Skip to content

Poor data driving decisions – but in 2021 this is not acceptable, especially in a small country such as NZ

22/04/2020

Derek Fisher at NZ Herald

How can we trust government decisions on the crisis when we read these sort of comments

Key data guiding critical decision-making on Covid19 is filled with gaps and even inaccuracies, with detailed information held only at a local level by regional health units.

It has also emerged there is no central database recording the length of time between a test being taken and the return of a result.

It means the daily updates of New Zealand’s coronavirus cases could be days old, with people who have been tested telling the Herald they waited almost a week to get results.

In 2021, in a small country with a centralised MOH we have this nonsenss. Shameful.

The Herald has been questioning the gaps in information for two weeks and it was only on Friday the Ministry of Health (MoH) responded, explaining data collection was handled by the 12 regional Public Health Units.

This is an explanation, but it is not a valid excuse. What is the army of analysts doing in Molesworth Street. They are always calling for data, to compile reports.

Why after all these years is there not one computer system. Hell there are hospital groups in the USA which are bigger than the various DHBs combined.

Stunning incompetence by successive governments for decades. Labour and National are both to blame.

This next comment, I accept the broad comment, but take exception to much of the substance. Let us consider the huge variance in scale, 4 countries, 2 legal systems and a massively larger health service and population. So, not convinced regarding NZ.

Data management has been an issue in other countries wrestling with the virus with concern in the United Kingdom over inconsistent data and warnings gaps in information may undermine faith in the Government’s response.

Of course gaps will cause concern.

The “current cases” list published on the MoH website records the date a confirmed or probable case of Covid-19 has been logged, the gender and age group of the individual with the disease, the DHB in which they are located and details of their international travel.

Of the 950 confirmed and probable cases recorded as of Saturday, there are 146 that carry no information as to whether there was international travel. On previous days, a number have been listed as “unknown” while others were left blank.

It’s not just new cases where information has yet to be obtained. Those without information on whether they travelled overseas — the main route for infection here — include two cases from February 28.

Of the 448 cases where it is known there was international travel, only 420 list the country the individual was in before returning here.

Other information missing included flight details — just 302 cases of the 448 involving international travel include a flight number.

However, many of the flight numbers listed are not the international flights to New Zealand but only domestic flights taken after arriving in the country.

This is just not acceptable. We are frequently told by the regime that we have a tough border control. This does not seem to be the case.

The article cites some examples, with details

Examples include a man aged 20-29 who flew back to New Zealand from Indonesia on March 19 whose only flight details are his journey from Auckland to Palmerston North. In that case, the man’s departure date for his journey is listed as April 19 2020 — two weeks in the future. His case is dated April 1.

Other examples include a woman aged 40-49 who returned to New Zealand from Egypt on March 9. Her only flight details are her onward journey from Auckland to New Plymouth.

Some flight numbers — such as Beka76 — don’t appear to match flight designation numbers for any airline. This was the flight listed alongside a Waikato man aged 50-59 who flew back from United Arab Emirates in mid-March.

We had some attempts at explaining the issues from the government

Director of Public Health Dr Caroline McElnay said the MoH information came from the National Notifiable Diseases database, EpiSurv, that held information about confirmed and probable cases of Covid-19, and those under investigation.

It was updated by public health staff in regional Public Health Units in line with notifiable disease law, she said. “Detailed reporting information can sometimes lag as we need to ensure the accuracy of detail around individual cases.

“Information reported in our stand-ups is current — confirmed cases are required to be notified to the Medical Officer of Health at the Public Health Unit and are typically notified with 24 hours of confirmatory testing.

“Details such as flight numbers may be known by the public health units, but not input into EpiSurv.”

A MoH spokeswoman, in an earlier statement, said: “Public Health Units may store other information elsewhere for case management, which is why there may be missing data.”

She said the “extensive case interviewing process” can involve more than one interview and could lead to delays inputting information.

OK, but this is not really satisfactory

Indeed a recent audit report of which more in another post has found major shortcoings in NZ contact tracing

The spokeswoman confirmed it did not hold data about the period of time between when an individual was tested and when the test was returned as positive.

“This is managed at a local level.” The Herald has interviewed people who have waited as long as a week for a test result to come back, although it is understood lab work is prioritised towards those most likely to produce a positive result.

It means it is unknown whether the numbers of confirmed and probable cases announced at the daily 1pm emergency briefings are from tests taken inside the last 24 hours or earlier.

Again this just seems not good enough. Other jusridictions appear to perform much better.

The Ministry of Health has listed 81 per cent of cases as linked to international travel or contact with someone who has travelled and 17 per cent as “under investigation”. Only 1 per cent of cases is listed as through community transmission.

Professor Shaun Hendy, director of University of Auckland-based research centre Te Pu¯naha Matatini, said the quality and timeliness of data had been an issue for his team in its work statistically modelling the behaviour and direction of the virus for the Government.

He said the modelling work was intended to give health and political leaders information to assist planning. Good-quality, fast data was needed to inform that, he said. He said it improved last week. “A system that wasn’t really designed to run at this speed is under stress,” he said. “It’s a big lesson for the health sector in how it manages data.”

“We ought to be seeing better data coming through soon. It will give our political leaders better information to make their decisions and for our health system to work more efficiently.”

So data has been deficient, important decisions affecting lives, livlihoods and the future of NZ are being made with data that is just not good enough.

Does not fill one with confidence

Comments are closed.

%d bloggers like this: