Some sensors might be external, either on your body or on Internet-connected appliances, but many might have to be internal. By combining sensor data with other data non-health streams such as location and current activity, artificial intelligence functions would watch for indications of developing disease, or in the case or already recognized diseases or assistive needs, could modify or control devices that change assistance levels or deliver therapeutics exactly as required.
The sophistication of these interventions would be tuned by access to your cumulative personal health records, as well as to developing best health practices based on big-data analysis of the growing health experience of millions of other people.
From a technological perspective, this future does not envision capabilities much beyond those already available. The area perhaps most in need of further development would be remote medical sensors.
But this vision would face a host of implementation barriers, including ready interchange of health data, privacy and regulatory concerns, and a consumer’s willingness to be ‘wired.’ But assuming that there will net benefits of improved quality of life and longevity, at an attractive cost to consumers and/or health payors, this idealized vision offers outer bounds within which a realistic scenario might emerge over time.
What does this say about the traditional, largely independent, categories of health products?
· Diagnostics and Medical devices
· Therapeutics, including pharmaceutical and biologicals
· Information technology
Let’s look at the diagnostics category. Currently, most diagnostic tests are performed at the request of a physician by specialty diagnostic and imaging laboratories. The testing process is expensive and generally performed once or a few times for each patient. The developing wave of new biosensors makes it very likely that the unit costs of tests will decrease dramatically –perhaps by an order of magnitude or more. This will adversely affect the business models of diagnostic testing labs; initially they will compensate by providing more tests, but testing of common parameters with a ‘medical tricorder’ will likely return to the physician’s office, and ultimately to consumers who would benefit from continuous monitoring.
This trend has already started for a few biological parameters such as blood glucose monitoring by diabetics. Pharmacies have replaced testing laboratories in the supply chain and they devote considerable shelf space to competing glucose monitoring devices. In this example, diagnostic laboratories have been largely disintermediated from performing the majority of repeated blood glucose tests. Meanwhile, the automated collection and transfer of data to a cloud repository is a feature of the latest generation of these devices, and continuous monitoring is in development (see for example GlySens ICGM).
In the case of pharmaceuticals, it has become clear that most drugs only work on a subset of patients, and that there may be considerable variations in the effective dose even among those patients. Similarly, ‘safe’ drugs can be harmful to a generally small subset of patients. As a result, drug selection and dosing is increasingly being coupled to diagnostic testing. While initially most such tests will be used to support a drug selection decision, continuous monitoring could offer the option to tune the dose, timing and duration of administration through feedback mechanisms.
We can already see the benefits of feedback and monitoring in a simple example. Pharmaceutical compliance is a significant issue as patients frequently do not following the dosing recommendations, often forget doses, and frequently stop taking the drug too soon. Many of the first mobile health applications were developed to address this issue – reminding patients to take their medication and tracking if and when they have – but without biosensors and administration devices, more sophisticated controls are not yet possible.
Some enabling themes are apparent in these examples:
a. Technological advances in biosensors will be highly disruptive
b. Therapeutic administration will be tightly linked to diagnostic testing and monitoring, and in a growing number of cases will be controlled by medical devices
c. Information technologies will be the glue that enable integrated feedback between measurement and delivery, as well as data collection, consolidation and interpretation
This May 15th MaRS will be holding the first HealthKick forum in Toronto to showcase 45 selected health-product ventures in front of potential investors. There will be three tracks:
1. Therapeutics: Biotechnology and Pharmaceuticals
2. Medical Devices and Diagnostics
3. Healthcare IT
But, I have to wonder if this will be the last time that we can conveniently and unequivocally assign new product developments into these traditional categories.