Skip to main content
CLIF Logo

CLIF Data Dictionary 2.2.0

Version CLIF v2.2.0 is a WIP implementation.

CLIF Data Dictionary

The CLIF Data Dictionary serves as a comprehensive guide to the Common Longitudinal ICU data Format, detailing the structure and purpose of each table within the framework. Designed to standardize and harmonize electronic health record data across multiple institutions, the dictionary outlines the entity-relationship model, variable definitions, and permissible values.

ERD

Beta Tables

The table purpose, structure, and field names for beta tables is complete and used in at least one federated CLIF project. The minimum Common ICU Data Elements (mCIDE) for category variables is defined. Actively testing the table’s practical use in projects. Breaking changes unlikely, but backward compatible updates in future minor versions possible

adt Beta

The admission, discharge, and transfer (adt) table is a start-stop longitudinal dataset that contains information about each patient’s movement within the hospital. It also has a hospital_id field to distinguish between different hospitals within a health system.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
hospital_idVARCHARAssign a unique ID to each hospital within a healthsystemNo restriction
hospital_typeVARCHARMaps hospital_id to a standardized list of hospital typesacademic, community, LTACH
in_dttmDATETIMEStart date and time at a particular location. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
out_dttmDATETIMEEnd date and time at a particular location. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
location_nameVARCHARLocation of the patient inside the hospital. This field is used to store the patient location from the source data. It is not used for analysis.No restriction
location_categoryVARCHARMaps location_name to a standardized list of ADT location categoriesed, ward, stepdown, icu, procedural, l&d, hospice, psych, rehab, radiology, dialysis, other
location_typeVARCHARMaps ICU type to a standardized list of ICU categoriesList of ICU categories in CLIF

Notes:

  • ADT represents the patient’s physical location, NOT the patient “status”.
  • Procedural areas and operating rooms should be mapped to Procedural. Pre/Intra/Post-procedural/OR EHR data (such as anesthesia flowsheet records from Labs, Vitals, Scores, Respiratory Support) are not currently represented in CLIF.


Example:

hospitalization_idhospital_idhospital_typein_dttmout_dttmlocation_namelocation_categorylocation_type
20010012ABCacademic2024-12-01 10:00:00+00:00 UTC2024-12-01 14:00:00+00:00 UTCB06Ficugeneral_icu
20010012ABCacademic2024-12-01 14:30:00+00:00 UTC2024-12-02 08:00:00+00:00 UTCB78Dward
20010015ABCacademic2024-11-30 16:45:00+00:00 UTC2024-12-01 12:00:00+00:00 UTCB06Ticugeneral_icu
20010015ABCacademic2024-12-01 12:30:00+00:00 UTC2024-12-02 07:00:00+00:00 UTCN23Eprocedural
20010020EFGcommunity2024-11-28 09:00:00+00:00 UTC2024-11-29 17:00:00+00:00 UTCB78Dward

code status Beta

This table contains only code status orders placed by clinicians. It is NOT equivalent to the code status display name in the EMR.

Unstable
ColumnTypeDescriptionPermissible Values
patient_idVARCHARUnique identifier for each patient, presumed to be a distinct individual.No restriction
start_dttmDATETIMEThe date and time when the specific code status was initiatedDatetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
code_status_nameVARCHARThe name/description of the code statusNo restriction
code_status_categoryVARCHARCategorical variable specifying the code status during the hospitalizationDNR, DNAR, UDNR, DNR/DNI, DNAR/DNI, AND, Full, Presume Full, Other

Notes:

  • The code_status_category set of permissible values is under development


Example:

patient_idstart_dttmcode_status_namecode_status_category
1234512024-12-01 08:30:00+00:00 UTCDo Not ResuscitateDNR
1234522024-12-02 14:00:00+00:00 UTCDo Not IntubateDNR/DNI
1234512024-12-03 10:15:00+00:00 UTCFull CodeFull

crrt therapy Beta

The crrt_therapy table captures Continuous Renal Replacement Therapy (CRRT) data, including different CRRT modalities, operational parameters, and fluid exchange details. The intermittent HD, peritoneal dialysis, PERT, and SLED tables are under development.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
device_idVARCHARUnique ID of the individual dialysis machine used (e.g., machine ACZ3RV91). Distinct from dialysis_machine_name, which stores the brand/model.No restriction
recorded_dttmDATETIMETimestamp when CRRT parameters were recordedDatetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
crrt_mode_nameVARCHARName of CRRT mode (e.g., CVVHDF)No restriction
crrt_mode_categoryVARCHARStandardized CRRT mode categoriesscuf, cvvh, cvvhd, cvvhdf, avvh
dialysis_machine_nameVARCHARBrand name for the dialysis machineNo restriction
blood_flow_rateFLOATRate of blood flow through the CRRT circuit (mL/min)150-350
pre_filter_replacement_fluid_rateFLOATRate of pre-filter replacement fluid infusion (mL/hr)0-10000
post_filter_replacement_fluid_rateFLOATRate of post-filter replacement fluid infusion (mL/hr)0-10000
dialysate_flow_rateFLOATFlow rate of dialysate solution (mL/hr)0-10000
ultrafiltration_outFLOATNet ultrafiltration output (mL/hr)0-500

Notes:

  • SCUF: Slow Continuous Ultrafiltration
  • CVVH: Continuous Veno-Venous Hemofiltration
  • CVVHD: Continuous Veno-Venous Hemodialysis
  • CVVHDF: Continuous Venous-Venous Hemodiafiltration
  • AVVH: Accelerated Veno-venous Hemofiltration also called ARRT or PIIRT

    CRRT Modalities and Parameter Usage:
CRRT ModalityBlood Flow RatePre-Filter Replacement RatePost-Filter Replacement RateDialysate Flow RateUltrafiltration Out
SCUFRequiredNot UsedNot UsedNot UsedRequired
CVVHRequiredRequiredRequiredNot UsedRequired
CVVHDRequiredNot UsedNot UsedRequiredRequired
CVVHDFRequiredRequiredRequiredRequiredRequired
AVVH (VVH)RequiredMay Be UsedMay Be UsedNot UsedRequired
AVVH (VVHD)RequiredNot UsedNot UsedMay Be UsedRequired
AVVH (VVHF)RequiredMay Be UsedMay Be UsedMay Be UsedRequired

Example:

hospitalization_iddevice_idrecorded_dttmcrrt_mode_namecrrt_mode_categorydialysis_machine_nameblood_flow_ratepre_filter_replacement_fluid_ratepost_filter_replacement_fluid_ratedialysate_flow_rateultrafiltration_out
201J02024-02-15 07:00:00+00:00 UTCCVVHDFCVVHDFNxStage by Baxter200.01000.0500.0800.01500.0
202J02024-02-16 09:15:00+00:00 UTCCVVHCVVHNxStage by Baxter180.01200.0300.0NA1300.0
203J02024-02-17 11:45:00+00:00 UTCSCUFSCUFNxStage by Baxter150.0NANANA800.0

hospital diagnosis Beta

Finalized billing diagnosis codes for hospital reimbursement, e.g. calculation of a Diagnosis Related Group (DRG). These diagnoses also do not have timestamps, as they are often finalized after discharge. The hospital_diagnosis table is appropriate for calculation of comorbidity scores but should not be used as input features into a prediction model for an inpatient event.

All other diagnosis codes for a patient are included under concept table patient_diagnosis which has start and end timestamps.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterMust match a hospitalization_id in the hospitalization table
diagnosis_codeVARCHARICD diagnosis codeValid ICD-9-CM or ICD-10-CM code
diagnosis_code_formatVARCHARFormat of the codeICD10CM or ICD9CM
diagnosis_primaryINTType of diagnosis: 1 = primary, 0 = secondary. If diagnoses are ranked, any rank of 2 or above is considered secondary.0, 1
poa_presentINTIndicator if the diagnosis was present on admission. Only two options are allowed: 1 = Yes (present on admission), 0 = No (not present on admission). No other values (such as Exempt, Unknown, or Unspecified) are permitted.0, 1

Example:

hospitalization_iddiagnosis_codediagnosis_code_formatdiagnosis_primarypoa_present
20010012I10ICD10CM11
20010012E11.9ICD10CM00
20010015250.00ICD9CM11
20010015401.9ICD9CM00
20010020J45.909ICD10CM11
20010020530.81ICD9CM01

hospitalization Beta

The hospitalization table contains information about each hospitalization event. Each row in this table represents a unique hospitalization event for a patient. This table is inspired by the visit_occurance OMOP table but is specific to inpatient hospitalizations (including those that begin in the emergency room).

