CFU-like data

[latlatina]

Hi,

I hope it's OK to ask - I'm looking for a specific type of information, and I want to make sure it is available before I go through training and the rest of the procedure...
Does anyone know if there is information regarding CFU, or any type of bacterial load?
If I understood correctly, the microbiologyevents contains info about species, but not their amount, or abundance. Is it correct?

Thanks

[alistairewj]

Yes AFAIK that's correct. microbiologyevents contains: test name, isolate number, organism, antibiotic tested, and dilution level.

[latlatina]

Thank you,
If I understood correctly, the dilution level corresponds to the antibiotics and not the bacteria, right? It's not a value I can use to calculate bacterial abundance (CFU)...?

[alistairewj]

Yes correct, the dilution is for the antibiotic.

Double checking myself, as usual there is some information in the deidentified free-text comment associated with microbiology cultures:

select comments, count(*) as n
from `physionet-data.mimiciv_hosp.microbiologyevents`
where lower(comments) like '%cfu%'
group by 1 order by 2 desc

comments	n
"< 10,000 CFU/mL. "	24087
"10,000-100,000 CFU/mL Commensal Respiratory Flora. "	4019
"NO GROWTH, <1000 CFU/ml. "	2882
">100,000 CFU/mL Commensal Respiratory Flora. "	1188
~1000 CFU/mL Commensal Respiratory Flora.	475
~2000 CFU/mL Commensal Respiratory Flora.	425
~3000 CFU/mL Commensal Respiratory Flora.	244

But the microbiologyevents table has 3,228,713 rows so this is a very small proportion.

[latlatina]

Thank you for making sure. I really appreciate it.