"Since those coming for blood tests do not constitute a representative sample of the Israeli population from a medical point of view. Such a survey will not allow scientific evaluations regarding possible error rates due to the fact that it is only a sample"
Prof. Danny Pepperman, the national statistician and director of the Central Bureau of Statistics, appealed to the government yesterday through the public, after his direct appeal was not answered. According to him, conducting the survey in the planned outline of the Ministry of Health will lead to biased and unrepresentative results, since those coming for blood tests do not constitute a representative sample of the Israeli population from a medical point of view. Such a survey will not allow scientific evaluations regarding possible error rates due to the fact that it is only a sample. Proper sampling is choosing a probability sample where each individual in the population has a positive (not necessarily equal) probability of being included in the sample. I demand that the decision-makers at the Ministry of Health hold an in-depth discussion on how the survey should be conducted, and act professionally and not conveniently!
In an official address to the Corona Cabinet, Prof. Pepperman said: "According to media reports, next week a national serological survey is going to begin among about 75,000 subjects on behalf of the Ministry of Health. As someone who has fought in every possible forum since near the beginning of the epidemic to have morbidity surveys, I of course welcome this, but I am very disturbed by the way the survey will be carried out, a sample of those who come to the health insurance funds anyway regardless of the survey for blood tests. Conducting the survey in such an outline will lead to biased and unrepresentative results, since those coming for blood tests do not constitute a representative sample of the Israeli population from a medical point of view, not to mention when it comes to the Corona period. For example, it is likely that at least some of those citizens who come for blood tests suffer from some medical problem, and perhaps precisely because of this they were more careful about contracting the disease. Beyond that, conducting the survey in the proposed format will not allow scientific evaluations regarding possible error rates due to the fact that it is only a sample. I will add that in a serological survey, one of the goals is to test interactions between individuals living in the same apartment, but the proposed sampling method will not allow this."
"I contacted the most senior officials in the Ministry of Health and other relevant government offices more than two weeks ago. I was alerted to the conduct of the survey in the proposed format and asked to have an in-depth discussion on the subject, but to my surprise and disappointment, I did not receive any response. According to the law, one of my duties as the national statistician is to advise the government ministries and guide them in carrying out statistical operations, and the government ministries are obliged to consult with me regarding official statistics that they produce and publish. Unfortunately, both things did not materialize and so I had no choice but to voice my position in public, because otherwise I feel that I will fail in my role."
The survey proposed by the Ministry of Health is called a "convenience sample" and there is no doubt that it is very convenient to sample in this way since the sampled come to give blood samples anyway. But with all due respect, you can't trade reliability for convenience. Proper sampling is choosing a probability sample where each individual in the population has a positive (not necessarily equal) probability of being included in the sample. The sample can be designed so that individuals living in the same apartment are sampled, and the estimation errors can be evaluated scientifically. The sample will also allow the calculation of estimates in different segments such as age and geographic location. This is how the CBS conducts over 60 surveys every year, and this is also how the sample was selected for the serological survey that is being conducted these days right in Bnei Brak, in collaboration with the Gartner Institute, which is responsible for carrying out the tests and interpreting them. Such a survey will require prior telephone coordination with the sampled, in which they will be asked to come to the health funds or designated complexes for the purpose of the test, and as the sampled avoid coming, it will be necessary to come to their home. Much less convenient, but this is how a statistically reliable survey should be conducted. It is important for me to say that, in my opinion, the best statisticians in Israel and the world, as well as experts in the health field, are the best partners. As mentioned, this is how the survey is carried out in Bnei Brak, and this is how it is also carried out in other countries in the world such as Italy and England.
I demand that the decision-makers at the Ministry of Health hold an in-depth discussion on how the survey should be conducted, and act professionally and not conveniently!
More of the topic in Hayadan:
Learn from the body to neutralize the corona
Researchers at the Technion, in Spain and in the UK present a new-old approach to disinfecting air and surfaces from the corona virus
Researchers from Bar-Ilan have identified an antigen that triggers an immune response towards finding a protein-based vaccine against the corona virus
A global survey proved: many corona patients experience damage to the senses of taste and smell
