Using macros in microsoft excel to facilitate cleaning of research data

Abstract

<b>Background:</b> Retrospective chart review studies may be delayed by inability to export clean clinical data from an electronic medical record (EMR) or data repository. Macros are pre-programmed procedures that can be used in Microsoft Excel to help streamline the process of cleaning clinical datasets. <b>Objectives:</b> To demonstrate how macros may be useful for researchers at community hospitals and smaller academic health centers that lack informatics support. <b>Methods:</b> Using an intrinsic function of our institution's EMR, vital signs and lab results from 20 individual hospitalizations were exported to a spreadsheet. Two macros were developed to sort through these datasets and output them into a specified format. The speed of macro-assisted data cleaning was compared to manual transcription. <b>Results:</b> Time spent on data cleaning was significantly reduced when using macro-assisted sorting compared to the manual approach for both vital signs (46.5 seconds versus 12.3 minutes per record, a 94% reduction; P < 0.001) and labs (13.7 seconds versus 2.6 minutes per record, a 91% reduction; P < 0.001). <b>Conclusions:</b>Macros offer a flexible and efficient tool for cleaning large sets of clinical data, particularly when an institution lacks informatics support or EMR functionality to export clinical data in an analysis-ready format.