About 1 in 20 Google searches are health-related and some 72 percent of internet users search online for health information, according to research from John Hopkins University in collaboration with Google.
That means devices people use to access the internet for self-diagnoses have become powerful data capturing tools as well as sources of engagement and empowerment for patients.
The entrance of Google specifically, but other tech firms more generally, into the healthcare market has and will continue to cause disruption. While data capture, processing and analysis are the tech giant’s unique selling point at this stage, it has plans to extend its reach into delivering healthcare and health insurance.
This post explores what Google’s entrance into the healthcare market means for clinical trials and how the tech company has already impacted the market.
Amazon, Apple and Alphabet Enter Healthcare
According to The Economist, the three big companies, Amazon, Apple and Google’s parent company Alphabet, are all getting involved in healthcare-related products and services provision.
Amazon: The company has partnered with Berkshire Hathaway and JPMorgan Chase to form a not-for-profit healthcare company that will offer its employees cheaper care than that offered by health insurers by selling drugs online.
Over the past three years, the company has been honing its devices and software to harness medical data and to develop clinical care products.
Apple: The organization is updating iPhone software with a health records feature so users can share and manage their personal medical records.
Alphabet: In addition to DeepMind, Google’s parent has US-based Cityblock Health and Verily, the last run by Andrew Conrad, Ph.D., who cofounded the National Genetics Institute.
The parent company’s development of AI has equipped it with predictive powers to determine the deaths of “hospitalised patients two days earlier than current methods,” according to the article.
It’s clear that this access to data and ability to analyze disease trends and prognoses could be powerful for clinical researchers working on optimal treatments.
Medical Market Disruptions
The Economist notes that many industries could be impacted by these large organizations entering the medical market. Currently, in the US, medical insurers UnitedHealth Group and CVS Health earned $185 billion and $178 billion respectively in 2016, which is higher than any tech firm besides Apple.
However, the article reports that the insurers lost 4 percent in their share price when Amazon announced its health venture.
The way treatment is carried out will also change. Cityblock Health will collect and analyze data to determine the areas most in need of healthcare but lacking infrastructure. It will then send healthcare practitioners to the poorest homes to treat patients and this will be covered by Medicaid.
The implications for clinical trials is huge. This technology could help to link trial recruiters with potential patients, which would result in clinical care being delivered to those most in need.
Making Predictive Analysis More Accurate
With all the data-processing power Google possesses, it’s no wonder that predictive analysis is where it can make a big impact.
In a study conducted by researchers from Google and the Universities of California San Francisco, Stanford and Chicago Medicine, the results showed how analyses of electronic health records (EHRs) could be used to improve predictive analysis of patients.
Kate Monica at EHR Intelligence writes that the EHR, analyzed using Fast Healthcare Interoperability Resources (FHIR), provided highly accurate analytics.
Datasets, which had been encrypted, included “demographics, provider orders, diagnoses, procedures, medications, lab values, vital signs, and flowsheet data,” Monica writes, resulting in data about 216,221 hospitalizations among 114,003 patients.
Instead of using custom datasets, the researchers were able to use “a single data structure” to predict health outcomes instead of requiring custom datasets for each new prediction.
This deep learning approach uses machine learning methods to make sense of data representations rather than task-specific algorithms. Through deep learning, the researchers managed to produce accurate predictions “ranging from mortality rates to readmissions, as well as length of stay and diagnoses,” Monica explains.
DeepMind Health’s Medical Ambitions
The team at DeepMind Health aims to apply the tech power of Google to healthcare, chiefly by providing more accurate analyses based on machine learning grappling with patient data and ensuring faster and more personalized treatment.
The Streams app, for example, which alerts healthcare providers of a patient’s deterioration, is saving nurses two hours each day, DeepMind reports.
While the focus is on equipping healthcare practitioners with tools to treat patients, DeepMind says its ambitions include “empowering patients to look after themselves and their families’ health, and supporting coordinated ongoing care around patients’ needs.”
With regard to data security, the company says all of the sensitive data is encrypted and is not and will not be linked “to Google accounts or services, or used for any commercial purposes like advertising or insurance.”
