuring data ends up
in consumable format is by proper training of end users and implementing data
stewards. According to Juran “data are of high quality if they are fit for use
in their intended operational decision making, and other roles (as cited in
Strome, 2013).” Improving communication with end users and sharing details on
analytical processes will improve end user understanding. As well, training
will also ensure that inappropriate workarounds are avoided in order to improve
the quality of data. Data stewards are individuals who maintain and monitor
current processes of how data is used, entered, and stored. Maintaining
consistency across multiple functions within a HCO will help alleviate
disparities and improve data quality. Moreover, data stewards work closely with
analytics teams to determine when systems should be replaced and how current
systems are being used. This provides
analytics teams with the necessary information needed to implement change as
needed (Strome, 2013).
With health information systems being used for different
functions across organizations from care management to revenue cycle
management. Organizations require strong governance strategies to improve
enterprise wide processes. Whereas, in the past the focus was how to improve
localized processes. The environment of healthcare requires that organizations
focus on the entire enterprise to improve overall quality and performance
initiatives. If organizations can work as an enterprise rather than in
isolation they will prosper in years to come (Strome, 2013)
References:
Kerr, K., Norris, T., & Stockdale, R. (2008). The strategic management of data quality in
healthcare. Health Informatics
Journal, 14(4), pp. 259-266. doi: 10.1177/1460458208096555
Marsh, R. (2004). Drowning
in dirty data? It’s time to sink or swim: A four-stage methodology for total
data quality management. Database
Marketing & Customer Strategy Management, 12(2), pp. 105-112. Retrieved
from: http://search.ebscohost.com.proxy.cc.uic.edu/login.aspx?direct=true&db=buh&AN=16249141&site=ehost-live
Strome, T.L. (2013). Data quality and governance. Healthcare
Analytics for Quality and Performance Improvement (pp. 6-7). Hoboken, New
Jersey: John Wiley & Sons Inc.
