NIHI: Using BigQuery, Data Studio, and Firebase to support mHealth research

About National Institute for Health Innovation, The University of Auckland

The National Institute for Health Innovation is part of the University of Auckland and Auckland UniServices Limited. NIHI measures its success through its contribution to improving people’s health in New Zealand and around the globe. The organization discovers, develops, tests and delivers innovative approaches to today’s most pressing health problems. Its work is focused on preventing disease, improving people’s health, reducing health inequities, and enabling the delivery of more effective and equitable healthcare.

Industries: Healthcare
Location: New Zealand

Tell us your challenge. We're here to help.

Contact us

With the use of Google Cloud and Firebase, the National Institute for Health Innovation is providing innovative health and wellbeing solutions through initiatives such as Ol@Or@, a culturally designed mHealth program co-designed and delivered in a collaborative partnership with Maori and Pasifika leaders and communities.

Google Cloud results

  • Automates previously manual report creation processes
  • Improves measurement of the effectiveness of mHealth research
  • Enables use of ML and AI to help with new innovations to reach populations

Helps the organization collaborate more effectively with communities and research partners to deliver high-quality research

A multi-disciplinary research unit based in the University of Auckland, New Zealand, the National Institute for Health Innovation (NIHI) aims to discover innovative solutions to the world’s leading health problems.

“We look at ideas, strategies, implementations, and interventions that can impact the health and wellbeing of people and populations,” says Gayl Humphrey, Research Fellow, Co-Lead Health Informatics and Technology Programme, NIHI.

NIHI has operated for 30 years and produced high-quality research that has been published in leading scientific journals such as the BMJ, The Lancet and The New England Journal of Medicine.

Humphrey works in an in-house team of about 45 people, ranging from professors to research assistants and developers, programmers and data managers.

“We are a diverse group with research interests and expertise in areas of cardiovascular health; food and nutrition policy security, sustainability, and labeling; the field of addiction including drugs, alcohol, and gambling; mental health; and health systems and behavioral change,” she says.

A focus on equity

A key priority for NIHI is health equity and ensuring the unit’s interventions improve the health and wellbeing of populations with high health needs. “For example, there is a worldwide obesity epidemic, and the impact of obesity does not impact equally across society in New Zealand or anywhere else, so we’re conscious of developing innovative interventions with equity front and center,” says Humphrey.

“This includes developing interventions based on ‘mHealth’—medicine and public health enabled by mobile apps and devices—that address issues of digital exclusion, such as digital and health literacy and digital access.”

The team at NIHI explores innovative solutions to support specific populations as part of its mHealth program. For example, NIHI has undertaken a feasibility study into using a chatbot to help patients find answers to simple questions they may have after being discharged from hospital.

The organization has also completed a study into whether a mobile app could help parents understand child growth charts and improve digital literacy around growth trajectories to help prevent childhood obesity. In addition, it has embedded a cognitive behavioral therapy program for people experiencing gambling problems or harm into a mobile app that uses elements of gamification.

NIHI office

Ability to drill deeper and understand more

NIHI relies on data, analysis, reporting, and visualization to help understand the impact of digital and mHealth programs.

The organization had previously relied on a labor-intensive, manual system that was effective for measuring research outcomes but was less able to deliver the granular insights it needed from the digital domain.

“We were limited in what we could routinely track digitally, and this made evaluating the impact of engagement with the mHealth program challenging,” says Humphrey. “We wanted to drill much deeper than the number of downloads of a mobile app, geographic distribution of the app users, and the operating system and phone type used.

“We wanted to measure the time, frequency, and intensity of use of the mHealth tool in much more detail, including variables such as repeat visits, multiple views, and the actions people undertook.

“In addition, we wanted to cross-match and review this information with our more traditional modes of collecting research data to see if we could understand more about aspects such as frequency of use and effectiveness of the dose-response of a particular program.”

In addition, Humphrey says, blending these mechanisms could help NIHI identify any gaps in its interventions it needed to understand.

“Ultimately we wanted to see in more detail where and how our interventions were having an impact and whether that impact was total or partial, or for one group or another.

“Plus, we wanted to be agile and iterative in our research and in our program design, so understanding these nuances early would enable a more pragmatic approach to our research.”

Improving partnerships with improved analytics

NIHI works closely with partner organizations and communities to shape, design, and implement interventions. This means the organization takes a collaborative approach to designing and generating insights into the outputs needed for project management.

“The impact of the research is important, but so is the research process,” says Humphrey. “In mHealth, capturing and visualizing administration data such as recruitment numbers, utilization of the mHealth intervention, and distribution of the intervention uptake is important.

“That was a particularly time-consuming, continuous process, particularly for projects involving multiple partners and when the research was fully digitalized in the mHealth domain.” NIHI often had to run different reports and present information in different ways to meet organizations’ specific needs.

Already a committed user of Google Analytics, NIHI turned to Google Cloud for the products and services that could deliver the deeper insights required. While the organization conducted only an informal review of the options available, Google Cloud’s treatment of data gave it confidence in its selection.

“We are very mindful of our responsibilities as researchers to our study participants, so we need to understand how information is handled, how we control access, and the importance of ownership,” explains Humphrey.

“We were reassured by the fact Google Cloud’s terms and conditions supported some of the privacy changes happening in the New Zealand health sector at the time, so there was a natural synergy for NIHI to adopt these tools.”

For example, NIHI had been collecting and storing all research data on servers in New Zealand. However, the increased availability of digital tools and the desire of organizations to understand usage around frequency and content views was prompting the use of solutions that gathered and presented data, without that data actually being hosted by the research team.

