In the week ahead Parliament will debate the lock down, and government will determine a back to work policy. To do so, they need to examine some of the numbers they are using carefully.
- Comparative deaths globally. It is quite wrong to say the UK after the US will have the most deaths. The government must adjust the death figures for population, which so far shows Belgium heading the lists, and the USA still relatively low
- The government needs to do more work on trying to get comparable death rates. Some other countries only list hospital deaths. There are different views on whether dying with Covid 19 is the same as dying of Covid 19. The UK figures for deaths have probably been boosted in recent days and weeks by counting more deaths where the patient died without a Covid 19 test as a Covid death, and by adding in non hospital deaths to the total. It is a bad idea to change the way you calculate numbers over time for a series when you are using the curve of that series over time to determine policy.
- Now there are so many more tests available the government needs to start testing a sample of the total population to get figures for how common this disease is, and to chart that over time.
- The accuracy of the tests. I have been given widely different figures for how many false negatives the tests might provide. There are apparently issues about how to secure a good sample so any disease does show up.
- How good are the returns reported centrally from Care Homes, as most of these are privately owned institutions that are not formally part of NHS management and reporting systems.
- Future capacity of the NHS. The government is right to want reassurance that the NHS can cope in future as it has done so far. There needs to be a rebased figure for Covid care beds and Covid Intensive care beds in the system in a world where there is also capacity for other serious medical conditions. Will the NHS move to identifying specialist Covid hospitals and units, to free more District General Hospitals to do everything else?
- The value of R or the reinfection rate. We were told this week it is currently 0.6 to 0.9. That is a very wide range. How can it be more precisely and accurately be discovered, where there are not sampled tests of the whole population over time? Doesn’t it need regular sampled testing to get it more accurate? As great stress is placed on R, it is crucial to get it right.
- Will the government publish the range of forecasts of deaths from the disease their experts have come up with, and show us the trend in these forecasts? That too is important in making a judgement.