By Shalen Young
Making business decisions for your health care organization based on the best available data, as opposed to intuition, seems like a no-brainer. But, even with a fully committed executive team, which is necessary for success, there are many pitfalls that can derail the process. While the data collection and analysis will mature and improve in time within committed organizations, the two most important initial questions to answer are: 1) Who within the organization will provide the data and analysis? and 2) How accurate is the data?
Who should own the data and be responsible for its manipulation and reporting? This decision often is not approached with the care and thought it is due. It is critical that the data come from a group unaffected by the decisions the organization will eventually make, so it is not manipulated to favor a specific outcome. Many organizations mistakenly leave data reporting up to individual departments instead of centralizing the function with dedicated staff, many times in order to avoid increasing headcount. In the long run this approach will cost significantly more, as resources will be misallocated due to biased data analysis.
Once a data analysis group is identified, the long and tedious process of data validation must begin. No company, regardless of size or complexity, has or will ever have perfect data. The degree to which it is accurate is another matter all together. While the data owner needs a good understanding of operations, it is imperative that the data be validated by the individual operational departments. When inaccuracies are discovered, processes must be redesigned to ensure more accurate data for future collection. There must be continuous education and revalidation of the data. Until your organization is comfortable with the accuracy of the data, use your data as a guide, while acknowledging known deficiencies.
International and multicultural organizations run into even more challenges when it comes to data quality. Language barriers can have a significant effect on educating front-line employees on data entry and process adherence. For instance, differentiating a patient visit from a patient encounter can be difficult to explain in English to native speakers, not to mention the difficulties of trying to accurately translate that to Arabic, Hindi, Chinese, etc. This makes process design critical to ensuring that data capture is difficult to do incorrectly. This may take several iterations, but focus on data quality is imperative to make accurate data-driven decisions.