Unstable
ColumnTypeDescriptionPermissible Values
patient_idVARCHARUnique identifier for each patient, presumed to be a distinct individual.No restriction
hospitalization_idVARCHARID variable for each patient encounterNo restriction
hospitalization_joined_idVARCHARUnique identifier for each continuous inpatient stay in a health system which may span different hospitals (Optional)No restriction
admission_dttmDATETIMEDate and time the patient is admitted to the hospital. All datetime variables must be timezone-aware and set to UTCDatetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
discharge_dttmDATETIMEDate and time the patient is discharged from the hospital. All datetime variables must be timezone-aware and set to UTCDatetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
age_at_admissionINTAge of the patient at the time of admission, in yearsNo restriction
admission_type_nameVARCHARType of inpatient admission. Original string from the source datae.g. Direct admission, Transfer, Pre-op surgical
admission_type_categoryVARCHARAdmission disposition mapped to mCIDE categoriesList of admission type categories in CLIF
discharge_nameVARCHAROriginal discharge disposition name string recorded in the raw dataNo restriction, e.g. home
discharge_categoryVARCHARMaps discharge_name to a standardized list of discharge categoriesHome, Skilled Nursing Facility (SNF), Expired, Acute Inpatient Rehab Facility, Hospice, Long Term Care Hospital (LTACH), Acute Care Hospital, Group Home, Chemical Dependency, Against Medical Advice (AMA), Assisted Living, Still Admitted, Missing, Other, Psychiatric Hospital, Shelter, Jail
zipcode_nine_digitVARCHARPatient 9 digit zip code, used to link with other indices such as ADI and SVINo restriction
zipcode_five_digitVARCHARPatient 5 digit zip code, used to link with other indices such as ADI and SVINo restriction
census_block_codeVARCHAR15 digit FIPS codeNo restriction
census_block_group_codeVARCHAR12 digit FIPS codeNo restriction
census_tractVARCHARFull 11 digit FIPS code. Eg. 13089022404 census_tract is the state (13) + the county (089) + the census tract (022404).No restriction
state_codeVARCHAR2 digit FIPS codeNo restriction
county_codeVARCHARFull 5 digit FIPS code. Eg. 13089 county_code is the state (13) + the county (089).No restriction
fips_versionVARCHARYear of the Census geography definitions used for the FIPS codes (e.g., 2010, 2020), indicating the tract and boundary set in effect at that time2000,2010, 2020

Notes:

  • If a patient is discharged to Home/Hospice, then discharge_category == Hospice.
  • Mortality Outcomes: Mortality is identified by discharge_category == "Expired". Some studies also include discharge_category == "Hospice" (i.e., both “Expired” and “Hospice”) when defining death outcomes.
  • The geographical indicators (zipcode_nine_digit, zipcode_five_digit, census_block_code, census_block_group_code, census_tract, state_code, county_code) should be added if they are available in your source dataset. zipcode_nine_digit is preferred over zipcode_five_digit, and census_block_code is ideal for census based indicators. The choice of geographical indicators may differ depending on the project.
  • If a patient is transferred between different hospitals within a health system, a new hospitalization_id should be created.
  • If a patient is initially seen in an ER in hospital A and then admitted to inpatient status in hospital B, one hospitalization_id should be created for data from both stays.
  • A hospitalization_joined_id can also be created from a CLIF table from contiguous hospitalization_ids.
  • Geo-based Indices (ADI, SVI):
    • ADI: Calculate the Area Deprivation Index at the census block-group level. Provide census_block_group_code when possible, or zipcode_nine_digit (9-digit ZIP) that can be cross-walked to a block group. Avoid using 5-digit ZIP or census-tract values for ADI—they are not validated and will add error.
    • SVI: The Social Vulnerability Index is published at census_tract (full 11-digit FIPS).


Example:

patient_idhospitalization_idhospitalization_joined_idadmission_dttmdischarge_dttmage_at_admissionadmission_type_nameadmission_type_categorydischarge_namedischarge_categoryzipcode_five_digitzipcode_nine_digitcensus_block_group_codelatitudelongitudefips_version
10100112345678223344552024-11-01 08:15:00+00:00 UTC2024-11-04 14:30:00+00:00 UTC65Direct admissionInpatientDischarged to Home or Self Care (Routine Discharge)Home6063760637000017031320200141.81030-87.596972020
10100287654321223344552024-11-04 15:00:00+00:00 UTC2024-11-07 11:00:00+00:00 UTC72Transfer from another hospitalAcute Care TransferTransferred to Acute Inpatient Rehab FacilityAcute Inpatient Rehab Facility4631146311000017031330100241.55030-87.301012020
10100311223344112233442024-10-20 07:45:00+00:00 UTC2024-10-22 10:20:00+00:00 UTC59Pre-op surgicalPre-opExpiredExpired6044660446000017031340100341.70010-87.603152020

labs Beta

The labs table is a long form (one lab result per row) longitudinal table.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
lab_order_dttmDATETIMEDate and time when the lab is ordered. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
lab_collect_dttmDATETIMEDate and time when the specimen is collected. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
lab_result_dttmDATETIMEDate and time when the lab results are available. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
lab_order_nameVARCHARProcedure name for the lab, e.g. “Complete blood count w/ diff”No restriction
lab_order_categoryVARCHARMaps lab_order_name to standardized list of common lab order names, e.g. “CBC”List of lab order categories in CLIF
lab_nameVARCHAROriginal lab component as recorded in the raw data, e.g. “AST (SGOT)”.No restriction
lab_categoryVARCHARMaps lab_name to a minimum set of standardized labs identified by the CLIF consortium as minimum necessary labs for the study of critical illness.List of lab categories in CLIF
lab_valueVARCHARRecorded value corresponding to a lab. Lab values are often strings that can contain non-numeric results (e.g. “> upper limit of detection”).No restriction
lab_value_numericDOUBLEParse out numeric part of the lab_value variable (optional).Numeric
reference_unitVARCHARUnit of measurement for that lab.Permissible reference values for each lab_category listed here
lab_specimen_nameVARCHAROriginal fluid or tissue name the lab was collected from as given in the source dataNo restriction
lab_specimen_categoryVARCHARFluid or tissue the lab was collected from, analogous to the LOINC “system” component.working CDE c(blood/plasma/serum, urine, csf, other)
lab_loinc_codeVARCHARLOINC code for the labNo restrictions
loinc_versionVARCHARVersion or release of the LOINC coding system used for lab_loinc_code, e.g. 1.0 or 1.1.No restriction

Notes:

  • All lab values must be reported using the lab’s reference units linked above in permissible values. Only the listed reference units are permissible for respective lab categories in the CLIF labs table. Sites must ensure that any raw laboratory values are converted to the reference units during the ETL process. Entries with other units must be transformed prior to loading into CLIF.
  • The lab_value field often has non-numeric entries that are useful to make project-specific decisions. A site may choose to keep the lab_value field as a character and create a new field lab_value_numeric that only parses the character field to extract the numeric part of the string.


Example:

hospitalization_idlab_order_dttmlab_collect_dttmlab_result_dttmlab_order_namelab_order_categorylab_namelab_categorylab_valuelab_value_numericreference_unitlab_specimen_namelab_specimen_categorylab_loinc_codeloinc_version
10010142023-05-01 07:00:00+00:00 UTC2023-05-01 07:15:00+00:00 UTC2023-05-01 08:00:00+00:00 UTCComplete blood count w/ diffCBCWBCwhite_blood_cell_count8.28.210^3/uLbloodblood/plasma/serum6690-21.0
10010142023-05-01 07:00:00+00:00 UTC2023-05-01 07:15:00+00:00 UTC2023-05-01 08:00:00+00:00 UTCComplete blood count w/ diffCBCHGBhemoglobin13.513.5g/dLbloodblood/plasma/serum718-71.0
10020252023-06-10 08:30:00+00:00 UTC2023-06-10 08:45:00+00:00 UTC2023-06-10 09:00:00+00:00 UTCBasic metabolic panelBMPSodiumsodium140140mmol/Lbloodblood/plasma/serum2951-21.0
10020252023-06-10 08:30:00+00:00 UTC2023-06-10 08:45:00+00:00 UTC2023-06-10 09:00:00+00:00 UTCBasic metabolic panelBMPPotassiumpotassium4.24.2mmol/Lbloodblood/plasma/serum2823-31.0
10030362023-07-15 06:45:00+00:00 UTC2023-07-15 07:00:00+00:00 UTC2023-07-15 07:30:00+00:00 UTCLiver function panelLFTAST (SGOT)ast3535U/Lbloodblood/plasma/serum1920-81.0
10030362023-07-15 06:45:00+00:00 UTC2023-07-15 07:00:00+00:00 UTC2023-07-15 07:30:00+00:00 UTCLiver function panelLFTALT (SGPT)alt2828U/Lbloodblood/plasma/serum1742-61.0

medication admin continuous Beta

This table captures medications administered at a rate over time, with NO set dose to be given. Examples include vasopressors, sedation, and paralysis drips. Multiple observations capture how the medication rate varies over time.

