CLIF-2.0 Data Dictionary

Below is the entity-relationship diagram (ERD) that provides an overview of the relational CLIF database structure.

Relational CLIF (RCLIF) is a database that is organized into clinically relevant column categories - demographics, objective measures, respiratory support, orders, and inputs-outputs. Below are sample templates for each table in R-CLIF. Here you can find detailed descriptions of each table and their fields.

You can use our custom GPT- CLIF Assistant to learn more about CLIF and develop analysis scripts.

CLIF maturity

CLIF is still under development and some parts of the format are more mature than others. CLIF will also need to evolve as the set of minimum Common Data Elements for studying critical illness expands or changes over time.

The consortium has two different maturity concepts: one for the overall ER model and one for the individual tables.

Overall Maturity Level for CLIF

• Experimental: Majority of critical illness and hospital course not represented in Entity-Relationship (ER) model, expect frequent breaking changes.
• Beta: Core ER model complete and breaking changes to the existing structure unlikely. Actively seeking feedback about new tables to add to the ER model to fully capture critical illness.
• Stable: Tested and recommended for general use. EHR data not currently represented in CLIF outside the scope of the format.
• Mature: Widely adopted across majority of consortium sites with majority of tables in stable or mature (see maturity levels for CLIF Tables). ER model very stable.
• Deprecated: No longer maintained.

The entity-relationship model for this project is currently at the Beta level for adult patients in a general medical intensive care unit. Major breaking changes to the existing structure are unlikely. The consortium is actively seeking feedback about new tables and fields to add to the ER model to achieve the goal of representing developing a minimum Common ICU Data Elements (mCIDE)

For pediatric patients, CLIF is in the Experimental maturity phase. CLIF is also Experimental for adult patients in specialty ICUs (e.g. cardiac intensive care unit, surgical intensive care unit, and neurointensive care unit).

Maturity Levels for CLIF Tables

There are two critical maturity elements for each CLIF table: 1) field structure and 2) Common ICU data Element development. Each CLIF table has one or more consortium physician-data scientists who are responsible for table design.

• Concept: Placeholder for future CLIF table. Majority of table structure and CDE elements incomplete. Expect breaking changes.
• Beta: Table structure and field names complete, but not fully tested. CDE for category variables underdevelopment. Actively seeking feedback.
• Stable: Tested and recommended for general use. CDE stable with permissible values for all category variables precisely defined and locked. Fully implemented at multiple consortium sites and used in a peer-reviewed publication.
• Mature: Adopted across a majority of the CLIF consortium sites and very stable.
• Deprecated: No longer maintained.

The CLIF-1.0 data dictionary is available here and is now deprecated

General inpatient tables

The data in these tables are typically of most electronic data warehouse systems and are not specific to critical illness. The CLIF versions of these general tables are designed to ensure that they clearly represent the minimum set of data required for critical illness research. Whenever possible, CLIF seeks compatibility with existing EHR data standards.

Patient

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

Variable Name	Data Type	Definition	Permissible Values
patient_id	VARCHAR	Unique identifier for each patient. This is presumed to be a distinct individual.
race_name	VARCHAR	Patient race string from source data	No restriction
race_category	VARCHAR	A standardized CDE description of patient’s race per the US Census permissible values. The source data may contain different strings for race.	Black or African American, White, American Indian or Alaska Native, Asian, Native Hawaiian or Other Pacific Islander, Unknown, Other
ethnicity_name	VARCHAR	Patient ethnicity string from source data	No restriction
ethnicity_category	VARCHAR	Description of patient’s ethnicity per the US census definition. The source data may contain different strings for ethnicity.	Hispanic, Non-Hispanic, Unknown
sex_name	VARCHAR	Patient’s biological sex as given in the source data.	No restriction
sex_category	VARCHAR	Patient’s biological sex.	Male, Female, Unknown
birth_date	DATETIME	Patient’s date of birth.	Date format should be %Y-%m-%d
death_dttm	DATETIME	Patient’s death date, including time.	Datetime format should be %Y-%m-%d %H:%M:%S
language_name	VARCHAR	Patient’s preferred language.	Original string from the source data
language_category	VARCHAR	Maps language_name to a standardized list of spoken languages	Under-development

Hospitalization

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).

