Control, treat, monitor
Patient behaviour should be at the center of medical device design, because if it isn’t the technology will not realise its goal of improving people’s health.
The data driven medical device market is booming. But despite their potential to reduce the cost and scale of chronic disease, medical devices tend to be primarily designed around the needs of the medical profession and not the patient.
Our thinking has been shaped initially by the completely unscientific sample of the people we know. For example, Joe Geldart, co-author of this post uses a device (shown below) to measure his blood glucose level to monitor Type-1 diabetes.
Like many devices with a similar function, it has an attractive shell, but its functionality and usability are limited:
- It only measures one thing, so there is no context to the data
- It doesn’t allow Joe to view data over time, therefore it’s not able to provide any insight
- It only allows to export data to a laptop via infrared
- It needs to be used in conjunction with a finger-pricker, as well as consumables (test strips, in a sealed container, and lancets), which all live in a case, so its quite fiddly and actually takes up quite a lot of space.
The same questions apply to the design of any data driven device, whether lifestyle and fitness oriented or those with a medical use: How to help the user make sense of the data that is presented to them? How accurate is it? How easily does it fit into their lifestyle? How affordable is it? How secure and private is it? Can the data be shared with third parties?
That led us to a simple requirement framework for any data driven device:
The optimum (most usable) data driven device (and associated apps and services) sits in the middle of this triangle. It is reliable and accurate; it enables the user to make sense of the data; it is affordable and fits into a patient’s lifestyle: it’s not intrusive, it stays charged for long periods of time etc.
Depending on the purpose of the product, emphasis might be given to one of these dimensions over the others.
Our hypothesis is that there are four broad categories of device based on user motivations.
Health conscious people with no specific condition to address aremotivated to improve their health and general lifestyle or fitness. They are not ‘patients’ and may be driven by a competitive urge, the desire to prevent a future illness and to reduce its probability. This group is characterised by voluntary action and self-motivation.
People with a chronic condition may use a device to control their condition and prevent it from becoming acute. An example would be Joe, a type 1 diabetes sufferer who monitors blood sugar levels. This group is likely to be operating under the directions of a medical professional and may be resistant to monitoring.
Acute care patients have an urgent short-term need for treating, for example, a kidney disease. As above, monitoring will be under the directions of a health professional. The level of compulsion is high (“accept this monitor or you may become gravely ill in the next few weeks”); the threat near-term, not on a distant horizon, and thus resistance will likely be lower and compliance with doctor’s orders higher.
Palliative care patients may be more or less resistant to wearing a device, whether to monitor the progress of a condition, collect data or deliver care, such as continuous pain relief.
Already we can see that the different motivations are likely to warrant a different approach to user experience.
If we plot our simple framework against our four users groups, we can start to form a hypothesis around the priority of device requirements for each group.
Health conscious people with no specific condition to address are likely to prioritise how the device fits into their lifestyle. ‘Expert users’ may also give great importance to the depth and accuracy of data.
For people with a chronic condition, all three dimensions potentially matter: the accuracy and depth of data is necessary to understand peaks, troughs and patterns in detail. Lifestyle fit and affordability may matter just as much to overcome potential resistance to monitoring.
Acute care patients are likely to focus on the depth and accuracy of data. Because of the short-term need for monitoring and near-term threat, affordability or lifestyle fit may be secondary.
For palliative care patients our assumption is that both lifestyle fit and data accuracy will prevail over affordability.
It seems to us that there is a significant design opportunity around the needs of patients with a chronic condition. First there is potential to prevent a condition becoming acute, directly or through complications, and help the patient to be as healthy as possible (and save on treatment). However this group might not always be compliant, which offers an interesting design challenge. Crucially there is potential to win over this group in the long term.
What, for example, would a blood sugar monitor be like that was designed around the user experience? It probably wouldn’t use infrared or USB to upload data but the patient’s smartphone. We plan to explore this in further articles.
We’re still exploring this area, so we’d love to hear from people with relevant ideas and research in this area. One thing we are pretty sure of: when it comes to health data and services, one size does not fitbit all.