However, not everyone has been supportive of the company’s tactics, particularly in terms of the data privacy of patients.
DeepMind had been working with the UK’s National Health Service (NHS) to develop software to make staff aware of patients at risk of kidney failure. However, the Information Commissioner’s Office decreed the agreement between the company and the NHS Trust unlawful. This was due to a failure to gain patient consent to use their data, journalist Nicole Kobie writes.
DeepMind received the healthcare data of around 1.6 million patients from hospitals run by a London-based NHS trust. Data included details of HIV-positive statuses, drug overdoses and abortions.
In response to the lack of patient consent issue, a spokesperson from one of the hospitals involved said the data had been encrypted as per standard industry practice.
Tech-Enabled Collaboration in the Cloud
Imagine the possibilities of healthcare providers being able to collaborate to help patients. This is what Google Cloud’s VP of Healthcare, Gregory J. Moore, M.D., Ph.D., says is possible.
To achieve this, data needs to flow better, which will ultimately enable developments in AI and machine learning to deliver more accurate insights into patients’ health and treat them more effectively.
Google’s Cloud Healthcare API, being trialled by Stanford School of Medicine, “provides a robust, scalable infrastructure solution to ingest and manage key healthcare data types—including HL7, FHIR and DICOM—and lets our customers use that data for analytics and machine learning in the cloud,” Moore says.
The aim is to “simplify data interoperability” for healthcare enterprises, he adds. Examples include:
- Apigee combines FHIR (Fast Healthcare Interoperability Resources) APIs with healthcare companies’ existing EHRs.
- BrightInsight uses data and software to optimize drugs, medical devices or combination products for pharmaceutical and medtech companies.
- Wuxi NextCODE is a “scalable genomics database management system” that will be accessible for clinical and research applications via the cloud.
- Not-for-profits Sutter Health and Augmedix, via the Google Cloud Platform, enables collaboration between patients and healthcare providers. Using Glass Enterprise Edition and Augmedix, health practitioners can access patient data as customized remote scribes. The result is less time behind a computer accessing EHRs and more time with patients.
Again, for clinical trials, this would be able to provide trial staff with plenty of patient and trial data as they interact with patients.
Another of Alphabet’s health companies, Verily Life Sciences, is collaborating on a clinical trial partnership with Duke University School of Medicine and Stanford Medicine. Bill Siwicki at Healthcare IT News reports that the company will gather phenotypic health data from approximately 10,000 participants over four years.
Dubbed Project Baseline, the study will be hosted on Google Cloud Platform. Its goal is to develop a baseline for health and data that can help reveal how people become sick and diseases develop.
Clinical trial patients at each trial site will have their data — clinical, imaging, self-reported, physical, environmental, behavioral, sensor, molecular, genetic, blood and saliva — collected via wearable devices and sensors, along with surveys and polls via smartphones.
Google’s Healthcare Strategy Emphasis on Machine Learning
Across Alphabet’s healthcare companies, the focus is on using data to improve healthcare, CB Insights reports.
In particular, the tech company’s strategy involves data generation through wearables, imaging and MRIs; disease detection through AI analysis of the data for any anomalies; and disease/lifestyle management through developing tools to help people manage their conditions.
Worry Over Tech Companies Not Playing By the Rules
Connected to data privacy concerns, which exist whenever tech companies branch into a new industry, is whether they will abide by the industry regulations and standards.
DeepMind launched its Streams app before seeking regulatory authority, technology writer Hal Hodson at New Scientist reports. Google’s response to questions of its ethics in so doing is that it has right to access the data as it is working with hospitals to provide “direct care” to patients, he adds.
This seems to fit the definition of direct care set out by the UK’s Caldicott guidelines for handling data: “a clinical, social or public health activity concerned with the prevention, investigation and treatment of illness and the alleviation of suffering of individuals.”
Data security and playing by the rules are essential particularly in the clinical research space as regulations are in place to ensure patient safety. Nevertheless, Google’s entry into the healthcare market along with other tech companies suggests disruptions to the medical market. This must surely encourage innovation from existing players.
Additionally, better data processing, predictive analysis and patient engagement will always be pursued points of progress in clinical research.