From Legerwood:

I am a bit puzzled by the numbers quoted as having been discharged in March. I have seen them quoted elsewhere too. Public Health Scotland collects data on the monthly numbers of delayed Discharges and the reports are published on the ISD Scotland web site. According to their Feb 2020 report there were 1627 patients delayed in hospital at the census point which is the last thursday in the month. The March 2020 report stated there were 1171 delayed patients at the census points. That is 456 patients discharged.

https://www.isdscotland.org/Health-Topics/Health-and-Social-Community-Care/Publications/index.asp?#2674

Apparently Health and Social Care also count delayed discharges. Double accounting?

ED: There is of course no evidence that these transferred patients caused the spread of disease in the homes nor that they were themselves infected there: https://talkingupscotlandtwo.com/2020/05/25/heralds-tom-gordon-has-become-a-parody-in-effort-to-blame-first-minister-for-care-home-deaths/

The first case of Covid19 in Scotland was confirmed on 1st March. It was the 16th March before every health board in Scotland bar Orkney and Western Isles had any cases. The number of confirmed cases at that point was 171. Therefore patients being discharged into care homes in the first half of the month had a very good chance of being discharged from hospitals that did not have any, or very few, Covid patients. In some ways you can see why the advice might have been not to test.

Furthermore on 12th March the test and trace was abandoned and only hospital patients with suspected Covid were to be tested. This may have been decided because PHE were having trouble supplying enough test kits. The test was one developed in the UK. It was in use by PHE from the off but Scotland was not provided with supplies of the test until Feb 10th. Until then all samples had to go to a PHE lab in London and even after tests were carried out in Scotland, at GRI and ERI labs, any positive results had to be sent to the PHE lab for confirmation.

Apart from the problems with supply of the test kits, and swabs for sampling, the test itself was not very accurate. About 70% – 75% accuracy. Around April 21-23rd labs were instructed to change to commercial tests which were much more accurate.

So a complex picture with all sorts of layers and certainly not as straight forward as the opposition parties are trying to make out in their endless pursuit of SNP bad.