DMD Project

A little background

The dental domain is complex, and we wanted to understand it from many different angles. One method successfully used was contextual inquiry, but we used a handful of other approaches as well.

Literature Review

With the help of our knowledgable clients, we reviewed literature on Medical and Dental records, Collaboration Aspects, Usability evaluation, safety, and errors, 3D and Advanced Imaging. Key findings included:

• Dental software today is not designed with the advantages of a computer in mind; instead, paper charts are duplicated exactly. There is plenty of oppertunity to integrate new technologies, concepts, and interactions. DMD uses 3D, among other technologies, to support dentists' work.
• Dental software must support the tasks of the individuals involved. The navigation of our system is set up to support a dentist in accessing and reviewing records. The first page for a patient, the overview page, is particularly important in providing dentists with the initial information they need.
• Dental software must become a more natural part of the environment, allowing them to utilize the computer effectively and efficiently during patient care. Many users found DMD intuitive to use partly because of the simple touch screen technology.
• Current commercial dental PMS packages do not successfully prevent user errors, match the real world, or provide consistency in terms of interface widgets or metaphors.

Software Review

To make sure we didn't commit the same sins as current dental software, we briefly reviewed eight dental software including Eaglesoft, Dentrix, PracticeWorks, DentiMax, Prime Dental, Softdent, ACE Dental, and Dental Symphony. Our conclusions:

• Most dental software is centered around administrative functions, with clinical functions added as an afterthought. There needs to be a clear separation between the two. Currently, DMD focuses on clinical functionality. If administrative functions are added, they should not interfere with the clinical part of the application.
• There are no standard icons, spatial or visual metaphors for many concepts commonly used in all dental software. Thus, we designed our own visual language, keeping it consistent throughout DMD. Part of this was examining how to represent hard and soft tissue charts, a topic that is expanded upon below.
• Many packages use nonstandard window management and dialog interaction techniques that are confusing and error-prone for first-time users. Even dentists familiar with the Windows interface found themselves confused by the many inconsistencies presented by the applications.
• Both feedforward and feedback are mostly of very poor quality. As the result of an action, information may change, appear, or disappear, in a completely unrelated or invisible location. In DMD, users know where they will end up when they perform an action.
• The hard tissue chart is the most visible clinical function in all of the dental software systems. While we have the hard tissue chart, more functionality is centered around radiographs, which dentists rely on for diagnosing.

Charting Comparison

One important visual element of a dental record is the chart. The chart shows a dentist, at a quick glance, the status of a set of teeth. However, there is no dental standard for these charts. Instead, each office seems to have their own preferred symbols and colors to represent decay, fillings, and other dental pathologies. We took two approaches towards making the perfect dental chart.

First, we sent out a survey to dentists around the country, asking them to fill out a chart for a fake patient we had created. This fake patient had a representative set of all the dental pathologies. The results from this survey would have allowed us to understand which symbols were most common and which were rare. As a second approach, we used each dental software to create one of these reference charts. The pictures are below, and they demonstrate the amount of variety between the applications -- not just in the symbols used, but even in which problems each application chose to show.

Cognitive Task Analysis

Our clients at the University of Pittsburgh Center for Dental Informatics provided us with data from their Cognitive Task Analysis (CTA) study. In these CTAs, dentists prepared treatment for patients based on a controlled set of information. Although the patient was not present, a wide range of data including charting, radiographs, intra-oral photos, and other written material about the patient's history was available. How the dentists went about preparing the treatment, and when they accessed each different peice of information, was analyzed and broken down into a spreadsheet. We took this data and re-visualized it, creating elaborate timelines of the information.

The timeline is a rich visualization that is a bit difficult to interpret alone. Although there is no common pattern in later stages of analysis, all three dentists do start out with patient info, then medical history, and most go on to dental history. This led to the creation of DMD's overview page for each patient, which dentists loved.

In the CTA data, dentists often switch between artifact types quickly and repeatedly, or look at several artifacts at the same time. In DMD, switching between information can be done with a single tap. The 'View by Tooth' tab enables dentists to see multiple types of information at one time, reducing the need to search for data.