The medication admin continuous table is a long-form (one medication administration record per) longitudinal table designed for continuous infusions of common ICU medications such as vasopressors and sedation (Boluses of these drugs should be recorded in med_admin_intermittent). Note that it only reflects dose changes of the continuous medication and does not have a specific “end_time” variable to indicate the medication being stopped. The end of a continuous infusion should be recorded as a new row with med_dose = 0 and an appropriate mar_action_name (e.g. “stopped” or “paused”).

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
med_order_idVARCHARMedication order ID. Foreign key to link this table to other medication tablesNo restriction
admin_dttmDATETIMEDate and time when the medicine was administered. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
med_nameVARCHAROriginal med name string recorded in the raw data which often contains concentration e.g. NOREPInephrine 8 mg/250 mLNo restriction
med_categoryVARCHARMaps med_name to a limited set of active ingredients for important ICU medications, e.g. norepinephrineList of continuous medication categories in CLIF
med_groupVARCHARLimited number of ICU medication groups identified by the CLIF consortium, e.g. vasoactivesList of continuous medication groups in CLIF
med_route_nameVARCHARMedicine delivery routee.g. IV, enteral
med_route_categoryVARCHARMaps med_route_name to a standardized list of medication delivery routes. Refer to notes.List of continuous route categories in CLIF
med_doseFLOATQuantity of active drug delivered per unit timeNumeric
med_dose_unitVARCHARUnit of dose in the format [active ingredient quantity]/[time] (e.g., mcg/min, mg/hr, units/hr, mcg/kg/min). Units are not standardized across drugs, and can be weight-based (e.g., mcg/kg/min). Boluses should be mapped to med_admin_intermittentNo restriction
infusion_rateFLOATAbsolute amount of volume administered over time, measured in [volume]/[time] (e.g., mL/hr). In contrast to med_dose which measures [quantity of drug]/[time] or [quantity of drug]/([time]*[body_weight])Numeric
infusion_rate_unitsVARCHARUnit of infusion rate in the format [volume]/[time] (e.g., mL/hr, mL/min)No restriction
mar_action_nameVARCHARMAR (medication administration record) action, e.g. stoppedNo restriction
mar_action_categoryVARCHARMaps mar_action_name to a standardized list of MAR actionsList of continuous action categories in CLIF
mar_action_groupVARCHARMaps mar_action_category to whether the action means the medication was administered or not.administered, not_administered, other

Example:

hospitalization_idadmin_dttmmed_namemed_categorymed_groupmed_route_namemed_route_categorymed_dosemed_dose_unitinfusion_rateinfusion_rate_unitsmar_action_namemar_action_group
7923912123-11-13 12:28:00+00:00 UTCPROPOFOL 10 MG/ML INTRAVENOUS EMULSIONpropofolsedationIntravenousNA75.0000mcg/kg/min45.0mL/hrNew Bagadministered
7923912123-11-13 13:49:00+00:00 UTCREMIFENTANIL CONTINUOUS IV (ANESTHESIA)remifentanilsedationNANA0.0500mcg/kg/min12.5mL/hrNew Bagadministered
7923912123-11-13 14:03:00+00:00 UTCPROPOFOL 10 MG/ML INTRAVENOUS EMULSIONpropofolsedationIntravenousNA0.0000mcg/kg/min0.0mL/hrStoppednot_administered
3709212123-02-12 03:07:00+00:00 UTCPHENYLEPHRINE 5 MG/50 ML (100 MCG/ML) IN 0.9 % SODIUM CHLORIDEphenylephrinevasoactivesIntravenousNA20.0000mcg/min10.0mL/hrNew Bagadministered
3709212123-02-12 03:14:00+00:00 UTCPHENYLEPHRINE 5 MG/50 ML (100 MCG/ML) IN 0.9 % SODIUM CHLORIDEphenylephrinevasoactivesIntravenousNA50.0000mcg/min25.0mL/hrRate Changeadministered
7023442123-04-27 04:30:00+00:00 UTCHEPARIN (PORCINE) 25,000 UNIT/250 ML IN 0.45 % SODIUM CHLORIDEheparinanticoagulationIntravenousNA18.0000Units/kg/hr18.0mL/hrNew Bagadministered

Notes:

  • Include combination medications when mapping medication names to respective categories. Eg. ACETAMIN-CALCIUM-MAG-CAFFEINE ORAL -> acetaminophen
  • Include trial drugs when mapping medication names to respective categories… Eg. ACETAMINOPHEN (IRB 140122) 325 MG ORAL TAB -> acetaminophen

medication admin intermittent Beta

This table captures medications administered as fixed doses at discrete time points. Examples include antibiotics, steroids, and other medications given as boluses or scheduled doses. Each row represents ONE observation for each medication administered.

This table has exactly the same schema as medication_admin_continuous. The consortium decided to separate the medications that are administered intermittently from the continuously administered medications. The mCIDE for medication_category for intermittent meds can be found here.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
med_order_idVARCHARMedication order ID. Foreign key to link this table to other medication tablesNo restriction
admin_dttmDATETIMEDate and time when the medicine was administered. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
med_nameVARCHAROriginal med name string recorded in the raw data which often contains concentration e.g. NOREPInephrine 8 mg/250 mLNo restriction
med_categoryVARCHARMaps med_name to a limited set of active ingredients for important ICU medications, e.g. norepinephrineList of medication categories in CLIF
med_groupVARCHARLimited number of ICU medication groups identified by the CLIF consortium, e.g. vasoactivesList of medication groups in CLIF
med_route_nameVARCHARMedicine delivery routee.g. IV, enteral
med_route_categoryVARCHARMaps med_route_name to a standardized list of medication delivery routesList of intermittent route categories in CLIF
med_doseFLOATQuantity taken in doseNumeric
med_dose_unitVARCHARUnit of dose. e.g. mcg.No restriction
mar_action_nameVARCHARMAR (medication administration record) action, e.g. stoppedNo restriction
mar_action_categoryVARCHARMaps mar_action_name to a standardized list of MAR actionsList of intermittent action categories in CLIF
mar_action_groupVARCHARMaps mar_action_category to whether the action means the medication was administered or not.administered, not_administered, other

Notes:

  • Continuous medications are included in this table when given as boluses
  • Intermittent medications can be given at different rates

microbiology culture Beta

The microbiology culture table is a wide longitudinal table that captures the order and result times of microbiology culture tests, the type of fluid collected, the component of the test, and the organism identified.

Unstable
ColumnTypeDescriptionPermissible Values
patient_idVARCHARUnique identifier for each patient, presumed to be a distinct individual.No restriction
hospitalization_idVARCHARID variable for each patient encounterNo restriction
organism_idVARCHARDistinct numerical identifier that each site creates which links a unique, non-missing organism_category that has a distinct patient_id, hospitalization_id, lab_order_dttm, lab_collect_dttm, lab_result_dttm, and fluid_category with its method_category == culture from the microbiology_culture table to an antibiotic_category and susceptibility_category from the microbiology_susceptibility tableNo restriction
order_dttmDATETIMEDate and time when the test is ordered.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00
collect_dttmDATETIMEDate and time when the specimen is collected.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00
result_dttmDATETIMEDate and time when the results are available.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00
fluid_nameVARCHARFluid name string from the raw data.No restriction.
fluid_categoryVARCHARFluid categories defined according to the NIH common data elements.CDE NIH Infection Site
method_nameVARCHAROriginal method names from the source data.No restriction
method_categoryVARCHARMaps method_name to a standardized list of method categories.culture, gram stain, smear
organism_nameVARCHAROrganism name from the raw data.No restriction.
organism_categoryVARCHARMaps organism_name to the standardized list of organisms under the structure of genus species.Organism species. Check this file for examples.
organism_groupVARCHARMaps organism_category to the standardized list of organisms under the NIH CDE structure.CDE NIH Organism
lab_loinc_codeVARCHARLOINC code.No restriction

