User Requirements
First responders are the target users for this product. The first responders primary objective during a mass trauma event is to prevent loss of life through administering medical aid to a patient. Their triage workflows often adhere to a strict procedure. The requirements below are a result from a contextual inquiry with a first responder.
- Triage patients through vitals assessment.
- Situational awareness is key. Understanding how many patients there are and where they are in relation to you.
- First assessment is always most difficult.
- Once a patient is treated in the field they are prepared for transport to a medical facility. Patients need to be transported within an hour of being treated.
- Communication is challenging. Easy way to transfer patient information to others in the field.
Technical Requirements
The hardware team developed the capability to report patient vitals in the filed with a sensor array that can be placed on each patient. The sensors have the capability to report vitals for heart rate, blood pressure, body temperature, respiration rate, oxygen levels and location information.
- Display vital sensor readings.
- Display geo location data.
- Display sensor reporting status (connected / not connected).
- Display sensor reporting delay (10 seconds since last reported).
- Develop a way for users to explore historical sensor data. Measurements reported since the array was activated.
- A mobile device will be the output format for the user interface. Assumed that first responders will wear the device their wrist or other means.
Sketching
I looked at ways to organize space in a horizontal and vertical layout options on a mobile device. Finding room for all of the data metrics proved challenging and the layout got cluttered quickly. Geospatial data could be added in the horizontal view.
First Iterations
The first designs focused on meeting the technical requirements and incorporating a triage encoding of patient’s vitals. The triage took on a color scale that matches triage practices in emergency rooms. Green is stable, Yellow is elevated, Red is needs attention, and Black is deceased.
User Feedback
- Actual sensor readouts of vitals are not helpful for triage. Trend data is useful. Are a patient's vitals getting better or worse?
- Would like to see trend data over time a defined time range. For example, 30 seconds (short), 2 mins (med), 10 mins (long).
- Sensor connectivity and reporting delay information is confusing.
- Triage and shock scores are useful for quick glance.
- Having access to more chart detail on the main screen would be useful.
Trend line Research
I reviewed trend line research from Edward Tufte to see what type of visualization could be used to represent trend information. Highlights from this research showed that data could still be understood without chart boundaries and only showing a trend line. Spark lines also showed good comprehension and could be represented with a normal range indicator.
One of the challenges with representing this data is that spark / trend up or down could be good or bad, it depends on the historical context. For example, a patient has a heart rate that is dropping down to within a normal range. That same patient could have a heart rate dropping to below normal range. Both examples have decreasing values but one is good and the other is problematic. The spark line visualization with a normal range indicator could work well for this problem.
Second Iterations
This round focused on displaying trend/spark line charts on the main overview page. It also only displayed vital information for the selected patient which freed up space to display more informative visualizations and geospatial data.
Future Considerations
This product is considered to be in the early stages of development and refinement. While it’s demonstrable there are still areas that need continued research. From a user experience perspective there are still many unknowns and use cases to define. Some examples are:
- Command and Control Use Case. Use within a command hierarchy where patients are assigned to first responders from a command and control center. This could reduce cognitive load on the first responders.
- Ergonomics. What hardware is best used for first responders to interact with the interface? A mobile phone, custom device (mobile phone with physical button add ons), using a wrist holster to secure the device to the first responder? Gesture interactions?
- Interaction Patterns. Could a voice interface be useful for basic interactions? For example, taking audio notes, photos, ect.
- Filed Test. A user study could yield tremendous information on how the hardware is used and where in the first responder’s workflow it could be most valuable.
