The 8 questions you need to answer about the data your smart device should collect.
In today’s digital world, data from smart, connected medical devices is growing exponentially and faster than organizations can keep up. At the same time, device users are suffering from data overload with a suite of devices that measure a variety of complex diagnostic and behavioral data. As a result, the pressure is greater than ever for MedTech companies to leverage this data to make faster, patient-centric decisions to improve the quality of care, decrease healthcare costs and improve outcomes.
But there’s a problem. There is a misconception in the digital medical world that your smart device should collect all the data it possibly can. The thinking is that even if that data is not useful now, maybe it will be one day. Not only is this strategy going to complicate the product development process, it is also going to put you in the same boat as the medical companies described at the beginning – drowning in data.
So, how do you determine what data to collect and where to draw the line? First, let’s take a step back. Before you can begin the process of deciding what data to collect, you need to have well defined problem statement. In other words, you need to understand your stakeholder’s problem and identify all of the potential sources of data that can help you solve it. Then the challenge begins to determine which of these data sources are going to be most useful and beneficial for your user.
Not all data is created equal.
In the medical industry, asking the right questions can not only save you time and money but can also help you save lives. Here are eight questions you need to answer about the data your smart device could collect:
This simply means the data about the data. What do you need in order to make the piece of data you are collecting relevant? For example, if your product is a digital glucometer, the diagnostic reading could have more value if it was paired with behavioral metadata including a time stamp or data indicating the last time the patient ate. Leveraged correctly, combining data points and sources can tell a different story that neither could tell on its own.
Does the data already exist or does it need to be generated? There are millions of data sources available to the MedTech world. If the data needed already exists from another source, your device will not need to create it.
There are three potential owners of data. Internal data is the data that you collect and therefore, your organization owns. This is typically the easiest for companies to use and integrate into their medical device solutions. But it gets trickier with external data. For example, if you’re reliant on another organization’s data in order to generate the information you want and they denies access for any reason, you’re in trouble. If you can’t own the data you’re using, then you need to make sure that at least you aren’t going to lose access to it. Also, if you are relying on data that users have provided to you, then you need to be aware that this data may not be completely accurate.
Data comes in two distinct forms – structured and unstructured – and each must be dealt with in a specific way. The fundamental difference between structured data and unstructured data, as you might expect, is that structured data is organized in a highly mechanized and manageable way. The challenge with structured data is ensuring that the format is compatible with other structured data sets. Unstructured data, by contrast, is raw and unorganized and therefore, will be more time consuming and costly to sort through and make sense of it.
Where is the data being stored – in the cloud or on an internal server? Among the advantages of data stored in the cloud are accessibility, lower costs and scalability. A cloud solution facilitates easier collaboration with users and partners. Information can be accessed anytime from anywhere as long as you have Internet access. Cloud solutions also lower IT investment and overhead costs, particularly for mid-size and smaller companies. However, cloud data is subject to scrutiny because basic updates can easily alter the user’s environment impacting FDA compliance. Internal servers, therefore, provide more features and controls for maintaining regulatory requirements and in some cases, security.
Data size is another important consideration. The larger the data set (gigabytes and beyond), the slower your medical device will process it. Size discrepancies in varying data sets can also reveal incompatibilities merging data which will cause headaches in product development. Size of the data will affect your product’s battery life, processing power, storage capacity and more.
The MedTech industry deals with sensitive information all of the time. It’s part of the job. Your internal data needs to be easily accessible to users, but as an organization, you also need to be complaint to HIPAA and other FDA rules and regulations. And if you are getting data from other sources, you need to make sure that data follows the same rules. Unfortunately, for the medical device world, data security is not a luxury, it’s a requirement.
Many medical applications rely on real-time data. This means the cycle from when the data query was made to when the data is obtained needs to be virtually instantaneous. The rate at which the data can be analyzed and displayed to the user will depend on a number of factors we already discussed (e.g. size, location) but real-time data will require more device processing power.
There is a reason why we think of a digital product strategy as an information strategy. The value of your digital medical device is directly linked to the value of the information your device provides. A holistic evaluation of the data you need to deliver this value will help you streamline operations, reduce costs and ultimately improve your quality of care. So don’t drown in data if you don’t have to!