Example:

patient_idhospitalization_idorganism_idorder_dttmcollect_dttmresult_dttmfluid_namefluid_categorymethod_namemethod_categoryorganism_nameorganism_categoryorganism_grouplab_loinc_code
12345HOSP12345ORG0012025-06-05 08:15:00+00:002025-06-05 08:45:00+00:002025-06-06 12:00:00+00:00AFB/FUNGAL BLOOD CULTUREBlood/Buffy CoatBlood culturecultureAcinetobacter baumaniiacinetobacter_baumaniiacinetobacter (baumanii, calcoaceticus, lwoffi, other species)
12345HOSP12345ORG0022025-06-05 08:15:00+00:002025-06-05 08:45:00+00:002025-06-06 12:00:00+00:00AFB/FUNGAL BLOOD CULTUREBlood/Buffy CoatBlood culturecultureCandida albicanscandida_albicanscandida albicans
67890HOSP67890ORG0032025-06-10 14:10:00+00:002025-06-10 14:35:00+00:002025-06-11 09:20:00+00:00BRAIN BIOPSY CULTUREBrainTissue culturecultureAspergillus fumigatusaspergillus_fumigatusasperguillus fumigatus

microbiology susceptibility Beta

This table is used to store the susceptibility results of the organisms identified in the Microbiology Culture.

Unstable
ColumnTypeDescriptionPermissible Values
organism_idVARCHARDistinct numerical identifier that each site creates which links a unique, non-missing organism_category that has a distinct patient_id, hospitalization_id, lab_order_dttm, lab_collect_dttm, lab_result_dttm, and fluid_category with its method_category == culture from the microbiology_culture table to an antibiotic_category and susceptibility_category from the microbiology_susceptibility tableNo restriction
antimicrobial_nameVARCHARName of the antimicrobialNo restriction
antimicrobial_categoryVARCHARCategory or class of the antimicrobial testedList of antimicrobial categories in CLIF
sensitivity_nameVARCHARName of the test result used to determine susceptibility (e.g., value of mcg/mL or MIC)No restriction
susceptibility_nameVARCHARName of the sensitivity interpretationNo restriction
susceptibility_categoryVARCHARStandardized category of susceptibility.susceptible, non susceptible, indeterminate, NA

Example:

organism_idantimicrobial_nameantimicrobial_categorysensitivity_namesusceptibility_namesusceptibility_category
1avycex (ceftazidime/avibactam)ceftazidime_avibactam8 MICsusceptiblesusceptible
1amoxicillin clavulanic acidamoxicillin_clavulanate< 0.1 ug/mLsusceptiblesusceptible
1meropenemmeropenemindeterminateindeterminateindeterminate
1ampicillin 500mgampicillin> 0.5 ug/mLresistantNA
2unasynampicillin_sulbactamsusceptible dose dependentsusceptiblesusceptible
2Ertapenemertapenemnot reportedNANA
2Vancomycin (non-Cdiff)vancomycin> 0.25 ug/mLnon susceptible, cautionnon_susceptible

patient Beta

This table contains demographic information about the patient that does not vary between hospitalizations. It is inspired by the OMOP Person table

Unstable
ColumnTypeDescriptionPermissible Values
patient_idVARCHARUnique identifier for each patient, presumed to be a distinct individual.No restriction
race_nameVARCHARPatient race string from source dataNo restriction
race_categoryVARCHARA standardized CDE description of patient race per the US CensusBlack or African American, White, American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, Unknown, Other
ethnicity_nameVARCHARPatient ethnicity string from source dataNo restriction
ethnicity_categoryVARCHARDescription of patient ethnicity per the US census definitionHispanic, Non-Hispanic, Unknown
sex_nameVARCHARPatient’s biological sex as given in the source dataNo restriction
sex_categoryVARCHARPatient biological sexMale, Female, Unknown
birth_dateDATEPatient’s date of birthDate format should be YYYY-MM-DD
death_dttmDATETIMEPatient’s death date, including time. Can be sourced from hospital records, or external sources like state vital records registriesDatetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
language_nameVARCHARPatient’s preferred languageOriginal string from the source data
language_categoryVARCHARMaps language_name to a standardized list of spoken languagesList of language categories in CLIF

Example:

patient_idrace_namerace_categoryethnicity_nameethnicity_categorysex_categorybirth_datedeath_dttmlanguage_namelanguage_category
132424Black or African-AmericanBlack or African AmericanNot Hispanic, Latino/a, or Spanish originNon-HispanicMale2145-05-09NAEnglishEnglish
132384WhiteWhiteNot Hispanic, Latino/a, or Spanish originNon-HispanicFemale2145-03-30NAEnglishEnglish
542367Black or African-AmericanBlack or African AmericanNot Hispanic, Latino/a, or Spanish originNon-HispanicMale2145-01-29NAEnglishEnglish
989862WhiteWhiteNot Hispanic, Latino/a, or Spanish originNon-HispanicFemale2145-11-06NAEnglishEnglish
428035More than one RaceOtherNot Hispanic, Latino/a, or Spanish originNon-HispanicMale2145-10-13NAEnglishEnglish

patient assessments Beta

The patient_assessments table captures various assessments performed on patients across different domains, including neurological status, sedation levels, pain, and withdrawal. The table is designed to provide detailed information about the assessments, such as the name of the assessment, the category, and the recorded values.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
recorded_dttmDATETIMEThe exact date and time when the assessment was recorded, ensuring temporal accuracy. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
assessment_nameVARCHARAssessment Tool Name. The primary name of the assessment tool used (e.g., GCS, NRS, SAT Screen).No restriction
assessment_categoryVARCHARMaps assessment_name to a standardized list of patient assessmentsList of permissible assessment categories in CLIF
assessment_groupVARCHARBroader Assessment Group. This groups the assessments into categories such as Sedation, Neurologic, Pain, etc.List of permissible assessment groups in CLIF
numerical_valueDOUBLENumerical Assessment Result. The numerical result or score from the assessment component.Applicable for assessments with numerical outcomes (e.g., 0-10, 3-15)
categorical_valueVARCHARCategorical Assessment Result. The categorical outcome from the assessment component.Applicable for assessments with categorical outcomes (e.g., Pass/Fail, Yes/No)
text_valueVARCHARTextual Assessment Result. The textual explanation or notes from the assessment component.Applicable for assessments requiring textual data

Example:

hospitalization_idrecorded_dttmassessment_nameassessment_categoryassessment_groupnumerical_valuecategorical_valuetext_value
123452024-12-01 08:15:00+00:00 UTCNUR RA GLASGOW ADULT EYE OPENINGgcs_eyeNeurological4NANA
123452024-12-01 08:15:00+00:00 UTCNUR RA GLASGOW ADULT VERBAL RESPONSEgcs_verbalNeurological5NANA
123452024-12-01 08:15:00+00:00 UTCNUR RA GLASGOW ADULT BEST MOTOR RESPONSEgcs_motorNeurological6NANA
123452024-12-01 08:15:00+00:00 UTCNUR RA GLASGOW ADULT SCORINGgcs_totalNeurological15NANA
678902024-12-01 10:30:00+00:00 UTCBRADEN ASSESSMENTbraden_totalNursing Risk18NANA
678902024-12-01 10:30:00+00:00 UTCSAT SCREENsat_delivery_pass_failSedationNAPassNA

patient procedures Beta

A long table of standardized procedural billing codes associated with the hospitalization, using the Healthcare Common Procedure Coding System (HCPCS). In CLIF version 2.1, the patient_procedures table includes only procedures that were actually performed or completed (not cancelled), and only contains professional billing codes—specifically, CPT codes billed by clinicians (HCPCS Level 1).

Hospital billing i.e., Products, supplies, and services that do not have CPT codes (HCPCS Level 2) are not included in this table.

