Week in Review

Border breach uncertainty traced to IT oversight

Analysis: A simple IT oversight leading to days of uncertainty over how many people left managed isolation without a test falls into a neat pattern of Ministry of Health technology failures, Marc Daalder reports

Just one question on a single form would have saved Ashley Bloomfield a week of headaches.

For days now, the Director-General of Health has undergone intense questioning from reporters and Opposition MPs wanting to know how many people left managed isolation without receiving a Covid-19 test, even though regulations Bloomfield himself had argued for meant they were supposed to return a negative result before returning to the community.

You'd think it would be simple, but on Monday Bloomfield let slip why the innocuous statistic has so evaded Ministry of Health staff: It was an IT oversight.

NHI the key

There are two databases, one containing information about people in managed isolation, such as their date of arrival, the country they came from and when they departed or are scheduled to depart. The other database contains medical information, the most important points of which are whether people were tested and what the test results were.

The problem? The two databases can't talk to each other.

When people were tested for Covid-19, Bloomfield said, their National Health Index (NHI) number wasn't collected. That number, which is used to coordinate different medical databases and is a unique health identifier for every New Zealander, would have been crucial to aligning the two databases.

In the absence of the NHI number, Ministry of Health staff have had to personally compare different entries in each database one-by-one, making sure that names, dates of birth and other identifying information match up in order to ensure the two entries are for the same person. This has meant that, until more than a week after the two cases from the UK were first reported, officials were unable to disclose how many of the 55 people released from managed isolation early on compassionate exemptions after June 9 received a test beforehand.

In the end, that number was shockingly low - just four of the 55 had been tested before leaving, and two of them didn't wait for their result before they were let out. Most of these people have since been tested, but Bloomfield said Wednesday that a handful won't be tested for medical reasons, because they are children, or are resisting doing so because they believe they have already had a test.

Normally, the Ministry of Health would be able to authoritatively prove that this last handful hadn't been tested, but the botched IT systems cast doubt on their numbers. Bloomfield said NHI numbers are now being collected.

It seems plausible that the rest of the numbers, which will take far longer, could stack up similarly. Between June 9 and 16, when people were supposed to be tested before leaving but often weren't, 2159 people left managed isolation, Bloomfield has said, including some of the above 55.

Of these, 800 received a test before leaving. Another 568 didn't receive a test before leaving - 210 of whom have since been tested and returned a negative result - 239 have a result pending and 119 won't be tested for medical reasons, because they are no longer in the country, or because they have refused testing. Unless Bloomfield is convinced the person poses a high risk to public health, he can't force people who have left managed isolation to take a test.

Alongside the 800 who received a test before leaving and the 568 who haven't, there are still 791 people who officials are trying to track down in person or match up between databases.

On the face of it, this should be an easy statistic to provide. But by leaving the NHI number off the testing form, the Government has kneecapped its own response.

Ministry plagued by tech problems

This isn't the first time the country's health response has been hampered by poor IT systems. For the first few months of the pandemic, Public Health Units (PHUs) which did contact tracing had no way of cross-referencing their work with that of other PHUs. If someone travelled from Wellington to Auckland and then tested positive, the two PHUs involved would have a mare of a time coordinating the contact tracing.

On March 27, three days into lockdown, the Government did the obvious thing and created a centralised database. But for more than a month thereafter, PHUs were unable to view the centralised database, which was governed by a specialised contact tracing unit at the Ministry's headquarters in Wellington.

In her audit of New Zealand's contact tracing system, epidemiologist and now-Labour candidate Ayesha Verrall wrote PHUs "were cautious about diverting contact tracing to the [centralised] NCCS in many situations, because once they did they lost visibility of the outcome for the contact". It took Verrall's report for the Government to allow PHUs to see the centralised database.

As early as Monday, three PHUs - Auckland's, Wellington's and Christchurch's - still weren't on the centralised database. As of Wednesday morning, just Auckland was the outlier.

The other tech problem the Ministry of Health has faced is more systemic - their database is severely outdated. 

The Privacy Impact Statement for the Government's contact tracing app found the Ministry of Health databases - the NHI and the National Enrolment Service (NES) - often have incorrect or outdated information.

"Consumers who have changed their contact details since they were last updated in the NHI or NES services, or who may not be residing at their usual address. The impact of this is that contact tracers may find it more difficult to contact the person concerned, delaying the process of managing their self-isolation, and close contact identification, or they may not be able to locate them at all," it states.

"If the infected person or a close contact has independently submitted their contact information via the [app] then these contact details could also be accessed by a contact tracer if they had not been able to locate them promptly via the standard processes."

Andrew Chen, a research fellow at the University of Auckland's Koi Tū - Centre for Informed Futures, theorised on the app's release that gathering updated contact information could be a prime reason for the app's release.

"Getting up-to-date contact details for people is actually huge. If they can actually do that for everybody, then that helps the contact tracers waste less time," he said.

"To some extent it feels like that's really what they're trying to do and the QR code functionality is to encourage people to download the app and give their details."

The app itself has been plagued by problems, though these result more from delays and poor communications than bugs in the programme itself. The two-month wait for a contact tracing app meant that a plethora of QR-based private sector apps had emerged by the time NZ COVID Tracer debuted. This has led to people scanning the wrong QR codes, as the app can only be used with official Government ones. Recent de-emphasising of the app - and contact tracing more generally - has meant the average user has scanned less than two QR codes in the past month.

Help us create a sustainable future for independent local journalism

As New Zealand moves from crisis to recovery mode the need to support local industry has been brought into sharp relief.

As our journalists work to ask the hard questions about our recovery, we also look to you, our readers for support. Reader donations are critical to what we do. If you can help us, please click the button to ensure we can continue to provide quality independent journalism you can trust.

With thanks to our partners