Bailiwick Express reported:
Plans to examine the entire population for signs of covid-19 have been put "on hold", the Health Minister has confirmed - leaving 150,000 antibody test kits ordered by Government just weeks ago lying on the shelf.
The news comes after the results of a pilot test scheme using kits provided by Healgen were released last week, suggesting that up to 3,300 islanders may have been infected with the virus.
The Health Minister had previously stated that, if the pilot scheme was successful, antibody testing would likely be rolled out to the whole population.
But when Express asked whether this was still the plan at a press conference this morning, Deputy Richard Renouf indicated that the government was no longer pursuing the idea.
He described the results of the pinprick blood test regime, released by Statistics Jersey on Friday, as “helpful and statistically valid”.
I have to say, rather surprisingly perhaps, that I am in almost total agreement with Richard Renouf.
However, I do think they should repeat the random sample again, at least twice more, to see how much variance is thrown up and whether the results match (or come close to matching).
There are also two things the report doesn’t show:
Plans to examine the entire population for signs of covid-19 have been put "on hold", the Health Minister has confirmed - leaving 150,000 antibody test kits ordered by Government just weeks ago lying on the shelf.
The news comes after the results of a pilot test scheme using kits provided by Healgen were released last week, suggesting that up to 3,300 islanders may have been infected with the virus.
The Health Minister had previously stated that, if the pilot scheme was successful, antibody testing would likely be rolled out to the whole population.
But when Express asked whether this was still the plan at a press conference this morning, Deputy Richard Renouf indicated that the government was no longer pursuing the idea.
He described the results of the pinprick blood test regime, released by Statistics Jersey on Friday, as “helpful and statistically valid”.
I have to say, rather surprisingly perhaps, that I am in almost total agreement with Richard Renouf.
However, I do think they should repeat the random sample again, at least twice more, to see how much variance is thrown up and whether the results match (or come close to matching).
There are also two things the report doesn’t show:
- I assume that none of the random sample had already been tested positive (296 cases) or negative (3,481 cases) for Coronavirus, as the number of tests conducted for antibodies was small. The sample consisted of 438 households and 855 individuals so the probabilities are pretty low even if we screen out children and assume a working population of 60,000. But it would be nice to have that in the report.
- The letter sent out was in English. Was any of the non-response from the population for whom English is not a first language? In fact, were there any language issues? That too would be useful to know.
If the numbers are so low, 3,300 out of 100,000, it throws the whole notion of a “Covid-19 Certificate” largely out of the window. The notion was that if it was high, it could be used as Germany have suggested to have a certificate of immunity and open up the economy. Unfortunately (or fortunately really!) the lockdown has reduced the spread so that is not the case, and it would be wasteful.
I have examined the statistics report with my mathematics head on, and it is very solid stuff – they do stratified weighting to look at total spread. The key worry is that only 63% of those contacted responded. Providing that is a random number who failed to respond, there is no problem, but if it was determined by other factors – and as the report states, “there was a significant level of non-response of those Islanders living in non-qualified accommodation.”
“Whilst the overall response rate to this survey was generally of an acceptable level, non-response from the non-qualified sector was particularly high. This may in part due to the methods employed to engage households in thissurvey and the time constraints imposed on the field work. As a result, weighting by tenure was not possible. Therefore, particular care should be taken when considering the results in the context of this particular tenure category.”
“It should be noted that it is not possible to completely control for non-response error as there may be characteristics that impact non-response that may also impact the prevalence rate in that population. This is a potential source of bias in the result that can be improved by achieving a higher level of response.”
I have examined the statistics report with my mathematics head on, and it is very solid stuff – they do stratified weighting to look at total spread. The key worry is that only 63% of those contacted responded. Providing that is a random number who failed to respond, there is no problem, but if it was determined by other factors – and as the report states, “there was a significant level of non-response of those Islanders living in non-qualified accommodation.”