Variable Name	Data Type	Definition	Permissible Values
patient_id	VARCHAR	Unique identifier for each patient, linking to the patient table	No restriction
hospitalization_id	VARCHAR	Unique identifier for each hospitalization encounter. Each hospitalization_id represents a unique stay in the hospital	No restriction
hospitalization_joined_id	VARCHAR	Unique identifier for each continuous inpatient stay in a health system which may span different hospitals (Optional)	No restriction
admission_dttm	DATETIME	Date and time the patient is admitted to the hospital	Datetime format should be %Y-%m-%d %H:%M:%S
discharge_dttm	DATETIME	Date and time the patient is discharged from the hospital	Datetime format should be %Y-%m-%d %H:%M:%S
age_at_admission	INT	Age of the patient at the time of admission, in years	No restriction
admission_type_name	VARCHAR	Type of inpatient admission. Original string from the source data	e.g. “Direct admission”, “Transfer”, “Pre-op surgical”
admission_type_category	VARCHAR	Admission disposition mapped to mCIDE categories	Under-development
discharge_name	VARCHAR	Original discharge disposition name string recorded in the raw data	No restriction, e.g. “home”
discharge_category	VARCHAR	Maps discharge_name to a standardized list of discharge categories	Home, 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_digit	VARCHAR	Patient’s 9 digit zip code, used to link with other indices such as ADI and SVI	No restriction
zipcode_five_digit	VARCHAR	Patient’s 5 digit zip code, used to link with other indices such as ADI and SVI	No restriction
census_block_code	VARCHAR	15 digit FIPS code	No restriction
census_block_group_code	VARCHAR	12 digit FIPS code	No restriction
census_tract	VARCHAR	11 digit FIPS code	No restriction
state_code	VARCHAR	2 digit FIPS code	No restriction
county_code	VARCHAR	5 digit FIPS code	No restriction

Notes:

1. If a patient is discharged to Home/Hospice, then discharge_category == Hospice.

2. 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.

3. If a patient is transferred between different hospitals within a health system, a new hospitalization_id should be created

4. 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

5. A hospitalization_joined_id can also be created from a CLIF table from contiguous hospitalization_ids

ADT

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.

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	ID variable for each patient encounter	No restriction
hospital_id	VARCHAR	Assign a unique ID to each hospital within a healthsystem	No restriction
in_dttm	DATETIME	Start date and time at a particular location	Datetime format should be %Y-%m-%d %H:%M:%S
out_dttm	DATETIME	End date and time at a particular location	Datetime format should be %Y-%m-%d %H:%M:%S
location_name	VARCHAR	Location of the patient inside the hospital. This field is used to store the patient location from the source data. This field is not used for analysis.	No restriction
location_category	VARCHAR	Maps location_name to a standardized list of ADT location categories	ed, ward, stepdown, icu, procedural, l&d, hospice, psych, rehab, radiology, dialysis, other

Note: 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.

Vitals

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

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	ID variable for each patient encounter.	No restriction
recorded_dttm	DATETIME	Date and time when the vital is recorded.	Datetime format should be %Y-%m-%d %H:%M:%S
vital_name	VARCHAR	This field is used to store the description of the flowsheet measure from the source data. This field is not used for analysis.	No restriction
vital_category	VARCHAR	Maps vital_name to a list standard vital sign categories	temp_c, heart_rate, sbp, dbp, spo2, respiratory_rate, map, height_cm, weight_kg
vital_value	DOUBLE	Recorded value of the vital. Ensure that the measurement unit is aligned with the permissible units of measurements.	temp_c = Celsius, height_cm = Centimeters, weight_kg = Kg, map = mm/Hg, spo2 = %. No unit for heart_rate, sbp, dbp, and respiratory_rate
meas_site_name	VARCHAR	Site where the vital is recorded. No CDE corresponding to this variable (Optional field)	No restrictions. Note: no _category CDE variable exists yet