Additionally, this table contains ICD-10-PCS procedure codes which are not used for clinician billing but can contribute to the calculation of DRGs for hospital reimbursement and can also appear in the hospital_diagnosis table.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
billing_provider_idVARCHARUniquely identifies the billingprovider associated with the procedure.No restriction
performing_provider_idVARCHARUniquely identifies the performing provider associated with the procedure.No restriction
procedure_codeVARCHAREncoded procedure identifier.Valid CPT, ICD-10-PCS OR HCPCS code
procedure_code_formatVARCHARCode format used.CPT, ICD10PCS, HCPCS
procedure_billed_dttmDATETIMEDate and time the procedure was billed (may differ from actual procedure time). All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)

Example:

hospitalization_idbilling_provider_idperforming_provider_idprocedure_codeprocedure_code_formatprocedure_billed_dttm
HOSP1001BP123PP45636556CPT2024-01-01 08:00:00+00:00 UTC
HOSP1001BP123PP78932551CPT2024-01-01 10:00:00+00:00 UTC
HOSP1002BP234PP890G0009HCPCS2024-01-05 09:30:00+00:00 UTC
HOSP1002BP234PP890G0008HCPCS2024-01-05 11:00:00+00:00 UTC
HOSP1003BP345PP90136620CPT2024-01-10 07:00:00+00:00 UTC

position Beta

The position table is a long form (one position per row) longitudinal table that captures all documented position changes of the patient. The table is designed for the explicit purpose of constructing the position_category CDE and identifying patients in prone position.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
recorded_dttmDATETIMEDate and time when the vital is recorded. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00
position_nameVARCHARDescription of the position from the source data. This field is not used for analysis.No restriction
position_categoryVARCHARMaps position_name to either prone or not prone.prone, not_prone

Example:

hospitalization_idrecorded_dttmposition_nameposition_category
842123-06-20 00:00:00+00:00 UTCSupine–turn Rnot_prone
842123-06-20 06:00:00+00:00 UTCSupine–turn Lnot_prone
842123-06-20 12:00:00+00:00 UTCSupine–backnot_prone
842123-06-20 16:00:00+00:00 UTCSupine–turn Rnot_prone
842123-06-20 20:00:00+00:00 UTCSupine–back;Supine–turn intolerantnot_prone
842123-06-20 22:00:00+00:00 UTCSupine–turn intolerant,microturn Lnot_prone
842123-06-20 00:00:00+00:00 UTCSupine–turn intolerant,microturn L;Supine–backnot_prone
842123-06-20 01:10:00+00:00 UTC30 Degreesnot_prone

respiratory support Beta

The respiratory support table is a wider longitudinal table that captures simultaneously recorded ventilator settings and observed ventilator parameters. The table is designed to capture the most common respiratory support devices and modes used in the ICU. It will be sparse for patients who are not on mechanical ventilation.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
recorded_dttmDATETIMEDate and time when the device settings and/or measurement was recorded. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00
device_nameVARCHARRaw string of the device.No restriction
device_idVARCHARUnique ID of the individual physical device used (e.g. ventilator ACZ3RV91), enables linkage to continuous waveform data. Distinct from vent_brand_name, which stores the brand or model.No restriction
device_categoryVARCHARMaps device_name to a standardized list of respiratory support device categoriesIMV, NIPPV, CPAP, High Flow NC, Face Mask, Trach Collar, Nasal Cannula, T Piece, Room Air, Other
vent_brand_nameVARCHARVentilator model name when device_category is IMV or NIPPVOptional
mode_nameVARCHARRaw string of ventilation mode (e.g., CMV volume control)No restriction
mode_categoryVARCHARStandardized list of modes of mechanical ventilationAssist Control-Volume Control, Pressure Control, Pressure-Regulated Volume Control, SIMV, Pressure Support/CPAP, Volume Support, Blow by, Other
tracheostomyINTIndicates if tracheostomy is present0 = No, 1 = Yes
fio2_setFLOATFraction of inspired oxygen set (e.g., 0.21)No restriction, see Expected _set values for each device_category and mode_category
lpm_setFLOATLiters per minute of supplemental oxygen set for patients NOT on positive pressure ventilationNo restriction, see Expected _set values for each device_category and mode_category
tidal_volume_setFLOATTidal volume set (in mL)No restriction, see Expected _set values for each device_category and mode_category
resp_rate_setFLOATRespiratory rate set (in bpm)No restriction, see Expected _set values for each device_category and mode_category
pressure_control_setFLOATPressure control set (in cmH2O)No restriction, see Expected _set values for each device_category and mode_category
pressure_support_setFLOATPressure support set (in cmH2O)No restriction, see Expected _set values for each device_category and mode_category
flow_rate_setFLOATFlow rate of air delivered to patients on non-invasive devives (in lpm)No restriction, see Expected _set values for each device_category and mode_category
peak_inspiratory_pressure_setFLOATPeak inspiratory pressure set (in cmH2O). This is equivalent to inspiratory positive airway pressure (IPAP) in non-invasive ventilation modes, and is not used in invasive ventilation modes.No restriction, see Expected _set values for each device_category and mode_category
inspiratory_time_setFLOATInspiratory time set (in seconds)No restriction, see Expected _set values for each device_category and mode_category
peep_setFLOATPositive-end-expiratory pressure set (in cmH2O)No restriction, see Expected _set values for each device_category and mode_category
tidal_volume_obsFLOATObserved tidal volume (in mL)No restriction
resp_rate_obsFLOATObserved respiratory rate (in bpm), as measured and recorded by the ventilator or non-invasive device. This value should not be pulled from the general vitals table, but should reflect measurements from the respiratory support device itself.No restriction
plateau_pressure_obsFLOATObserved plateau pressure (in cmH2O)No restriction
peak_inspiratory_pressure_obsFLOATObserved peak inspiratory pressure (in cmH2O)No restriction
peep_obsFLOATObserved PEEP (in cmH2O)No restriction
minute_vent_obsFLOATObserved minute ventilation (in liters)No restriction
mean_airway_pressure_obsFLOATObserved mean airway pressureNo restriction

Notes:

Expected setting values for each device_category and mode_category

  • device_category == “IMV”
ventilator_settingAssist Control-Volume ControlPressure Support/CPAPPressure ControlPressure-Regulated Volume ControlSIMVVolume Support
fio2_setEEEEEE
tidal_volume_setEEPE
resp_rate_setEEEE
pressure_control_setEP
pressure_support_setEE
flow_rate_setPP
inspiratory_time_setPEP
peep_setEEEEEE

E = Expected ventilator setting for the mode, P = possible ventilator setting for the mode.

  • device_category == “NIPPV”

mode_category is Pressure Support/CPAP and the fio2_set, peep_set, and either pressure_support_set OR peak_inspiratory_pressure_set (IPAP) is required.

  • device_category == “CPAP”

mode_category is Pressure Support/CPAP and the fio2_set and peep_set are required.

  • device_category == “High Flow NC”

mode_category is NA and the fio2_set and lpm_set are required.

  • device_category == “Face Mask”

mode_category is NA
lpm_set is required.
fio2_set is possible.

  • device_category == “Trach Collar” or “Nasal Cannula”

mode_category is NA
lpm_set is required

Example:

hospitalization_idrecorded_dttmdevice_namedevice_iddevice_categorymode_namemode_categoryvent_brand_nametracheostomyfio2_setlpm_settidal_volume_setresp_rate_setpressure_control_setpressure_support_setflow_rate_settidal_volume_obsresp_rate_obsplateau_pressure_obspeak_inspiratory_pressure_obspeep_obsminute_vent_obsmean_airway_pressure_obs
123452024-12-01 08:00:00+00:00 UTCVentilatorDEV001IMVCMV Volume CtrlAssist Control-Volume ControlVent A10.5040500181555045018202559.012.0
123452024-12-01 09:00:00+00:00 UTCVentilatorDEV001IMVSIMVSIMVVent A10.45354802018855470202128610.514.0
678902024-12-01 10:30:00+00:00 UTCHFNCDEV002High Flow NCN/AOtherN/A00.3060NANANANA60NANANANANANANA
678902024-12-01 11:00:00+00:00 UTCCPAPDEV003CPAPCPAPPressure Support/CPAPCPAP X00.4050NANANA10NANANANANA8NANA

vitals Beta

The vitals table is a long-form (one vital sign per row) longitudinal table.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
recorded_dttmDATETIMEDate and time when the vital is recorded. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
vital_nameVARCHARDescription of the flowsheet measure from the source data. Not used for analysis.No restriction
vital_categoryVARCHARMaps vital_name to a list of standard vital sign categories.temp_c, heart_rate, sbp, dbp, spo2, respiratory_rate, map, height_cm, weight_kg
vital_valueFLOATRecorded value of the vital. Measurement unit should match the vital category.temp_c = Celsius, height_cm = Centimeters, weight_kg = Kg, map = mmHg, spo2 = %. No unit for heart_rate, sbp, dbp, respiratory_rate
meas_site_nameVARCHARSite where the vital is recorded. Optional field with no associated category.No restriction

