The measure developer must explicitly identify types of data and how to aggregate or link these data so the calculation of the measure is reliable and valid. The measure developer should proceed carefully when merging data from different sources or systems to prevent errors. Some potential areas where problems may occur include

• Difficulty in determining which data represent duplicates
• Different units of measurement used by the different data sources (e.g., different age groups, different time frames)
• Different quality controls used by data sources

It may be necessary to clean the merged data. If the measure developer finds inaccurate, incomplete, or unreasonable data, they should correct data errors or omissions.

Electronic Clinical Quality Measures (eCQMs)

For eCQMs, conduct preliminary feasibility assessments to confirm availability of the information within an electronic health record (EHR), ideally in a structured format. The feasibility assessments should include both the data model and how various EHR systems map and store the data elements. Also review the specifications on either the draft external document (prior to into into Measure Authoring Development Integrated Environment [MADiE]) or on exported files, and other documentation for criteria such as

• The eCQM header includes appropriate information in the data fields or contains preferred terms. 
• Correct mapping of the measure data elements to the correct category and datatype in the Quality Data Model (QDM). 
• Each QDM data element is subsequently linked to an appropriate value set(s) or direct reference code.
• Value sets and direct reference codes used represent the most current Interoperability Standards Advisory recommendations. 
• The addition of CMS additional supplemental data elements.
• Testing of the eCQM logic in MADiE.