Fast reading isn't always good reading

A welcome research week before the onslaught of semester 2 begins. It's so long since I had any sustained time to work on my PhD that it takes me a day to remind myself of some of the main themes of my research so far. I start reading through a stack of journal articles only to find that I must have already read them, because they're covered with underlining and my scribbled comments. Worryingly, I have no memory of any of the content of these articles, or even of having read them, so I go through them again and make notes on the main themes and arguments. Not exactly an efficient use of time but as I'm discovering, one of the most difficult aspects of doing a PhD in the gaps between the rest of my job is that the process of research is very disjointed. I read fast because I have limited time, but how useful is this if I can't remember what I've read? There must be a better way...