Labs

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

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	ID variable for each patient encounter.	No restriction
lab_order_dttm	DATETIME	Date and time when the lab is ordered.	Datetime format should be %Y-%m-%d %H:%M:%S
lab_collect_dttm	DATETIME	Date and time when the specimen is collected.	Datetime format should be %Y-%m-%d %H:%M:%S
lab_result_dttm	DATETIME	Date and time when the lab results are available.	Datetime format should be %Y-%m-%d %H:%M:%S
lab_order_name	VARCHAR	Procedure name for the lab, e.g. “Complete blood count w/ diff”	No restriction
lab_order_category	VARCHAR	Maps lab_order_nameto standardized list of common lab order names, e.g. “CBC”	CDE under development
lab_name	VARCHAR	Original lab component as recorded in the raw data, e.g. “AST (SGOT)”.	No restriction
lab_category	VARCHAR	Maps 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_value	VARCHAR	Recorded 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_numeric	DOUBLE	Parse out numeric part of the lab_value variable (optional).	Numeric
reference_unit	VARCHAR	Unit of measurement for that lab.	Permissible reference values for each lab_category listed here
lab_specimen_name	VARCHAR	Original fluid or tissue name the lab was collected from as given in the source data	No restriction
lab_specimen_category	VARCHAR	fluid or tissue the lab was collected from, analogous to the LOINC “system” component.	working CDE c(blood/plasma/serum, urine, csf, other).
lab_loinc_code	VARCHAR	LOINC code for the lab	No restrictions

Note: 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.

Patient Assessments

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.

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	Primary Identifier. Unique identifier linking assessments to a specific patient hospitalization.
recorded_dttm	DATETIME	The exact date and time when the assessment was recorded, ensuring temporal accuracy.	Datetime format should be %Y-%m-%d %H:%M:%S
assessment_name	VARCHAR	Assessment Tool Name. The primary name of the assessment tool used (e.g., GCS, NRS, SAT Screen).	No restriction
assessment_category	VARCHAR	Maps assessment_name to a standardized list of patient assessments	List of permissible assessment categories here
assessment_group	VARCHAR	Broader Assessment Group. This groups the assessments into categories such as “Sedation,” “Neurologic,” “Pain,” etc.	List of permissible assessment groups here
numerical_value	DOUBLE	Numerical Assessment Result. The numerical result or score from the assessment component.	Applicable for assessments with numerical outcomes (e.g., 0-10, 3-15).
categorical_value	VARCHAR	Categorical Assessment Result. The categorical outcome from the assessment component.	Applicable for assessments with categorical outcomes (e.g., Pass/Fail, Yes/No).
text_value	VARCHAR	Textual Assessment Result. The textual explanation or notes from the assessment component.	Applicable for assessments requiring textual data.

Provider

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

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	Unique identifier for each hospitalization, linking the provider to a specific encounter	No restriction
provider_id	VARCHAR	Unique identifier for each provider. This represents individual healthcare providers	No restriction
start_dttm	DATETIME	Date and time when the provider’s care or involvement in the patient’s case began	Datetime format should be %Y-%m-%d %H:%M:%S
stop_dttm	DATETIME	Date and time when the provider’s care or involvement in the patient’s case ended	Datetime format should be %Y-%m-%d %H:%M:%S
provider_role_name	VARCHAR	The original string describing the role or specialty of the provider during the hospitalization	No restriction
provider_role_category	VARCHAR	Maps provider_role_name to list of standardized provider roles	under development

Admission Diagnosis

Record of all diagnoses associated with the hospitalization. Expect breaking changes to this table as we seek to align it with existing diagnosis ontologies

Variable Name	Data Type	Definition	Permissible Values
patient_id	VARCHAR	Unique identifier for each patient	No restriction
diagnostic_code	DOUBLE	numeric diagnosis code	valid code in the diagnositic_code_format
diagnosis_code_format	VARCHAR	description of the diagnostic code format	icd9 ,icd10
start_dttm	DATETIME	date time the diagnosis was recorded	Datetime format should be %Y-%m-%d %H:%M:%S
end_dttm	DATETIME	date time the diagnosis was noted as resolved (if resolved)	Datetime format should be %Y-%m-%d %H:%M:%S

Medication Admin Intermittent

This table has exactly the same schema as medication_admin_continuous described below. The consortium decided to separate the medications that are administered intermittently from the continuously administered medications. However, the CDE for medication_category remains undefined for medication_admin_intermittent.

Medication Orders

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.