“The current review of privacy legislation in New Zealand aims to explore digital data more explicitly and offer requirements and guidance,” says Humphrey. “Because issues such as security and access are key considerations for health and research data, Google Cloud’s high standards in this area delivered excellent synergies and made adoption straightforward.”

NIHI elected to adopt the BigQuery data warehouse, Google Data Studio visualization service and the Firebase mobile and web application development platform to meet its requirements.

The organization saw that Firebase and Google Cloud services opened up opportunities to operate more efficiently and explore new mHealth interventions, such as combining apps with sensors and text messages to deliver multifaceted interventions. “For Firebase alone, we wanted to use features and services such as Firebase Cloud Messaging and Firebase Crashlytics,” says Manoj Alwis, Mobile Application Developer, NIHI.

“Firebase Cloud Messaging now allows us to do some ‘out of system’ push notifications, which is extremely valuable for us as researchers. When something new is happening or something different is needed, we can communicate to end users without having to incur, for example, the costs of an SMS message. Some of our studies involve thousands of people, so this ability to tailor a message outside of the planned programme is extremely important.”

Gayl Humphrey, Research Fellow, Co-Lead Health Informatics and Technology Programme, National Institute for Health Innovation

A range of benefits

The organization completed the move to Google Cloud and Firebase for its mHealth studies in early 2019 and is achieving a range of benefits from the project.

“Firebase Cloud Messaging now allows us to do some ‘out of system’ push notifications, which is extremely valuable for us as researchers,” says Humphrey. “When something new is happening, or something different is needed, we can communicate to end users quickly and seamlessly without having to consider the costs of, for example, sending an SMS message. Some of our studies involve thousands of people, so this ability is extremely important.”

Meanwhile, Firebase Crashlytics enables NIHI to quickly track and correct any issues with app performance, allowing the organization to respond to problems before they become a significant issue for its research.

The new architecture allows the organization to create clear and well-presented reports more easily and implement a reporting framework to enable filtering of variables, without undertaking extensive manual processes

“Our partners can also view dashboards and filter the data and reports to meet their needs, which may be different from our research-oriented focus,” says Humphrey. “For example, we use the SCRIPT app to support digital decision-making for antibiotic prescribing. The app is based on current evidence and enables prescribers to work through simple decision trees to determine the best first-line antibiotic for adults or children.”

SCRIPT is used by a range of groups, including doctors, pharmacists, and infectious disease teams. Therefore, Humphrey says, NIHI needs a versatile, automated reporting capability.

Once the organization builds a reporting dashboard, the research team can easily filter, analyze, and review data. This reduces the time for data analysts to create bespoke reports, releasing them to focus on complex research data analyses.

“Through BigQuery and Data Studio, we could demonstrate whether the promotion campaign for the cellulitis pathway was effective by visually presenting information such as how many people viewed the pathway, the roles those people had, and the uptake by different service areas such as surgery or general medicine.”

Gayl Humphrey, Research Fellow, Co-Lead Health Informatics and Technology Programme, National Institute for Health Innovation

Routine reporting is also important for NIHI’s partners, enabling them to view data in ways relevant to them and their communities—including showing complex outputs in easily digestible visual formats.

“By using Data Studio to present data simply, our partners can deliver strong, compelling messages to the communities they work with,” adds Humphrey.

NIHI also used BigQuery and Data Studio to help evaluate a new treatment pathway for cellulitis, a bacterial skin infection, for inclusion in the SCRIPT app.

“Through BigQuery and Data Studio, we could demonstrate whether the promotion campaign for the cellulitis pathway was effective by visually presenting information such as how many people viewed the pathway, the roles those people had, and the uptake by different service areas such as surgery or general medicine,” says Humphrey. “This enabled the hospital team to determine what additional support was needed, and where and on whom to focus their efforts. This insight helped maximize resources and minimize waste.”

“We are very interested in insights that could help us to learn how to tailor and adapt interventions to be delivered just when people need them—we call it the ‘just-in-time’ principle. Google Cloud, as a suite of tools, has enormous potential to help us be creative in these areas.”

Gayl Humphrey, Research Fellow, Co-Lead Health Informatics and Technology Programme, National Institute for Health Innovation

AI and ML open new possibilities

With Firebase and Google Cloud providing powerful support to NIHI, the organization is turning to AI and machine learning services to explore new mHealth intervention mechanisms.

“We are exploring using Dialogflow as part of a new chatbot concept for one of our research studies’ says Alwis.

NIHI is also evaluating the potential of TensorFlow to build machine learning models and apply machine learning to data obtained from wearable sensors and geolocation devices to help improve healthcare outcomes.

“We are very interested in insights that could help us to learn how to tailor and adapt interventions to be delivered just when people need them—we call it the ‘just-in-time’ principle,” says Humphrey. “If we can deliver support just-in-time, we may prevent someone having a cigarette or entering that gambling zone, by providing tools and strategies to use right then to manage that urge or support an action.

“Google Cloud, as a suite of tools, has enormous potential to help us be creative in these areas,” she concludes. “We are looking forward to continuing to work with the organization to realize our ambitions.”

Tell us your challenge. We're here to help.

Contact us

About National Institute for Health Innovation, The University of Auckland

The National Institute for Health Innovation is part of the University of Auckland and Auckland UniServices Limited. NIHI measures its success through its contribution to improving people’s health in New Zealand and around the globe. The organization discovers, develops, tests and delivers innovative approaches to today’s most pressing health problems. Its work is focused on preventing disease, improving people’s health, reducing health inequities, and enabling the delivery of more effective and equitable healthcare.

Industries: Healthcare
Location: New Zealand