Why are we tinkering with the vaccine manufacturers' directions?

The long-awaited COVID-19 vaccine roll-out is underway at last, in some countries at least. The Pfizer/BioNTech and Moderna products are already being administered, and the cheaper and easier-to-transport Oxford/AstraZeneca one is expected to follow in very short order. However, the roll-out is proceeding at a much slower pace than originally anticipated or promised. 

Partly as a result of this, some jurisdictions, notably the UK and some provinces in Canada, are taking the unilateral decision to concentrate on administering the first dose to as many vulnerable people as possible, even if this comes at the risk of delaying, or even missing, the second booster dose. Quebec has also changed tack and is re-allocating planned second doses to other people's first doses. Parts of the USA is also currently debating whether to go down a similar road. 

The UK has gone one step further and says it plans to ensure that the second dose is administered within 12 weeks of the first, despite the very clear recommendations of the manufacturers that a second dose must be administered within three or four weeks (depending on the particular vaccine).

It is maybe understandable why vaccine coordinators might want to make make this decision - a well-intentioned attempt to get as many people at least partially treated as possible as quickly as possible - but it still seems like a rather cavalier approach. Pfizer, for example, has been very up-front in saying that the second dose is absolutely essential for long term protection, and that the first dose alone may only offer protection for a week or two, and that  the two doses three weeks apart (not 12, not 20, not 1) are the only way to ensure the 95% efficacy rates people are expecting. There are some hopes that a first dose may be as much as 60-70% effective, but Pfizer says that this has not been rigorously tested, and little over 50% is more likely. Scientists are split on the policy.

Surely, if a vaccine, like any other medication, comes with specific instructions, they should be followed to the letter (like Pfizer's onerous and restrictive -70° storage instructions). Pfizer and BioNTech know their product much better than the British or Canadian governments. If they say to administer the second dose after three weeks, then that is exactly what we should be doing. We should not be second-guessing the manufacturers, not should we be cutting corners where people's lives are at stake.

UPDATE

The UK has taken another quantum leap towards anti-science by giving the green light to mix-and-match vaccines. 

For some reason, despite the complete lack of evidence and data that would support it, and contrary to specific warnings from the US Centers for Disease Control and Prevention and other scientists that the authorized COVID vaccines "are not interchangeable", the UK medical authorities are saying that, if a second dose of a vaccine that a patient has received is not available (which is quite conceivable due to its other policies - see above), then they should just be given a dose of any other vaccine as a substitute. They say that, "it is likely the second dose will help to boost the response to the first dose", based on no evidence whatsoever as far as I can see.

Say, what? Even as a non-scientist, that sounds wrong to me. Where is the UK coming from, or going to, with this stuff?

Subsequent clarifications indicate that this is expected to be an extremely unlikely eventuality, for example where a patient is not sure which vaccine they had been given, and that the government is not actually recommending the mixing of vaccines: "Every effort should be made to give them the same vaccine, but where this is not possible it is better to give a second dose of another vaccine than not at all."

UPDATE UPDATE

Quebec is now looking to follow the UK on this 12-week policy, a move that comes with substantial risks. For example, no-one really knows how fast the immunity conferred by the first dose drops off; encountering people with incomplete immunizations may lead the virus to mutate further in order to evade the vaccines; it is even possible that future supply of the vaccines may be withheld if countries are not using them as recommended. 

None of these outcomes are to be welcomed. So, why is Quebec being allowed too even consider it?

UPDATE UPDATE UPDATE

The National Advisory Committee on Immunization, the Canadian body that, at least theoretically, oversees the vaccine rollout, has now got involved in this debate, and has concluded that provinces are justified in delaying the second dose until 42 days after the first, but that any further delay after that (such as as the 90 days that Quebec is now trying to justify) becomes increasingly risky. 

They add that, particularly given that the vaccines are now being given to the most vulnerable populations, the closer to the recommended 21/28 days the better. Under no circumstances, therefore, should any province be delaying for 90 days.