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	Unique identifier for each hospitalization, linking medication orders to a specific encounter	No restrictions
med_order_id	VARCHAR	Unique identifier for each medication order	No restrictions
order_start_dttm	DATETIME	Date and time when the medication order was initiated	Datetime format should be %Y-%m-%d %H:%M:%S
order_end_dttm	DATETIME	Date and time when the medication order ended or was discontinued	Datetime format should be %Y-%m-%d %H:%M:%S
ordered_dttm	DATETIME	Date and time when the medication was actually ordered	Datetime format should be %Y-%m-%d %H:%M:%S
med_name	VARCHAR	Name of the medication ordered	No restrictions
med_category	VARCHAR	Maps med_name to a list of permissible medication names	Combined CDE of medication_admin_continuous and medication_admin_intermittent , under development
med_group	VARCHAR	Limited number of medication groups identified by the CLIF consortium
med_order_status_name	VARCHAR	Status of the medication order, e.g. “held”, or “given”	No restrictions
med_order_status_category	VARCHAR	Maps med_order_status_name to a standardized list of medication order statuses	Under-development
med_route_name	VARCHAR	Route of administration for the medication	No restrictions, Examples include Oral, Intravenous
med_dose	DOUBLE	Dosage of the medication ordered	Numeric
med_dose_unit	VARCHAR	Unit of measurement for the medication dosage	Examples include mg, mL, units
med_frequency	VARCHAR	Frequency with which the medication is administered, as per the order	Examples include Once Daily, Every 6 hours
prn	BOOLEAN	Indicates whether the medication is to be given “as needed” (PRN)	0 (No), 1 (Yes)

Critical illness specific tables

Respiratory Support

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.

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	ID variable for each patient encounter
recorded_dttm	DATETIME	Date and time when the device started	Datetime format should be %Y-%m-%d %H:%M:%S
device_name	VARCHAR	Includes raw string of the devices. Not used for analysis	No restriction
device_category	VARCHAR	Maps device_name to a standardized list of respiratory support device categories	IMV, NIPPV, CPAP, High Flow NC, Face Mask, Trach Collar, Nasal Cannula, Room Air, Other
vent_brand_name	VARCHAR	Ventilator model name when device_category is IMV or NIPPV	No restriction
mode_name	VARCHAR	Includes raw string of the modes, e.g. “CMV volume control”	No restriction
mode_category	VARCHAR	Maps mode_name to a standardized list of modes of mechanical ventilation	Assist Control-Volume Control, Pressure Control, Pressure-Regulated Volume Control, SIMV, Pressure Support/CPAP, Volume Support, Other
tracheostomy	BOOLEAN	Indicates if tracheostomy is present	0 = No, 1 = Yes
fio2_set	DOUBLE	Fraction of inspired oxygen set in decimals (e.g. 0.21)	No restriction, see Expected _set values for each device_category and mode_category
lpm_set	DOUBLE	Liters per minute set	No restriction, see Expected _set values for each device_category and mode_category
tidal_volume_set	DOUBLE	Tidal volume set (in mL)	No restriction, see Expected _set values for each device_category and mode_category
resp_rate_set	DOUBLE	Respiratory rate set (in bpm)	No restriction, see Expected _set values for each device_category and mode_category
pressure_control_set	DOUBLE	Pressure control set (in cmH2O)	No restriction, see Expected _set values for each device_category and mode_category
pressure_support_set	DOUBLE	Pressure support set (in cmH2O)	No restriction, see Expected _set values for each device_category and mode_category
flow_rate_set	DOUBLE	Flow rate set	No restriction, see Expected _set values for each device_category and mode_category
peak_inspiratory_pressure_set	DOUBLE	Peak inspiratory pressure set (in cmH2O)	No restriction, see Expected _set values for each device_category and mode_category
inspiratory_time_set	DOUBLE	Inspiratory time set (in seconds)	No restriction, see Expected _set values for each device_category and mode_category
peep_set	DOUBLE	Positive-end-expiratory pressure set (in cmH2O)	No restriction, see Expected _set values for each device_category and mode_category
tidal_volume_obs	DOUBLE	Observed tidal volume (in mL)	No restriction
resp_rate_obs	DOUBLE	Observed respiratory rate (in bpm)	No restriction
plateau_pressure_obs	DOUBLE	Observed plateau pressure (in cmH2O)	No restriction
peak_inspiratory_pressure_obs	DOUBLE	Observed peak inspiratory pressure (in cmH2O)	No restriction
peep_obs	DOUBLE	Observed positive-end-expiratory pressure (in cmH2O)	No restriction
minute_vent_obs	DOUBLE	Observed minute ventilation (in liters)	No restriction
mean_airway_pressure_obs	DOUBLE	Observed mean airway pressure	No restriction

