A lot of effort goes into making healthcare information systems (such as electronic records, etc) easy for end-users to navigate and comprehend. The better the usability, the more likely systems are to be accepted and used in meaningful ways. Although much effort goes into enhancing the usability of these “front-end” systems, there isn’t much focus on the usability of the analytics tools that leaders of all levels within healthcare organizations (HCOs) use to guide decision-making.
With the proliferation of healthcare information technology (HIT) and information used for decision-making, many more “non-analytical” users are accessing analytics in the form of dashboards, reports, predictive algorithms, and even simulation models. It is no longer a given that another data expert, but rather a nurse, physician, or other clinical expert, will be a potential user/consumer of analytics. Usability of analytics for these users, then, becomes an important issue because they are unlikely to have the time or inclination to decipher information that is not comprehensible and accessible,.
Much has been written on the topic of the usability of information technology. When building and deploying analytical tools for decision-makers, there are a few key areas that I focus on:
- Accessibility – do the end-users (i.e., decision-makers) know that the information they need exists, do they know where the information they need is located (i.e., what tool, report, or dashboard contains the information), and do they know how to (or are they able to) access that information? (Click here for a recent article focusing on accessibility of information).
- Analysis – is the data summarized and analyzed in appropriate ways (for example, are the necessary statistical methods used and are the graphs, charts, and other visualizations employed correct for the type of data), and is the level of analysis detailed enough (or is it too detailed) for the intended use?
- Presentation – is the required information presented in a clean, uncluttered layout, are graphical elements consistent (i.e., use of red, yellow, and green to indicate “bad”, “poor”, and “acceptable” performance levels), and are all elements of the information labeled adequately and correctly?
Something that I have noticed is that the key to usability isn’t how flashy the visualization is, but how quickly users can assimilate what is presented to extract the intended message(s). When improving healthcare quality and operational efficiency, basing decisions on correct analyses is of paramount importance; flashiness may only detract from the core purpose.
The effort to enhance the usability of analytics is well worth the time. By ensuring that the accessibility, analysis, and presentation of information is meeting the needs of end-users (by ultimately improving their decision-making capacity), analytics will become an even more important tool for improving healthcare.