Example:

hospitalization_idrecorded_dttmvital_namevital_categoryvital_valuemeas_site_name
200100122024-12-01 08:00:00+00:00 UTCHEIGHTheight_cm170.0unspecified
200100122024-12-01 08:15:00+00:00 UTCWEIGHTweight_kg70.0unspecified
200100122024-12-01 08:30:00+00:00 UTCPULSEheart_rate72.0unspecified
200100122024-12-01 08:45:00+00:00 UTCBLOOD PRESSURE (SYSTOLIC)sbp120.0unspecified
200100122024-12-01 08:45:00+00:00 UTCBLOOD PRESSURE (DIASTOLIC)dbp80.0unspecified
200100122024-12-01 08:50:00+00:00 UTCRESPIRATORY RATErespiratory_rate16.0unspecified
200100122024-12-01 09:00:00+00:00 UTCTEMPERATUREtemp_c36.8unspecified
200100122024-12-01 09:15:00+00:00 UTCSPO2spo298.0unspecified
200100132024-12-01 09:30:00+00:00 UTCMEAN ARTERIAL PRESSURE (MAP)map85.0arterial

Concept Tables

A planned future CLIF table that has yet to be used in a federated project. The table structure and CDE elements are in draft form. Permissible values of category variables may still need to be defined. Seeking conceptual feedback. Significant changes to all aspects of the table are possible.

clinical trial Concept

This table captures whether a patient was enrolled in any clinical trial during their hospitalization. It enables longitudinal tracking of trial participation, including trial identifiers, arm assignment, and key consent, randomization, and withdrawal timestamps. This structure supports research into the effects of experimental therapies and interventions on patient outcomes.

Unstable
ColumnTypeDescriptionPermissible Values
participant_idVARCHARUnique identifier for each clinical trial participant from clinical trial management software, used for pipelines to electronic data capture software. NOT the same as patient_idNo restriction
patient_idVARCHARUnique identifier for each patient, presumed to be a distinct individual.No restriction
hospitalization_idVARCHARID variable for each patient encounter.No restriction
trial_idVARCHARUnique identifier for the clinical trial (e.g., institutional trial ID, NCT number).No restriction
trial_nameVARCHARDescriptive name of the clinical trial.No restriction
arm_idVARCHARIdentifier indicating which arm of the trial the participant was enrolled in.No restriction
consent_dttmDATETIMETimestamp of participant consent for the clinical trial. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
enrollment_dttmDATETIMETimestamp of participant enrollment in the clinical trial. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
randomized_dttmDATETIMETimestamp of participant randomization in the clinical trial. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
withdrawal_dttmDATETIMETimestamp of trial withdrawal if applicable. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)

Example:

participant_idpatient_idhospitalization_idtrial_idtrial_namearm_idconsent_dttmenrollment_dttmrandomized_dttmwithdrawal_dttm
PT1234567890234567T-001Early Vasopressor in Septic Shock (EVSS)Arm A2024-06-01 10:15:00+00:00 UTC2024-06-01 14:00:00+00:00 UTC2024-06-01 15:00:00+00:00 UTC
PT2345678901345678NCT04321Lung Protective VentilationStandard2024-06-11 12:00:00+00:00 UTC2024-06-11 13:15:00+00:00 UTC
PT3456789012456789T-017Multi-center COVID-19 Anticoagulation TrialHigh Dose2024-07-15 09:10:00+00:00 UTC2024-07-15 09:45:00+00:00 UTC2024-07-15 10:00:00+00:00 UTC2024-07-20 08:30:00+00:00 UTC

ecmo mcs Concept

The ECMO/MCS table is a wider longitudinal table that captures the start and stop times of ECMO/MCS support, the type of device used, and the work rate of the device.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
recorded_dttmDATETIMEDate and time when the device settings and/or measurement was recordedDatetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
device_nameVARCHARName of the ECMO/MCS device used including brand information, e.g. CentrimagNo restriction
device_categoryVARCHARMaps device_name to a standardized mCIDEList of device categories in CLIF and outlier thresholds by device category
mcs_groupVARCHARMaps device_category to a standardized mCIDE of MCS typesList of MCS groups in CLIF
ecmo_configuration_categoryVARCHARCategorical variable designating the ECMO configuration type as defined by the cannulation strategyvv, va, va_v, vv_a, etc.
control_parameter_nameVARCHARString that captures the measure of work rate of the device, e.g., RPMs, Impella Power, etc.No restriction
control_parameter_categoryVARCHARMaps control_parameter_name to a standardized list of control parameter categoriesList of control parameter categories in CLIF
control_parameter_valueFLOATThe value of the control parameter (numeric).Numeric values
flowFLOATBlood flow in L/min.Numeric values in L/min
sweep_setFLOATGas flow (L/min) set. Applies to ECMO only.Numeric values in L/min
fdO2_setFLOATFraction of delivered oxygen set. Applies to ECMO only.Numeric values (0-1)

Example:

hospitalization_idrecorded_dttmdevice_namedevice_categorymcs_groupecmo_configuration_categorycontrol_parameter_namecontrol_parameter_categorycontrol_parameter_valueflowsweep_setfdO2_set
210022024-02-18 09:10:00+00:00 UTCImpella 5.5impella_5_5impella_lvadNULLP5impella_performance54.8NULLNULL
210032024-02-22 11:40:00+00:00 UTCAbbott IABPiabpiabpNULLIABP Ratioiabp_ratio3NULLNULLNULL
210042024-02-23 14:30:00+00:00 UTCHeart Mate III LVADheartmate_3durable_lvadNULLRPMsrpm42003.5NULLNULL
210052024-02-27 16:25:00+00:00 UTCCentrimag ECMO VVcentrimag_ecmoecmovvRPMsrpm350031NULL

intake output Concept

The intake_output table is long form table that captures the times intake and output events were recorded, the type of fluid administered or recorded as “out”, and the amount of fluid.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
intake_dttmDATETIMEDate and time of intake. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
fluid_nameVARCHARName of the fluid administered.No restriction
amountDOUBLEAmount of fluid administered (in mL)Numeric values in mL
in_out_flagINTIndicator for intake or output (1 for intake, 0 for output)0 = Output, 1 = Intake

Example:

hospitalization_idintake_dttmfluid_nameamountin_out_flag
10012024-01-01 08:00:00+00:00 UTCNormal Saline5001
10012024-01-01 10:30:00+00:00 UTCUrine3000
10022024-01-05 09:15:00+00:00 UTCDextrose2501
10022024-01-05 14:00:00+00:00 UTCUrine4000
10032024-01-10 07:45:00+00:00 UTCLactated Ringer’s6001
10032024-01-10 12:00:00+00:00 UTCDrainage2000

invasive hemodynamics Concept

The invasive_hemodynamics table records invasive hemodynamic measurements during a patient’s hospitalization. These measurements represent pressures recorded via invasive monitoring and are expressed in millimeters of mercury (mmHg).

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
recorded_dttmDATETIMEThe date and time when the measurement was recorded. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
measure_nameVARCHARDescription of the site or context of the invasive hemodynamic measurement.Free text (e.g., Right Atrium)
measure_categoryVARCHARCategorical variable specifying the type of invasive hemodynamic measurement.cvp, ra, rv, pa_systolic, pa_diastolic, pa_mean, pcwp, cardiac_output_thermodilution , cardiac_output_fick
measure_valueDOUBLEThe numerical value of the invasive hemodynamic measurement in mmHg.Positive decimal values (e.g., 5.00, 25.65)

Notes:

  • All measure_value entries should be recorded in mmHg.
  • The measure_category field ensures standardization of invasive hemodynamic data:
    1. CVP - Central Venous Pressure
    2. RA - Right Atrial Pressure
    3. RV - Right Ventricular Pressure
    4. PA_systolic - Pulmonary Artery Systolic Pressure
    5. PA_diastolic - Pulmonary Artery Diastolic Pressure
    6. PA_mean - Pulmonary Artery Mean Pressure
    7. PCWP - Pulmonary Capillary Wedge Pressure


Example:

hospitalization_idrecorded_dttmmeasure_namemeasure_categorymeasure_value
123452024-12-01 08:30:00+00:00 UTCCVPCVP12.50
123452024-12-01 09:00:00+00:00 UTCPulmonary Artery-SysPA_systolic25.00
123452024-12-01 09:30:00+00:00 UTCWedgePCWP18.75

key icu orders Concept

The key_icu_orders table captures key orders related to physical therapy (PT) and occupational therapy (OT) during ICU stays. It includes details about the hospitalization, the timing of the order, the specific name of the order, its category, and the status of the order (completed or sent).