Expected *_set values for each device_category and mode_category

device_category == “IMV”

ventilator setting	Assist Control-Volume Control	Pressure Support/CPAP	Pressure Control	Pressure-Regulated Volume Control	SIMV	Volume Support
fio2_set	E	E	E	E	E	E
tidal_volume_set	E			E	P	E
resp_rate_set	E		E	E	E
pressure_control_set			E		P
pressure_support_set		E			E
flow_rate_set	P				P
inspiratory_time_set	P		E		P
peep_set	E	E	E	E	E	E

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 and lpm_set is required.

Medication Admin Continuous

The medication admin continuous table is a long-form (one medication administration per row) longitudinal table. 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.

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	ID variable for each patient encounter
med_order_id	VARCHAR	Medication order ID. Foreign key to link this table to other medication tables
admin_dttm	DATETIME	Date and time when the medicine was administered	Datetime format should be %Y-%m-%d %H:%M:%S
med_name	VARCHAR	Original med name string recorded in the raw data which often contains concentration e.g. “NOREPInephrine 8 mg/250 mL”
med_category	VARCHAR	Maps med_name to a limited set of active ingredients for important ICU medications, e.g. “norepinephrine”	List of continuous medication categories in CLIF
med_group	VARCHAR	Limited number of ICU medication groups identified by the CLIF consortium, e.g. “vasoactives”	List of continuous medication groups in CLIF
med_route_name	VARCHAR	Medicine delivery route	e.g. IV, enteral
med_route_category	VARCHAR	Maps med_route_name to a standardized list of medication delivery routes	Under-development
med_dose	DOUBLE	Quantity taken in dose
med_dose_unit	VARCHAR	Unit of dose
mar_action_name	VARCHAR	MAR (medication administration record) action, e.g. “stopped”
mar_action_category	VARCHAR	Maps mar_action_name to a standardized list of MAR actions	Under-development

Position

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.

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	ID variable for each patient encounter. This table only includes those encounters that have proning documented ever.
recorded_dttm	DATETIME	Date and time when the vital is recorded.	Datetime format should be %Y-%m-%d %H:%M:%S
position_name	VARCHAR	This field is used to store the description of the position from the source data. This field is not used for analysis.	No restriction
position_category	VARCHAR	Maps position_name to either prone or not prone.	prone, not_prone

Dialysis

The dialysis table is a wider longitudinal table that captures the start and stop times of dialysis sessions, the type of dialysis performed, and the amount of dialysate flow and ultrafiltration.

Variable Name	Data Type	Definition	Permissible values
hospitalization_id	VARCHAR	ID variable for each patient encounter
start_dttm	DATETIME	Start date and time of dialysis session	Datetime format %Y-%m-%d %H:%M:%S
stop_dttm	DATETIME	Stop date and time of dialysis session	Datetime format %Y-%m-%d %H:%M:%S
dialysis_type_name	VARCHAR	Name of dialysis type	No restriction
dialysis_type_category	VARCHAR	Maps dialysis_type_name to a list of standardized dialysis types	intermittent, peritoneal, crrt
crrt_mode_name	VARCHAR	Name of the CRRT mode, e.g. “CVVHD”	No restriction
crrt_mode_category	VARCHAR	Maps crrt_mode_name to a standardized list of CRRT modes	under development
fluid_removal_amt	DOUBLE	Amount of fluid removed during dialysis	Numeric
dialysate_flow_rate	DOUBLE	Rate of dialysate flow	Numeric

ECMO_MCS

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.

Variable Name	Description
hospitalization_id	Unique identifier for the hospitalization event.
start_dttm	Date and time when ECMO/MCS support started.
end_dttm	Date and time when ECMO/MCS support ended.
device_name	Name of the ECMO/MCS device used including brand information, e.g. “Centrimag”
device_category	Maps device_name to a standardized list of ECMO or MCS
device_metric_name	String that captures the measure of work rate of the device, e.g., RPMs.
device_rate	Numeric value of work rate, e.g., 3000 RPMs.
flow	Blood flow in L/min.
sweep	Gas flow rate in L/min.

