National data analytics platform speeds up research, ensures safety of personal info

SINGAPORE – A local study made the surprising finding that, compared with pregnant women with gestational diabetes, those without gestational diabetes actually incurred higher hospital expenses in the last three months of their pregnancy.

The latter spent more than an average of $1,300 on hospitalisation in the later stages of their pregnancies, possibly for conditions that were not detected earlier.

The research team from A*Star analysed its own study data of the mother-child pairs, as well as data it obtained from national databases through a national analytics platform – the Trusted Research and Real World-Data Utilisation and Sharing Tech (TRUST) platform.

This platform, which ensured the data was anonymised and protected before it was made available to the team, has been used by researchers in some 70 projects since it was implemented in November 2022.

It allows local researchers to draw on anonymised data within a formal governance framework to more quickly conduct studies, rather than repeating the traditional process of recruiting new participants to collect and analyse new data.

The A*Star research team postulated that women with gestational diabetes mellitus (GDM), which develops during pregnancy and may resolve after delivery, typically received closer health monitoring, resulting in other health risks being detected and managed earlier.

On the contrary, mothers who did not have GDM and received standard antenatal screening may develop complications in the later stages of pregnancy, such as hypertension or delivery issues, which may require hospital care.

To validate this hypothesis, the A*Star team is now tapping the TRUST platform again to analyse more data from the National Electronic Health Record system. This new phase of the study is still ongoing.

Earlier, the research team, led by Dr Mukkesh Kumar, head of the data management platform at A*Star Bioinformatics Institute, and Dr Evelyn Loo, a principal scientist at the A*Star Institute for Human Development and Potential, had analysed the data of 1,166 mother-child pairs through the TRUST platform.

These were from the Growing Up in Singapore Towards healthy Outcomes (GUSTO) cohort study – Singapore’s first and most comprehensive longitudinal birth cohort study, which has followed more than 1,000 mother-child pairs since 2009, and collects extensive data on pregnancy, genetics, nutrition, growth, brain development and well-being.

The A*Star team said its work is likely to be Singapore’s first such study on GDM-related healthcare expenditure, and the findings suggest that women without GDM may also need closer monitoring during late pregnancy.

Other findings of the study were less unexpected.

For instance, the team found that mothers with GDM had higher healthcare expenditure for hospital, polyclinic and emergency department visits in the first year after delivery while they were followed up to ensure they regained normal insulin functions. They spent an average of $1,400 more than mothers who did not have GDM, before government subsidies.

Dr Kumar said future findings can aid health planners to monitor cost burden and disparities among different population groups.

TRUST was co-developed by the Ministry of Health (MOH), Ministry of Digital Development and Information, Government Technology Agency and the national healthtech agency Synapxe.

As at the third quarter of 2025, it has 13 partners, including all three healthcare clusters, five autonomous universities and research institutions such as A*Star, Duke-NUS Medical School and Singapore-ETH Centre.

The 48 datasets currently available through TRUST include population data such as socio-economic, birth and death data; clinical data; lifestyle data such as those collected through wearable devices; chronic disease screening data; and genomic data such as those collected through Singapore’s

.

The TRUST platform has safeguards to ensure data security and individual privacy, aligned with the internationally recognised “Five Safes” governance framework.

They include the Data Access Committee that evaluates requests from researchers and will only approve those that fulfil public interest, have scientific or clinical value, and can bring benefits to society.

The committee is chaired by Mr Philip Ong, deputy secretary (development) at MOH, with representatives from the healthcare clusters, research institutes, ethical and legal experts, and two members of the public from patient advocacy groups.

After the request is approved, data will be made available to researchers who have been endorsed by their institution and cleared to be granted rights to access the data for analysis.

Any breaches of privacy or misuse of the data will be penalised.

Before researchers can use the data, all personal details are removed and pseudonyms are used to ensure individuals cannot be identified.

Data is not permanently stored on the platform and is available for only a limited period for the researchers to complete their analysis.

There are also safeguards against unauthorised data disclosure or export, including the use of a virtual desktop interface with internet separation and copy-paste functions disabled.

Activity on the TRUST platform is also logged and monitored for any irregular activities, such as suspicious login attempts.

The MOH Office for Healthcare Transformation has set up a TRUST Office to oversee the running and continued development of TRUST.

Ms Koh Mingshi, director of the TRUST Office, told The Straits Times that TRUST will continue to develop and provide new datasets, functions and tools, such as data related to climate change and urban resiliency, which can impact one’s health.

“Evidence-based decision-making is applicable not only in the health sector. Also, the determinants of a person’s health go far beyond health itself,” she said.

Ensuring data consistency so that it is easier for different researchers, including those from different domains and even internationally, to utilise and analyse data stored in the same format is also a key area to work on.

The TRUST Office is working with its partners to adopt the Observational Medical Outcomes Partnership model and data standards.

The development team is also looking into ways to make processed datasets that are frequently used to be more readily available on the platform.

Referring to how research projects like GUSTO can go beyond its initial research scope to look at areas such as unhealthy eating habits in children, Ms Koh said: “This is the kind of thing we want to do, which is to ensure whatever investments we bring in, it is not just used within that area, but we have the ability to use data in a smarter and safer way, and a more secure manner, which allows us to do a lot more.”