I'm blaming a high varroa count with mine,
There should be very few colony failures due to varroa. The problem I suspect is a poor efficacy with apiguard, possibly coupled with doing it too early. I'm guessing you used apiguard?
Just think numbers and double up each month. Remember the DEFRA 'double each month' is likely on the optimistic side as there would have been little drone brood in most hives for the larger part of the intervening time for their increase.
Mite numbers should still be below a critical level if a) the treatment was done in September, b) there was not an excessive number of mites at treatment time, c) few have been picked up on robbing bees (or bees drifting from adjacent collapsing colonies) and d) the treatment was effective.
Apiguard advertised it as high efficacy, quoting all sorts of climates, but with no results shown for the UK. Go figure why. I expect in poor conditions, like in this last September, for some it's efficacy was in the seventies. So don't blame the mite - blame the treatment is my suggestion.
In my view oxalic acid treatment is to reduce the need for other treatments in the next foraging season and has very little to do with preventing colony mortality in the next two or three months.
Ther is too much hype around oxalic as a 'colony must be treated' (as it will save it from becoming a dead-out before spring). If the mite loading was high it has already done the damage to the over-wintering bees.
Think of too many mites (lack of other IPM treatments) and only a seventy percent kill rate. Now that would be far too many mites. You might save the colony by treating in January, but the damage was done long ago and the spring build-up will likely be tortuous and slow.
Agree or not, that is my take on it.
RAB