Part of DiaTrack is the ability to view patient information in different forms. DiaTrack introduces a visualization that shows the progression of treatments and the corresponding changes in patient metrics. This is done by superimposing line charts that show selected measurements over time against varied backgrounds the represent courses of treatment.
Treatment courses are depicted as horizontal blocks of time with informative coloring. The coloring scheme is based on two factors: the intensity of the treatment and the compliance of the patient. The following chart shows the color scheme for treatment blocks:
By characterizing treatments as more or less aggressive, a horizontal graph showing treatment intensity over time (with patient compliance) can be depicted like this:
As treatments change, becoming more aggressive, the colors shift from greens to blues. Patient compliance with the treatment (as self-reported by the patient) is depicted by increasingly dark hues. The darker the hue of the treatment block the more compliant the patient has been with the treatment.
Selected diagnostic measurements can then be imposed on the treatment/compliance chart in order to give a quick visual impression of how the patient is responding to the shifting periods of treatment and compliance. Here is an example of how such a graph might look:
In this graph, each dot represents a visit of some kind where the selected measurement has been collected. Additional metrics can be selected. Because this graph is intended to be used as an intuitive summary, there are no grids and scales. The scale is relative to ranges observed in many patients over time. Superimposing an additional type of metric, therefore, does not introduce scaling issues. The metrics legend is provided elsewhere on the page.
A relative metric scale requires some explanation. For most patient metrics there is a “normal” range within which values do not themselves raise a medical concern. Both above and below that range is a problem. For instance, if your pulse is 180 that is a problem, but it is also a problem if it is 30. Ideally, pulse will range between some upper and lower values.
The problem is that we want to be able to graph values that have different normal and limit ranges. If BMI is on the same graph as HbA1C, then the scale means something different for each line. By placing a normal range on the visualization chart using dotted lines, we can scale each metric independently within the range so that each graph has consistent meaning. For BMI, the “normal” range is between 18.5 and 25.0. The extremes are 15.0 and 40.0. So a BMI value is scaled differently depending on which part of the BMI scale it falls into, giving emphasis to abnormal values on the chart. And HbA1C will be scaling between the extremes of 4.0 and 12.0 with a normal range of 7.0 to 7.5.
The reason that this is acceptable is that the visualization chart is merely an intuitive tool. The chart does not exist to show precise values. It exists to show values deviating from a normal range. By using independent scales for each metric type, we can graph more than one patient metric without losing the value of the chart.
TODO: Show the popup boxes when cursor hovers over a point or treatment block.