
It was, as all studies are, a collaboration of multiple people. But Pedro has been excited about this one from the beginning, as it was an opportunity for him to use his increasing skills in machine learning, artificial intelligence (AI), and natural language processing (NLP) in the field of medicine.
I’m posting a link here because I mentioned it before, and a couple people expressed interest. Here is the study objective:
“This research protocol describes the development of a stakeholder-informed NLP model to identify patients with exposure to firearm violence using clinical notes from a large, multistate network of ambulatory health care centers by classifying text as indicative as exposure to firearm violence or not.”
Said another way, this study used computers to scrape the patient notes across multiple healthcare facilities in different states (all private information was, of course, unidentifiable so that no data involved can ever be linked to a specific person). The computers were trained to find information related to potential firearm violence in the clinical notes. The reason the investigators wanted to look in notes is because sometimes information about firearms is not part of the current treatment, and therefore isn’t noted in the main diagnosis and treatment section. A purely hypothetical example might be that a mother brings in a baby for a wellness check. The baby is healthy, but the mother does express worry that someone in the household doesn’t lock up their firearm. This study would have found that note.
The study is valuable in its information about firearms, of course, but also it is groundbreaking in its use of computers to do the work. Setting a computer to this task is much more efficient and more reliable than asking a human being to read records and report findings.
Pedro is eager to learn more and work more in this field. On his free time, he has been taking university courses from MIT and from Carnegie Mellon in NLP as they relate to the field of medicine. He consistently keeps his eye on new opportunities to participate in more studies like the one above. Here’s the link to the study on JMIR Publications.
Well done Pedro 👏🏼👏🏼 I hope the use in a practical situation will be available soon to those who will use it. Although it is a pity that it has to be used at all
Thanks Brian, and I agree on both counts.
Congratulations to Pedro. The work would be well beyond me
Social work looks different in 2025. I’m so glad for people like Pedro who find this type of work enthralling.
Compliments to Pedro on a work well done. Daughter #2 has been doing work on gun violence in Central America. I’ll see if I can send her the link…
👏🏻
Oh how interesting to have such an overlap.
I couldn’t share the link, but I downloaded the pdf and sent it to my daughter. Not sure it falls in line with what she does…
I hate to rain on anybody’s parade, but is Pedro carrying his ID? NO matter how angry it would make him? Why take risks?
He has his usual wallet on him most of the time. Nothing more. He is unassailable on this topic, however. Pedro has a deep faith in the US Justice system (one I do not share right now). He knows how risky it is and still declares, “I dare them!”
I know you mentioned that before… I wouldn’t trust the US Justice system right now. The judges are doing the best they can, but the Supremes? They’ve all but abandoned all power to the Sultan. They don’t even justify their decisions any more…