Misleading Healthcare Graph

As healthcare is so dominant in the information, I want to show an instance of a confusing and misleading graph from a hospital. Figure one demonstrates the benefits for the very first 6 inquiries from a survey of the patient experience measures from the Hospital of Special Surgery. This demonstrates that a single graph can be acceptable by itself but puzzling or even misleading when positioned in a group of graphs. The particular issue with this group is the use of distinct scales, ranging in highest values from 70% to a hundred%.

Figure one. Misleading Group of Graphs Supply: http://www.healthgrades.com/hospital-directory/new-york-ny-manhattan/hospital-for-special-surgery-hgst6cf97b36330270#Satisfaction Click to enlarge

Every single panel presents the percentage of patients at the Hospital for Specific Surgery that gave the hospital a rating of 9 or 10 and compares that worth to that of the nationwide regular of hospitals.  Note that each bars in every panel have the exact same scale so that the comparison inside of panels is legitimate. A reader may possibly also be interested in the comparisons of ratings of distinct inquiries for the exact same hospital for illustration, whether or not individuals are more pleased that the hospital employees often explained medicines or that the individuals felt that their discomfort was constantly nicely managed. The bar for explaining medicines is longer than the bar for controlling soreness for both the Hospital for Unique Surgery and for the nationwide common. Nonetheless, the labels display that individuals are much more pleased with the management of soreness. These figures are visually misleading since their scales differ.

Groups of graphs ought to use steady scales, colours, order of classes and other graphical elements. At times this is not feasible simply because the values in the different panels vary so considerably. See The Elements of Graphing Information by William S Cleveland or my Creating Much more Efficient Graphs for guidance on handling this scenario. Just because a single panel is acceptable does not mean the graph with several panels is also acceptable.

Misleading Healthcare Graph