Chris Gianacas wrote:
Hi Duncan,
My experience - a few years old but still current I think - is that direct access to the general PBS/MBS datasets for secondary use research is effectively impossible. One way around is to go via 45-and-up, as they have individual consent and many other datasets in addition to PBS/MBS all already linked, but this comes at a cost. Not sure where the MHR secondary use framework is at but that may be worth considering too.
Best of luck!
Hi Chris
Thanks for your comments. I am still trying with great frustration just to get my whole PBS/MBS history through official channels
I have made remarkable progress with the help of one remarkably efficient pharmacist (and database) and a few very helpful GPs delving into ancient records. That's those who haven't retired or closed down completely. A couple of full hospital records have helped too. A few practices have closed or deleted everything - EDIT just remembered an interesting aside in case people think I forget which forum we are in. You are doing historical longitudinal research when information attached to items may have changed over time - if your database just has current information attached to historical records it effects possible analysis too. You need to source that changed information elsewhere. At least it appears that the pharmacies are up to date though :)
I have serious concerns about the status of said data.
If, as I hypothesise, there are such serious issues with (alleged - doing my own legals too, but try getting a legal practice interested) historical human rights abuses indirectly caused by medications and mistreatments, there are very important legal reasons for people individually and collectively to get prompt access to their complete medical histories
Unless people can get extensions on limitations every week or month's delay accessing important evidence is problematic
If anyone in the PBS/MBS data area can shed any light on the status of said data I would be very grateful
Time is of the essence, as they say
But of course, as I hinted, even if all that data were available at a population level it would be a seriously resource intensive analysis
I have some vague memories of the requirements of such analyses. Human and computational
The considerable evidence may remain circumstantial and anecdotal it appears :)