Physician Data Analytics – Impact of Data Quality – Part 1

By Richard Howe, PhD, Executive Director of NTREC
Last month, I discussed five steps for a practice to migrate toward population health. One area that has a major impact on both practice operations and future analytics is data quality.

Over the past year, I have participated in a work group called the ONC-HIMSS Patient Matching Community of Practice, co-led by the Office of the National Coordinator for Health IT (ONC) and HIMSS. This group focused on patient identification and matching. After reviewing the issue from many perspectives, one of the conclusions was the quality/integrity of patient demographic data at the point of patient registration is key in correct patient matching. Next month, I will present more details from the patient matching work group.

Obviously, as electronic health records (EHR) get to the point of being readily shared across organizations and physician practices, correctly identifying the patient is the most critical step. You certainly do not want to be reviewing clinical information from two different patients – all along assuming the data is from the same patient. This could lead to huge clinical errors.

In a recent article (“Data Quality: The Key for Integrated Analytics “, in Healthcare IT News, July 9, 2015), several examples of data quality issues that may happen in an EHR were noted. These include:
• Erroneous patient identifiers, such as a missing social security number, misspelled name, incorrect sex, or transposed date of birth; 

• A standard numerical metric, such as blood pressure, written in text in encounter notes rather than in appropriate structured fields; 

• Generic diagnosis codes entered quickly or out-of-habit instead of more specific and actionable diagnosis codes appropriate to the patient. Correct coding is going to be even more critical as we move into ICD-10 coding in the Fall of 2015;
• Inconsistent entry of standard codes, such as National Drug Catalog (NDC) for drugs, derailing bulk analysis.

The article went on to describe a few key action items to improve data quality:
• Get rid of unstructured “free text” notes and maximize use of structured data;
• Use the correct structured data element. For example, this could mean using the correct NDC, LOINC and now ICD-10 codes;
• Know your reporting system. By reviewing the information you want from the output reporting side of the system, you will quickly determine if you are collecting the correct data elements in the correct format to produce the correct reports.

In summary, using standard patient demographic data elements, using standard data structures and coding, and training all staff on how to improve quality of the data input (both demographic and clinical), will greatly enhance your EHR system to support your practice from both an operational and a data analytics point of view. There really is gold in your data!