No factor in isolation causes an accident, and this was something that Professor James Reason, of Manchester University, as one of the leading experts in Human Error always stressed including the active and latent failures that are always present in accidents, just consider the Herald of Free Enterprise, Quinteshill, Bhopal, Erebus, Piper Alpha disasters.
Fireflybob is correct to suggest that there will be more than one factor, not least the human factors consideration in having a locked flight deck door, but if you take the Reason Model (Swiss Cheese) further by considering the SHELL model, and by that I mean the interfaces that will always exist between the Liveware(pilots), Software (proceedures, SOPs), Hardware (the aircraft), and the Environment (Wx,Altitude).
Hence in this disaster:
Liveware-Liveware (No one detected that were issues with the co-pilot, for example other crew members, medical officers, and the company)
Liveware-Hardware (no means of Captain getting back into the flightdeck)
Liveware-Software (company culture, rapid turnaround, leaving am inxeperienced pilot at the controls, the need for a crew member to leave the flight deck on such a short flight, when perhaps this should have been done on the turnaround, perhaps a need to amend the SOP here).
Liveware-Environment (possibly this doesn't come into play, unless there was a loss of pressurisatiion).
This isn't a comprehensive list, just the start as facts become clear, but if any pilot seriously thinks there is only one factor in this accident, then I kindly suggest that they haven't recieved the Human Factors/CRM training they deserve.