Ayesha Verrall (University of Otago)
A software system built for screening kiwis for bowel cancer is now powering New Zealand's Covid-19 contact tracing efforts.
Deloitte won a deal to plan and design the National Screening Solution (NSS) in mid 2018 and it was developed on a Salesforce platform.
Today, director general of health Dr Ashley Bloomfield said the Covid-19 contact tracing service was transforming what was a local, manual process into a "national automated system with scale".
According to a report released today by Dr Ayesha Verrall of the University of Otago, the new National Close Contact Service, which traces Covid-19 transmissions, started operations using manual processes on 24 March.
The national contact tracing solution (NCTS) was developed by repurposing the Salesforce system, piloted on the 27 March and used to process all calls from 6 April.
National Bowel Screening Program director Stephanie Chapman said in late 2018 that the technology would be a significant asset for the New Zealand health system, supporting, not only bowel screening, but other population health initiatives and screening programs long-term.
“We need a high-quality product with robust privacy and security features and the adaptability to support other health services in the future.
"I believe investment in achieving such features at this foundational stage will pay off in the long-term."
How right she is proving to be.
However, gaps remain in the system, Verrall also found.
The platform stores case and contact details linked by exposure events, and supports contact management.
It also provides links to existing health information sources, primarily for sourcing contact details and the unique identifier from the National Health Index.
Training in use of the new system for the contact tracing process was completed for all 200 NCCS users on 6 April.
Further development was needed, however, Verrall found, because the crucial measurement for contact tracing success was the time from symptom onset to contact isolation.
Data on this measurement was not available during her audit because the current monitoring system will not provide visibility of the upstream events relating to case management.
"The time taken to be assessed clinically, tested and notified of results should be considered components of a single system, and measured and managed accordingly," she recommended.
"This will require information systems that make some clinical, laboratory and primary health unit (PHU) processes visible at a single point within the Ministry of Health.
"Contacts traced through PHUs, with their various different contact information management systems, will also need to be captured."
Today, minister of health David Clark said the NCTS platform would be further developed to manage the delivery of a Covid-19 vaccine when one becomes available.
The NCCS was an impressive service, especially considering it had been established in just weeks, Verrall said.
However it was also not a suitable nor desirable system for managing all contacts.
"The NCCS also has limited use in certain important situations, such as in the event of a large complex cluster or specific scenarios that require intense involvement of Medical Officers of Health," Verrall wrote.
At present the only centrally visible performance indicators relate to the completion of tracing for contacts referred to the NCCS.
"However, this does not capture the upstream events that impact the timeliness of contact tracing, like case referral processes and testing times.
"Nor does it capture contact tracing activity in public health units."
Measuring performance indicators to drive improvement was an "urgent priority", Verrall wrote, proposing a set of indicators for that purpose.
Verrall also reference the ongoing development of a smartphone app to aid contact tracing -- and that project's possible limitations.
"As it is not yet completed and a number of key aspects are under consideration, it cannot be meaningfully evaluated as part of this audit," she wrote.
"However, it is possible to comment on some aspects of the public health impact of whatever product is developed.
"First, it needs to be available quickly, piloted and continuously improved. Second, potential impact will not be realised unless it is acceptable to a large proportion of the population and enjoys high uptake.
"Less than a fifth of the Singaporean population downloaded the trace together app in ten days, which, assuming random mixing, means only one in 25 exposures will be captured by the app and public health impact will be negligible."
An app would also only produce incremental improvement in the time to isolate contacts as many other steps were involved and needed to be managed.
Finally, an app cannot replace the option to interact with a real person as many contacts will develop illness, have welfare needs, or face issues with accessibility.
