Like many Australians in lockdown during the Covid lockdowns, Xin Yin Ooi found herself “obsessed” with weekly reports from health authorities, looking at case numbers, vaccinations, hospitalizations and deaths.
As a data analyst, the Sydney woman was naturally interested in the statistical breakdowns provided in the NSW weekly health report, so when she noticed an odd figure in the table, she decided to ask for the underlying numbers to verify the work for herself.
Maybe alone? The data had been “deleted”.
“It was great,” said Ms. Oh
“In the early days of Omicron, our prime minister, health minister, chief health officer, I remember them vividly every day in a press conference saying, ‘The data shows that two doses are not enough, it takes three to treat with Omicron’. They said, ‘the data, data’; it’s unbelievable that they deleted the data.”
Earlier this month, the New Wales Civil and Administrative Tribunal (NCAT) ruled in favor of NSW Health in a freedom of information battle with Ms Ooi, who filed a public information request (GIPA) for the first time. last year
‘long way’
Ms Ooi had asked NSW Health to provide “de-identified patient level information” supporting the two notices issued in the first week of February 2022.
Table five in that report was a breakdown of hospitalizations, ICU admissions and deaths between November 26, 2021, and February 5, 2022, sorted by vaccination status: one, two or three or more “effective doses”, not “effective”. dose” or unknown.
Table six was the “proportion of cases with a serious outcome” of the ICU or death, destroyed by age and vaccination, but only shows “three or more effective doses”, “effective doses” or “less than two effective doses”. «. “.
Ms. Ooi also asked for some “other relevant information that could help me understand this data”, saying in March 2022 that “the first goal of my analysis is to combine the last two columns in the table into six circles. [a] the most elegant type of vaccination”.
“This table is actually very important – it’s investigating the causes when Omicron comes to NSW, a new virus at the time, everyone knows what it’s doing,” he said.
“What’s different is that you see, from no weight to one dose, is actually an increase” [in hospitalisation] from 1 percent to 2.7 percent if you look at that column. I am a data analyst by profession, I wanted to know why, so I did GIPA. I began the long journey of pursuit.
NCAT After these hearings
NSW Health denied the request the following month, saying it did not have the requested information. Ms Ooi applied for a review by the Information Commissioner, who accepted the NSW Health decision in June.
The matter then came to NCAT for an administrative review in December.
Ooi published last week more broadly Ms Twitter thread outlining their concerns following the NCAT decision in favor of Health NSW.
“In 2021, our government told us that Covid-19 is an unprecedented pandemic that requires an unprecedented solution that affects the entire population with a few exceptions: the elderly, the young, the middle-aged, everyone, with the Covid-19 vaccine. I just got provisional approval from the TGA,” he wrote.
“They told us this is the way out of the pandemic. That vaccine ‘stops the expansion’. Later, our government said that if they did not stop the spread of vaccination, we would be less likely to be hospitalized, end up in the ICU, or die. Logically, NSW Health, which led our response to Covid, is very interested (and obliged) to verify the effectiveness of the vaccination program, which achieves its intended purpose.
At the hearing, in which Ms Ooi represented herself, Dr Christine Selvey from the NSW influenza health’s Covid branch explained how the weekly reports were drawn up so that there was no “monitoring media” behind which the records were made.
‘complex process’
Dr Selvey told the court that the former Public Health Response Branch (PHRB), which produced previous departmental restructuring reports in March 2022, drew data from three distinct sources and drew a link: the Conditions Information Management System (NCIMS); The Australian Immunization Register (AIR) and Flow Portal, a live database of hospital patients in the NSW health system.
“As Dr. Selvey explained, the process of extracting data from each source and linking the total data was a very complex process and took several hours,” said NCAT Senior Member Elizabeth Bishop SC, your council.
“Once the data was extracted, the analyst ran the code to create tables with the results that were then reviewed by epidemiologists. This process took several hours more.”
NSW Vesalius staff used R, a statistical programming language, to perform the analysis required to produce the tables.
“A temporary data (dataframe) is created that is kept inside the machine’s RAM and exists only in the R session,” Ms.
“No intermediate tables are created and only necessary events are produced, after which the session ends and the data table is cleared.”
Dr Selvey told the court that no records had been kept in the meantime because “the purpose of the analysis was to release a snapshot at a certain time reflected in published records” and “he was known to keep patterns”.
“Furthermore, if further analysis is required at a later date, more recent information will be provided… at that time,” he said.
Privacy Policy
He added that “access to AIR data is only permitted for very limited purposes and to be retained for no more than the prescribed purpose”, and that “the associated patient information is extremely confidential and the retention of those records would result in confidentiality”. cares”.
GIPA’s application was refused because NSW Health argued that to provide the requested information, it would have to go through the same complicated process again and “create a new record”, which is outside the scope of the legislation.
Ms Bishop said she was “satisfied on the balance of probabilities” that NSW Health “did not retain the patient-level data tables five and six, and therefore did not have the information requested by the applicant”, and also found that “the central data tables are not being held (in any way) and consequently neither in they are not in the possession of the accused”.
“For the sake of completeness, I note that the respondent claimed that it was not necessary to create a new record to fulfill the information request of the applicant, although it could theoretically create a new record containing the requested information,” he said.
“To do this would require a data re-extraction exercise (including non-responding AIR data), complex data compilation and analysis, interpretation and creation of a new record. Furthermore, it is doubtful whether the exact same data could be extracted again, since the linked data tables five and six are being published there would be spots”.
‘A wrong custom;’
However, Ms Ooi argued the process presented NSW Health with “a major liability issue”.
“We’ve had three years of this pandemic and they say they never report the data they used to report, which they continued to produce every week,” he said.
“There’s no verification, we can’t verify that it’s right because they’ve deleted everything.”
He said he would always keep that data in his work so he could go back and “verify it was accurate” if something went wrong.
“Look for some data analyst, is this going to do?” she said.
“Create a complex analysis that takes hours to produce an output and never save your analysis, never except for some intermediate work, just publish the output? You can’t track it, there’s no tracking. It’s just bad practice, especially in the pandemic and vaccination program, which is all the more serious
A NSW Health spokesman declined to comment.
not black and white
Associate Professor James Trauer, head of Monash University’s Epidemiological Modeling Unit, said that, while he declined to comment on the legal details of the case, there appeared to be a “technical dispute” over the data.
“I agree with the broader point that Australian jurisdictions have publicly provided more disaggregated surveillance data during the pandemic,” he said in an email after the NCAT decision and Ms Ooi reviewed the Twitter thread.
“We know that unvaccinated people are represented in deaths and hospitalizations, but we need to know more about the individual characteristics of these people to calculate the true effectiveness of the vaccine. More details on this matter would have been helpful.
He noted that NSW “is leading the country in providing this information, so the issue is not limited to NSW”.
“But Australia is far behind other countries, such as the UK,” he said. “We need to think about how we can improve this for future pandemics. The Australian CDC could be the primary vehicle to make this happen.”
Professor Trauer added: “I see that most of Xin Yin Ooi’s tweets focus on identifying information. In reality, there is no simple black and white division between knowing and knowing information. Even aggregated file information can be linked to confidential information.”
