ANY_ANTIBIOTICS

Standardizations by Source

eICU (infusionDrug, medication):
- Antibiotics identified by regex on drugname
- infusionDrug: bactrim|cipro|flagyl|metronidazole|zithromax|zosyn|(((amika|cleo|ofloxa)|(azithro|clinda|tobra|vanco)my)c|(ampi|oxa|peni|pipera)cill|cefazol|levaqu|rifamp)in
- medication: cipro|flagyl|maxipime|metronidazole|tazobactam|zosyn|cef(azolin|epime)|(((azithro|clinda|vanco)my|ofloxa|vanco)c|levaqu|piperacill|roceph)in
HiRID (pharma*):
- Antibiotics identified by pharmaid in a predefined list.
MIMIC-III (prescriptions, inputevents_mv):
- Antibiotics identified by regex on prescriptions: (column DRUG) aztreonam|bactrim|cephalexin|chloramphenicol|cipro|flagyl|metronidazole|nitrofurantoin|tazobactam|rifampin|sulfadiazine|timentin|trimethoprim|(amika|gentami|vanco)cin|(amoxi|ampi|dicloxa|naf|oxa|peni|pipera)cillin|(azithro|clarithro|erythro|clinda|strepto|tobra|vanco)mycin|cef(azolin|tazidime|adroxil|epime|otetan|otaxime|podoxime|uroxime)|(doxy|mino|tetra)cycline|(levofl|moxifl|ofl)oxacin|macro(bid|dantin)|(una|zo)syn
- inputevents_mv: ITEMID in a predefined list
MIMIC-IV (inputevents):
- Antibiotics identified by regex on prescriptions: (column DRUG) aztreonam|bactrim|cephalexin|chloramphenicol|cipro|flagyl|metronidazole|nitrofurantoin|tazobactam|rifampin|sulfadiazine|timentin|trimethoprim|(amika|gentami|vanco)cin|(amoxi|ampi|dicloxa|naf|oxa|peni|pipera)cillin|(azithro|clarithro|erythro|clinda|strepto|tobra|vanco)mycin|cef(azolin|tazidime|adroxil|epime|otetan|otaxime|podoxime|uroxime)|(doxy|mino|tetra)cycline|(levofl|moxifl|ofl)oxacin|macro(bid|dantin)|(una|zo)syn
- inputevents: ITEMID in a predefined list
SICdb (medication):
- Antibiotics identified by DrugID in a predefined list
UMCdb (drugitems):
- Antibiotics identified by itemid in a predefined list

Extraction Table

Dataset	Source Table(s)	Filter / Condition	Item Column(s)
eICU	`infusionDrug`	`drugname` matches antibiotics regex	`drugname`
	`medication`	`drugname` matches antibiotics regex	`drugname`
HiRID	`pharma`	`pharmaid` in antibiotics list	`pharmaid`
MIMIC-III	`prescriptions`	`DRUG` matches antibiotics regex	`DRUG`
	`inputevents_mv`	`ITEMID` in antibiotics list	`ITEMID`
MIMIC-IV	`inputevents`	`itemid` in antibiotics list	`itemid`
SICdb	`medication`	`DrugID` in antibiotics list	`DrugID`
UMCdb	`drugitems`	`itemid` in antibiotics list	`itemid`

Pseudocode Steps

Initialize paths and dataset references.
Extract antibiotic administration data from each dataset as per the table above:
- Load the relevant source table(s).
- Apply the specified filter/condition (using the regex or ID lists above).
- Select the item columns.
- Add a global ICU stay ID.
Combine all datasets:
- Concatenate all antibiotic records.
- For each global ICU stay ID, set RECEIVED_ANY_ANTIBIOTICS = True if any record exists.
- Sort by global ICU stay ID.
Return the final DataFrame with standardized RECEIVED_ANY_ANTIBIOTICS per ICU stay.