“Whilst the overall response rate to this survey was generally of an acceptable level, non-response from the non-qualified sector was particularly high. This may in part due to the methods employed to engage households in thissurvey and the time constraints imposed on the field work. As a result, weighting by tenure was not possible. Therefore, particular care should be taken when considering the results in the context of this particular tenure category.”
“It should be noted that it is not possible to completely control for non-response error as there may be characteristics that impact non-response that may also impact the prevalence rate in that population. This is a potential source of bias in the result that can be improved by achieving a higher level of response.”
The other thing which has occurred recently is a private enterprise setting up their own testing centre. As the government don't want to waste antibody tests, and the business has agreed to share results in terms of numbers, rather than shutting it down, they should (1) check the test is properly accredited and (2) ensure the test facility is not like to become a "hotspot" of contamination. What you don't want are people showing symptoms turning up with attendant risk!
The Government has taken a line given by the usual official (anonymous) spokesperson who says: "We urge Islanders to be extremely cautious before giving their blood in a setting that we cannot verify meets clinical standards for care."
Well why on earth don't they verify it rather than being moaning minnies!
Deputy Gregory Guida comments:
"This is an enzyme-linked immunosorbent assay which detects the presence of two antibodies in the blood. One antibody will be present at the tail end of the illness and the other mostly after recovery. This test can basically tell you if you -have had- Covid 19, in which case, you are -less likely- to catch it again. I don't know from which laboratory these are sourced, but so far the tests available on the market have had about 90% sensitivity and 99% specificity. That means that if you have a negative result you still have a 10% chance of having had the virus and if you test positive, there is a less than 1% chance that you haven't had the virus and something else was detected instead. I believe the test will cost around £76. As less than 3% of the population have had the virus so far, I would only test if you have had unequivocal symptoms and strongly suspect you have been ill with Covid 19."
As noted by the private company, this is a test which is being used by the Australian government,
But for people who have just been told to self-isolate with what appear to be symptoms and who have not been tested, or who have been in close proximity to someone who has tested positive, and want to see if they have had it asymptomically, it is a good idea.
The more data we can get on the spread, the better, and if it provides a degree of reassurance, well and good. The government has said they won't accept any data, but I hope it gets published anyway. Data is the key to understanding risk.
But what should also be realised, both with a private test, and indeed with the government test, is that a positive result does not mean that any lessening of lockdown and hygiene measures should be taken by those testing positive for having had the virus. Just like the wearing of facemasks, a false sense of security must be guarded against.
The Government has taken a line given by the usual official (anonymous) spokesperson who says: "We urge Islanders to be extremely cautious before giving their blood in a setting that we cannot verify meets clinical standards for care."
Well why on earth don't they verify it rather than being moaning minnies!
Deputy Gregory Guida comments:
"This is an enzyme-linked immunosorbent assay which detects the presence of two antibodies in the blood. One antibody will be present at the tail end of the illness and the other mostly after recovery. This test can basically tell you if you -have had- Covid 19, in which case, you are -less likely- to catch it again. I don't know from which laboratory these are sourced, but so far the tests available on the market have had about 90% sensitivity and 99% specificity. That means that if you have a negative result you still have a 10% chance of having had the virus and if you test positive, there is a less than 1% chance that you haven't had the virus and something else was detected instead. I believe the test will cost around £76. As less than 3% of the population have had the virus so far, I would only test if you have had unequivocal symptoms and strongly suspect you have been ill with Covid 19."
As noted by the private company, this is a test which is being used by the Australian government,
But for people who have just been told to self-isolate with what appear to be symptoms and who have not been tested, or who have been in close proximity to someone who has tested positive, and want to see if they have had it asymptomically, it is a good idea.
The more data we can get on the spread, the better, and if it provides a degree of reassurance, well and good. The government has said they won't accept any data, but I hope it gets published anyway. Data is the key to understanding risk.
But what should also be realised, both with a private test, and indeed with the government test, is that a positive result does not mean that any lessening of lockdown and hygiene measures should be taken by those testing positive for having had the virus. Just like the wearing of facemasks, a false sense of security must be guarded against.
Posts
No comments:
Post a Comment