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
order_dttmDATETIMEDate and time when the order was placed. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
order_nameVARCHARName of the specific order (e.g., PT Evaluation, OT Treatment).No restriction
order_categoryVARCHARCategory of the order.Under-development. Some examples include: pt_evaluation, pt_treat, ot_evaluation, ot_treat
order_status_nameVARCHARStatus of the order.sent, completed

Example:

hospitalization_idorder_dttmorder_nameorder_categoryorder_status_name
123452024-12-15 10:00:00+00:00 UTCPT Initial EvaluationPT_evaluationcompleted
678902024-12-16 14:30:00+00:00 UTCOT Follow-up TreatmentOT_treatsent
543212024-12-16 08:00:00+00:00 UTCPT Mobility SessionPT_treatcompleted
987652024-12-15 11:15:00+00:00 UTCOT Cognitive AssessmentOT_evaluationsent

medication orders Concept

This table records the ordering (not administration) of medications. The table is in long form (one medication order per row) longitudinal table. Linkage to the medication_admin_continuous and medication_admin_intermittent tables is through the med_order_id field.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
med_order_idVARCHARUnique identifier for each medication orderNo restriction
order_start_dttmDATETIMEDate and time when the medication order was initiated. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
order_end_dttmDATETIMEDate and time when the medication order ended or was discontinued. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
ordered_dttmDATETIMEDate and time when the medication was actually ordered. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
med_nameVARCHARName of the medication orderedNo restriction
med_categoryVARCHARMaps med_name to a list of permissible medication namesCombined CDE of medication_admin_continuous and medication_admin_intermittent
med_groupVARCHARLimited number of medication groups identified by the CLIF consortiumNo restriction
med_order_status_nameVARCHARStatus of the medication order, e.g. held, givenNo restriction
med_order_status_categoryVARCHARMaps med_order_status_name to a standardized list of medication order statusesUnder-development
med_route_nameVARCHARRoute of administration for the medicationNo restriction. Examples include Oral, Intravenous
med_doseDOUBLEDosage of the medication orderedNumeric
med_dose_unitVARCHARUnit of measurement for the medication dosageExamples include mg, mL, units
med_frequencyVARCHARFrequency with which the medication is administered, as per the orderExamples include Once Daily, Every 6 hours
prnBOOLEANIndicates whether the medication is to be given as needed (PRN)0 = No, 1 = Yes

Example:

hospitalization_idmed_order_idorder_start_dttmorder_end_dttmordered_dttmmed_namemed_categorymed_groupmed_order_status_namemed_order_status_categorymed_route_namemed_dosemed_dose_unitmed_frequencyprn
123454567892023-10-01 14:00:00+00:00 UTC2023-10-02 14:00:00+00:00 UTC2023-10-01 13:30:00+00:00 UTCNorepinephrine 8 mg/250 mLnorepinephrinevasoactivesactiveongoingIntravenous8.0mg/mLContinuous0
123464567902023-10-01 16:00:00+00:00 UTC2023-10-02 10:00:00+00:00 UTC2023-10-01 15:30:00+00:00 UTCVancomycin 1 g IVvancomycinantibioticsactiveongoingIntravenous1.0gEvery 12 hours0
123474567912023-10-02 08:00:00+00:00 UTC2023-10-03 08:00:00+00:00 UTC2023-10-02 07:30:00+00:00 UTCFurosemide 40 mg IVfurosemidediureticsdiscontinueddiscontinuedIntravenous40.0mgOnce Daily0
123484567922023-10-02 12:00:00+00:00 UTC2023-10-02 18:00:00+00:00 UTC2023-10-02 11:45:00+00:00 UTCInsulin Regular 100 units/mL SCinsulinendocrineheldheldSubcutaneous100.0units/mLAs Needed1
123494567932023-10-03 08:00:00+00:00 UTC2023-10-03 20:00:00+00:00 UTC2023-10-03 07:30:00+00:00 UTCAcetaminophen 1 g POacetaminophenanalgesicsactiveongoingOral1.0gEvery 6 hours0
123504567942023-10-03 10:00:00+00:00 UTC2023-10-03 18:00:00+00:00 UTC2023-10-03 09:45:00+00:00 UTCHeparin 5,000 units SCheparinanticoagulantactiveongoingSubcutaneous5000.0unitsEvery 8 hours0
123514567952023-10-03 14:00:00+00:00 UTC2023-10-03 22:00:00+00:00 UTC2023-10-03 13:30:00+00:00 UTCMorphine Sulfate 2 mg IVmorphineanalgesicsactiveongoingIntravenous2.0mgAs Needed1
123524567962023-10-03 20:00:00+00:00 UTC2023-10-04 08:00:00+00:00 UTC2023-10-03 19:45:00+00:00 UTCDexamethasone 10 mg IVdexamethasonesteroidsactiveongoingIntravenous10.0mgOnce Daily0

microbiology nonculture Concept

The microbiology non-culture table is a wide longitudinal table that captures the order and result times of non-culture microbiology tests, the type of fluid collected, the component of the test, and the result of the test.

Unstable
ColumnTypeDescriptionPermissible Values
patient_idVARCHARUnique identifier for each patient, presumed to be a distinct individual.No restriction
hospitalization_idVARCHARID variable for each patient encounterNo restriction
result_dttmDATETIMEDate and time when the non-culture result was obtained. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00
collect_dttmDATETIMEDate and time when the sample was collected. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00
order_dttmDATETIMEDate and time when the test was ordered. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00
fluid_nameVARCHARName of the fluid sample.No restriction
fluid_categoryVARCHARFluid categories defined according to the NIH common data elements.CDE NIH Infection Site
method_nameVARCHAROriginal method names from the source data.No restriction
method_categoryVARCHARMaps method_name to a standardized list of method categories.pcr
micro_order_nameVARCHARString name of microbiology non-culture test.No restriction
organism_categoryVARCHARMaps the organism name in micro_order_name to the standardized list of organisms under the structure of genus species.Organism species. Check this file for examples.
organism_groupVARCHARMaps organism_category to the standardized list of organisms under the NIH CDE structure.CDE NIH Organism
result_nameVARCHARResult name from the raw data.No restriction
result_categoryVARCHARCategory of the test result.Check list of result categories
reference_lowDOUBLEReference low value.No restriction
reference_highDOUBLEReference high value.No restriction
result_unitsVARCHARUnit of the test result.No restriction
lab_loinc_codeVARCHARLOINC code.No restriction

Example:

patient_idhospitalization_idorder_dttmcollect_dttmresult_dttmfluid_namefluid_categorymethod_namemethod_categorymicro_order_nameorganism_categoryorganism_groupresult_nameresult_categoryreference_lowreference_highresult_unitslab_loinc_code
1121212025-06-15 09:05:00+00:002025-06-15 09:30:00+00:002025-06-15 13:45:00+00:00BLOODblood/buffy coatPCRpcrneisseria quantitative pcr, bloodneisseria_spneisseria (gonorrhoea, meningitidis, other species)100,000 copies/uL of neisseria detecteddetectedcopies/mL39528-5
2323322025-06-16 11:15:00+00:002025-06-16 11:40:00+00:002025-06-16 15:25:00+00:00cerebrospinal fluidmeninges and csfPCRpcrcsf hsv pcrherpes_simplex_virusherpes simplex (hsv1, hsv2)no herspes simplex DNA measurednot_detectedIU/mL16954-2
2323322025-06-17 10:00:00+00:002025-06-17 10:20:00+00:002025-06-17 14:05:00+00:00fecesfeces/stoolPCRpcrstool c. diff toxinclostridioides_difficileclostridium difficiledefault in test for C. difficile toxin analysisindeterminatecopies/mL34712-0

patient diagnosis Concept

The patient_diagnosis table provides a record of all diagnoses assigned to a patient.

