For months now I have been awaiting the outcome of the ping-pong between my boss and the guy in Beirut who sent us 250 samples to analyse. I'm sure I blogged about that project, but it's been a while. Anyway, the bossguy wants the results to be presented to the patients the way we would present them to our patients - with a real results sheet, interpreted and signed by the lab experts.
That means having the name of the patient on the report, as well as other usual identification details like their date of birth and sex.
But the other guy, he's got a research project going. What is it to us what the samples are called? None of our business. Anonymisation in research is a good thing.
Well, sure, and if we're just going to send a table back to Beirut with a mutation attached to a sample code, and that's it, ok. But one of the things our southern collaborators are bugging us about is the analysis taking so long while the patients are waiting for their results. And if the patients are waiting for results from us, we want those results to be as proper as possible.
No, no, the Lebanese want to take responsibility for getting the right result to the right patient. Just write a report with the code, and they'll add the name.
Doesn't work for us. No names, no reports.
I've had the bulk of these reports ready since October, waiting for the back and forth to end.
In the meantime, we have had to give them summary tables of our progress, in order for them to show their financial backers some progress, and to allow presentations and papers to be prepared. For us, these were strictly research-use results, but of course they went back to their patients with the information. So instead of getting a report where the results are interpreted and put into context by cancer genetics experts with recommendations about how to proceed, they got the bare bones.
Who comes out ahead there?
I don't know what they did with some of the stickier results, the "unclassified variants", but over the holidays we did at last receive a table with names across from the codes.
Of course, they put the whole name in one column, so you have to guess what's the family name, what's the given name (in your own culture that seems obvious enough, but to me, Majed and Haroum and Issa could go either way). Plus, most were 3-part names, and it isn't clear if the middle one is a person's middle name (second given name), or if it's a 2-part last name, or a maiden name. So I was not well advanced by that.
Turns out, the second name is the given name of either the woman's father or her husband. OK. It's still unclear to me if that is a usual way of referring to people in that part of the world, or if it's just the way their computer has them listed (as indicated in a semi-explanatory email).
This is all I'm going to get, however, for the identification of these cases. Birthdate, forget it. Sex I can verify on the pedigrees.
The boss is satisfied to write the reports now.
What I'm still not convinced of is whether we are actually more certain now that the right report is going to the right person. All I've done is taken a name from a list and added it to my report, which was generated according to the code. I will have someone check that I haven't made any mistakes (yes, you're welcome Yannick!), but how do I know the list is right? That there wasn't some mixup in the codes?
It is good to have a report with a name on it.
But is it really more secure for me to add it than for my collaborator to?