I’m guessing maybe I’m not understanding what type of blood donation this data set is for (unless it’s platelet donation) because I’m wondering how some of the donors were able to donate so often. My understanding is that most organizations won’t let you donate more than once every 8 weeks, and for some other organizations it’s every 12 weeks.
For example in the training data set:
There are some cases like ID 474 (11 times in 11 months), ID 497 (8 times in 12 months) which seem too often. Then there are a lot of cases like 391, 471, and 377 which are listed as having donated twice but it says first donation and last donation are both 2 months ago. Then there is ID 313 who donate 5 times but first and last donations were both 14 months ago.
I don’t know much about donating blood but are these platelet donations or is it really possible to donate blood 5 times in one month?
As far as I can tell, the original paper that uses that data isn’t available outside of a paywall, so it may be a challenge to get more detail (though, I know some people have luck with #icanhazpdf on twitter.)
Sorry that’s not very satisfying, but we’d definitely be curious to hear any results you turn up. Thanks for bringing it up!
You 100% need to wait 8 weeks between whole blood donations so it is not possible that this dataset is for whole blood donations only. I think it’s both platelets AND whole blood since the data source information says they “selected 748 donors at random from the donor database” so that includes people who donate any type of blood related thing!
Also, some people donate both. you can donate platelets 72 hours after donating whole blood. If you do the platelets first, 72 hours later, you can donate whole blood.
Based on eligibility requirements there are certainly more procedures going on that is possible (assuming these are all whole blood donations) just by looking at the “months since last donation” and number of donations. It is showing that some donors have given once a month which biologically is not enough time for the body to reproduce the blood lost. The most logical explanation is that these donors are platelet donors, which are more diligent donors, as a platelet donation is a lot longer (up to two hours) than a whole blood donation, thus more dedication and propensity is required on the donors part. This attribute should make them more “predictable”. Also, if the donors were picked at random, just as there is a chance that platelet donors are in the training set, that also stands that donors who have donated double red cells could be in the training set. Which in that case their eligibility is twice that of a whole blood donation and their frequency would be quite different than a whole blood donor but not an accurate representation of there diligence to donate.
The only other explanation is they tried to donate but were deferred due to low iron count fever etc but still received a donation mark because they tried and then tried to donate soon there after again. I believe this is the more logical explanation because typically when a person donates platelets the volume is counted more than a whole blood donation so the donors that have a frequency that is not inline with eligibility requirements for whole blood should have a higher cc metric compared to whole blood donors.