Intake_Output

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.

Variable Name	Description
hospitalization_id	Unique identifier for the hospitalization event.
intake_dttm	Date and time of intake.
fluid_name	Name of the fluid administered.
amount	Amount of fluid administered (in mL).
in_out_flag	Indicator for intake or output (1 for intake, 0 for output).

Therapy_Details

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.

Variable Name	Description
hospitalization_id	Unique identifier for the hospitalization event.
session_start_dttm	Date and time when the therapy session started.
therapy_element_name	Name of the therapy element.
therapy_element_category	Category of the therapy element.
therapy_element_value	Value associated with the therapy element.

Microbiology Culture

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.

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	ID variable for each patient encounter.
order_dttm	DATETIME	Date and time when the test is ordered.	Datetime format should be %Y-%m-%d %H:%M:%S
collect_dttm	DATETIME	Date and time when the specimen is collected.	Datetime format should be %Y-%m-%d %H:%M:%S
result_dttm	DATETIME	Date and time when the results are available.	Datetime format should be %Y-%m-%d %H:%M:%S
fluid_name	VARCHAR	Cleaned fluid name string from the raw data. This field is not used for analysis.	No restriction. Check this file for examples
fluid_category	VARCHAR	Fluid categories defined according to the NIH common data elements.	CDE NIH Infection Site
component_name	VARCHAR	Original component names from the source data.	No restriction
component_category	VARCHAR	Maps component_name to a standardized list of component categories	culture, gram stain, smear
organism_name	VARCHAR	Cleaned organism name string from the raw data. This field is not used for analysis.	No restriction. Check this file for examples](https://github.com/clif-consortium/CLIF/blob/main/mCIDE/mCIDE_mapping_examples/clif_vocab_microbiology_name_organism_ucmc.csv)
organism_category	VARCHAR	Maps organism_name to the standardized list of organisms in the NIH CDE	CDE NIH Organism

Sensitivity

This table is used to store the susceptibility results of the organisms identified in the Microbiology Culture table and may be renamed to Microbiology_Susceptibility

Variable Name	Data Type	Definition	Permissible Values
culture_id	VARCHAR	Unique identifier linking to the culture from which the sensitivity test was performed
antibiotic	VARCHAR	Name of the antibiotic tested for sensitivity	Examples include Penicillin, Vancomycin
sensitivity	VARCHAR	The result of the sensitivity test, indicating the organism’s resistance or susceptibility	Resistant, Intermediate, Susceptible
mic	DOUBLE	Minimum Inhibitory Concentration (MIC) value, which measures the lowest concentration of an antibiotic needed to inhibit growth

Microbiology_Nonculture

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.

Variable Name	Description
hospitalization_id	Unique identifier for the hospitalization event.
result_dttm	Date and time when the non-culture result was obtained.
collect_dttm	Date and time when the sample was collected.
order_dttm	Date and time when the test was ordered.
fluid_name	Name of the fluid sample.
component_category	Category of the component tested.
result_unit_category	Unit category of the test result.
result_category	Category of the test result.

Procedures

A longitudinal record of each bedside ICU procedure performed on the patient (e.g. central line placement, chest tube placement). Note that this table is not intended to capture the full set of procedures performed on inpatients.

Variable Name	Data Type	Definition	Permissible Values
hospitalization_id	VARCHAR	Unique identifier for each hospitalization, linking the procedure to a specific encounter
procedure_name	VARCHAR	Name of the procedure performed on the patient	Examples include “Central Line Placement”
procedure_category	VARCHAR	Maps procedure_name to a list of standardized procedures	CDE under development
diagnosis	VARCHAR	The diagnosis or reason for performing the procedure
start_dttm	DATETIME	Date and time when the procedure was initiated

Future proposed tables

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

• Code_status: longitudinal record containing changes in the patient’s code status during the hospitalization over time
  hospitalization_id
start_dttm
code_status_name
code_status_category with levels c(DNR, UDNR, DNR/DNI, Full, Presume Full, Other)

• 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.