Unstable
ColumnTypeDescriptionPermissible Values
patient_idVARCHARUnique identifier for each patient, presumed to be a distinct individual.No restriction
hospitalization_idVARCHARID variable for each patient encounterNo restriction
diagnosis_codeVARCHARICD-10-CM from clinical documentationValid ICD-10-CM code
diagnosis_code_formatVARCHARICD10CM (clinical data typically ICD-10 only)ICD10CM
source_typeVARCHARSource of diagnosis: problem_list, medical_history, encounter_dx (optional)problem_list, medical_history, encounter_dx
start_dttmDATETIMEWhen condition started (user-defined)Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
end_dttmDATETIMEWhen condition ended (NULL if ongoing)Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)

Example:

patient_idhospitalization_iddiagnosis_codediagnosis_code_formatsource_typestart_dttmend_dttm
PAT1001HOSP1001I10ICD10CMproblem_list2024-01-01 08:00:00+00:00NULL
PAT1001HOSP1001E11.9ICD10CMencounter_dx2024-01-01 08:00:00+00:002024-01-10 12:00:00+00:00
PAT1002HOSP1002J18.9ICD10CMmedical_history2024-01-05 09:30:00+00:00NULL
PAT1003HOSP1003N17.9ICD10CMencounter_dx2024-01-10 07:00:00+00:002024-01-15 10:00:00+00:00

place based index Concept

The place_based_index table is designed to store geospatial or community-level indices linked to a patient’s hospitalization.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
index_nameVARCHARThe name of the index (e.g., Area Deprivation Index, Social Vulnerability Index).No restriction
index_valueDOUBLEThe numerical value of the index for the given hospitalization.Numeric
index_versionVARCHARVersion of the index used (e.g., ADI 2019, SVI 2020).No restriction

Notes:
Source indices from validated and publicly available datasets.

Example:

hospitalization_idindex_nameindex_valueindex_version
1001Area Deprivation Index85.2ADI 2019
1001Social Vulnerability Index0.72SVI 2020
1002Area Deprivation Index67.5ADI 2019

provider Concept

Continuous start stop record of every provider who cared for the patient.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
provider_idVARCHARUnique identifier for each provider. This represents individual healthcare providersNo restriction
start_dttmDATETIMEDate and time when the provider’s care or involvement in the patient’s case began. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
stop_dttmDATETIMEDate and time when the provider’s care or involvement in the patient’s case ended. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
provider_role_nameVARCHARThe original string describing the role or specialty of the provider during the hospitalizationNo restriction
provider_role_categoryVARCHARMaps provider_role_name to list of standardized provider rolesUnder development

Example:

hospitalization_idstart_dttmstop_dttmprovider_role_nameprovider_role_category
10010142023-05-01 08:00:00+00:00 UTC2023-05-01 20:00:00+00:00 UTCAttending PhysicianAttending
10010142023-05-01 08:00:00+00:00 UTC2023-05-02 08:00:00+00:00 UTCResident PhysicianResident
10010142023-05-01 08:00:00+00:00 UTC2023-05-03 08:00:00+00:00 UTCNurse PractitionerNurse Practitioner
10020252023-06-10 09:00:00+00:00 UTC2023-06-10 21:00:00+00:00 UTCCritical Care SpecialistCritical Care
10020252023-06-10 09:00:00+00:00 UTC2023-06-11 09:00:00+00:00 UTCRespiratory TherapistRespiratory Therapy
10030362023-07-15 07:30:00+00:00 UTC2023-07-15 19:30:00+00:00 UTCAttending PhysicianAttending
10030362023-07-15 07:30:00+00:00 UTC2023-07-16 07:30:00+00:00 UTCCharge NurseNurse
10040472023-08-20 10:00:00+00:00 UTC2023-08-20 22:00:00+00:00 UTCPhysical TherapistTherapy

therapy details Concept

The therapy_details table is a wide longitudinal table that captures the details of therapy sessions. The table is designed to capture and categorize the most common therapy elements used in the ICU.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
session_start_dttmDATETIMEDate and time when the therapy session started. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
therapy_element_nameVARCHARName of the therapy element.No restriction
therapy_element_categoryVARCHARCategory of the therapy element.No restriction
therapy_element_valueVARCHARValue associated with the therapy element.No restriction

Example:

hospitalization_idsession_start_dttmtherapy_element_nametherapy_element_categorytherapy_element_value
10012024-01-01 08:00:00+00:00 UTCPhysical TherapyRehabilitation45.0
10012024-01-01 10:00:00+00:00 UTCRespiratory TherapyRespiratory Support3.0
10022024-01-05 09:30:00+00:00 UTCOccupational TherapyRehabilitation60.0
10022024-01-05 11:00:00+00:00 UTCSpeech TherapyRehabilitation30.0
10032024-01-10 07:00:00+00:00 UTCVentilation SupportRespiratory Support2.5

transfusion Concept

This table provides detailed information about transfusion events linked to specific hospitalizations.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounterNo restriction
transfusion_start_dttmDATETIMEThe date and time the transfusion of the blood component began. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
transfusion_end_dttmDATETIMEThe date and time the transfusion of the blood component ended. All datetime variables must be timezone-aware and set to UTC.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
component_nameVARCHARThe name of the blood component transfused.E.g., Red Blood Cells, Plasma, Platelets
attribute_nameVARCHARAttributes describing modifications to the component.E.g., Leukocyte Reduced, Irradiated
volume_transfusedDOUBLEThe volume of the blood component transfused.Numeric, e.g., 300
volume_unitsVARCHARThe unit of measurement for the transfused volume.E.g., mL
product_codeVARCHARISBT 128 Product Description Code representing the specific blood product.E.g., E0382

Example:

hospitalization_idtransfusion_start_dttmtransfusion_end_dttmcomponent_nameattribute_namevolume_transfusedvolume_unitsproduct_code
1234562024-12-03 08:30:00+00:00 UTC2024-12-03 10:00:00+00:00 UTCRed Blood CellsLeukocyte Reduced300mLE0382
7890122024-12-04 14:00:00+00:00 UTC2024-12-04 16:30:00+00:00 UTCPlateletsIrradiated250mLP0205
4567892024-12-05 12:15:00+00:00 UTC2024-12-05 13:45:00+00:00 UTCPlasma200mLF0781

validated diagnosis Concept

The validated_diagnosis table captures clinician-validated diagnostic labels for research purposes. This table is designed for studies requiring confirmed diagnoses through chart review or consensus adjudication, enabling high-quality phenotyping for research cohorts.

Unstable
ColumnTypeDescriptionPermissible Values
hospitalization_idVARCHARID variable for each patient encounter.No restriction
diagnosis_nameVARCHARFree-text name of the diagnosis (clinician-entered or source name).No restriction
diagnosis_categoryVARCHARMaps diagnosis_name to a standardized list of diagnosis categories.e.g., sepsis, arf, aki, stroke, etc.
diagnosis_start_dttmDATETIMEDate and time when the diagnosis is deemed clinically active.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
diagnosis_end_dttmDATETIMEDate and time when the diagnosis was resolved or deemed no longer active (optional).Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)
validation_methodVARCHARMethod of clinician validation used for this label.manual_chart_review, consensus_panel, automated_with_review
diagnosis_statusVARCHARFinal reviewer determination.confirmed, ruled_out, uncertain
reviewer_idVARCHARAnonymized ID of the clinician or reviewer responsible for this label.No restriction
review_timestampDATETIMETime when the label was finalized or adjudicated.Datetime format should be YYYY-MM-DD HH:MM:SS+00:00 (UTC)

Notes:

  • This table is intended for research diagnoses that have undergone validation, not routine clinical diagnoses
  • The diagnosis_category should map to a controlled vocabulary appropriate for the research context
  • Multiple reviewers can validate the same diagnosis by creating separate rows with different reviewer_id values

Example:

hospitalization_iddiagnosis_namediagnosis_categorydiagnosis_start_dttmdiagnosis_end_dttmvalidation_methoddiagnosis_statusreviewer_idreview_timestamp
12345Septic shocksepsis2024-12-01 08:00:00+00:00 UTC2024-12-03 14:00:00+00:00 UTCmanual_chart_reviewconfirmedREV0012024-12-15 10:30:00+00:00 UTC
12345Acute kidney injuryaki2024-12-01 12:00:00+00:00 UTCNULLconsensus_panelconfirmedREV0022024-12-16 09:00:00+00:00 UTC
67890Acute respiratory failurearf2024-11-28 06:00:00+00:00 UTC2024-11-30 18:00:00+00:00 UTCautomated_with_reviewuncertainREV0012024-12-10 14:15:00+00:00 UTC
54321Ischemic strokestroke2024-12-05 02:30:00+00:00 UTCNULLmanual_chart_reviewruled_outREV0032024-12-18 11:45:00+00:00 UTC

Future Proposed Tables

These are tables without any defined structure that the consortium has not yet committed to implementing.

Clinical Decision Support: This table will capture the actions of clinical decision support tools embedded in the EHR. The table will have the following fields: cds_name, cds_category, cds_value, cds_trigger_ddtm.