The Cohort Requester is a spark application server for querying FHIR data with a set of criteria and return a count or list of patients that match the criteria.
Fill in the configuration file src/main/resources/application.conf
with the appropriate values.
Or use the following minimal needed environment variables:
# The URL of the FHIR server
export FHIR_URL=http://localhost:8080/fhir
Build the project with maven:
mvn clean package
Run the application server:
java \
--add-opens=java.base/java.lang=ALL-UNNAMED \
--add-opens=java.base/java.lang.invoke=ALL-UNNAMED \
--add-opens=java.base/java.lang.reflect=ALL-UNNAMED \
--add-opens=java.base/java.io=ALL-UNNAMED \
--add-opens=java.base/java.net=ALL-UNNAMED \
--add-opens=java.base/java.nio=ALL-UNNAMED \
--add-opens=java.base/java.util=ALL-UNNAMED \
--add-opens=java.base/java.util.concurrent=ALL-UNNAMED \
--add-opens=java.base/java.util.concurrent.atomic=ALL-UNNAMED \
--add-opens=java.base/sun.nio.ch=ALL-UNNAMED \
--add-opens=java.base/sun.nio.cs=ALL-UNNAMED \
--add-opens=java.base/sun.security.action=ALL-UNNAMED \
--add-opens=java.base/sun.util.calendar=ALL-UNNAMED \
--add-opens=java.security.jgss/sun.security.krb5=ALL-UNNAMED \
-jar target/cohort-requester.jar
Launch a query with the following command:
curl -X POST http://localhost:8090/jobs -H "Content-Type: application/json" -d '{
"input": {
"cohortDefinitionSyntax": "{\"sourcePopulation\":{\"caresiteCohortList\":[118]},\"_type\":\"request\",\"request\":{\"_type\":\"andGroup\",\"_id\":0,\"isInclusive\":true,\"criteria\":[{\"_type\":\"basicResource\",\"_id\":1,\"isInclusive\":true,\"resourceType\":\"Patient\",\"filterFhir\":\"active=true&gender=female\",\"criteria\":[],\"dateRangeList\":[],\"temporalConstraints\":[]},{\"_type\":\"basicResource\",\"_id\":2,\"isInclusive\":true,\"resourceType\":\"Observation\",\"filterFhir\":\"code=http://snomed.info/sct|224299000\",\"criteria\":[],\"occurrence\":{\"n\":1,\"operator\":\">=\"},\"dateRangeList\":[],\"temporalConstraints\":[]}],\"dateRangeList\":[],\"temporalConstraints\":[]},\"temporalConstraints\":[]}",
"mode": "count"
}
}'
The job query format is as follows :
{
"input": {
"cohortDefinitionSyntax": "<cohort definition syntax>",
"mode": "<mode>",
"modeOptions": { // optional mode options
"details": "<details>" // list of criteria ids separated by commas or "all", this will activate a detailed count of the patients per criteria
},
"callbackUrl": "<callback url>" // optional callback url to retrieve the result
}
}
with mode
being one of the following values:
count
: Return the number of patients that match the criteria of thecohortDefinitionSyntax
create
: Create a cohort of patients that match the criteria of thecohortDefinitionSyntax
and cohortDefinitionSyntax
being a JSON string that represents the criteria described in the query format section.
If a callbackUrl
is provided (else the result will be printed in the logs), the response will be sent to the provided URL with the following format:
type RESPONSE = {
_type: string; // the job mode (count or create)
message: string; // the message of the response (usually the job status)
groupId: string; // the cohort id in case of a cohort creation job
count: number; // the count of patients in case of a count job
extra: Map<string, string>; // extra information about the count job (like details count per organization or per criteria)
stack: string; // the stack trace in case of an error
}
The query format of the cohortDefinitionSyntax
field is as follows (described in typescript
):
type REQUEST = {
sourcePopulation: {
caresiteCohortList?: Array<Integer> // Optional list of caresite ids that will be used to filter the resource with the fhir param `_list`
},
_type: "request",
request: CRITERIA,
resourceType?: string // The fhir resource type of the cohort (Patient being the default)
}
type CRITERIA = {
_type: "andGroup" | "orGroup" | "nAmongM" | "basicResource",
_id: Integer, // Unique identifier of the criteria
isInclusive: Boolean // If true, the criteria is inclusive, if false, the criteria is exclusive
}
type GROUP_CRITERIA = CRITERIA & {
_type: "andGroup" | "orGroup" | "nAmongM", // the type of the group "and" or "or" or "at least n matches among subcriterias"
criteria: Array<CRITERIA>, // list of subcriterias (can be basicResource or group)
temporalConstraints: Array<TEMPORAL_CONSTRAINT>, // list of temporal constraints between the subcriterias
nAmongMOptions: Array<OCCURENCE_CONSTRAINT> // list of occurrence constraints for the nAmongM type
}
type BASIC_RESOURCE = CRITERIA & {
_type: "basicResource",
resourceType: string, // The fhir resource type to filter
filterFhir: string, // The fhir query string to filter the resource
filterSolr?: string, // The solr query string to filter the resource (if using solr resolver)
occurrence?: OCCURENCE_CONSTRAINT, // The occurrence constraint of the resource
patientAge?: PATIENT_AGE, // The age constraint of the patient
dateRangeList?: Array<DATE_RANGE>, // The date range constraint of the resource
encounterDateRange?: DATE_RANGE // The date range constraint of the related encounter
}
type PATIENT_AGE = {
minAge?: string, // with the format {year}-{month}-{day}, eg. for someone of 21 years old this would be 21-0-0
maxAge?: string, // with the format {year}-{month}-{day}, eg. for someone of 21 years old this would be 21-0-0
datePreference?: string[], // resource reference field of the date to calculate the age
dateIsNotNull?: boolean
}
type DATE_RANGE = {
minDate?: string, // with the standard format {year}-{month}-{day}, eg. for 2024-01-25
maxDate?: string, // with the standard format {year}-{month}-{day}, eg. for 2024-01-25
datePreference?: string[], // resource reference field of the date to filter
dateIsNotNull?: boolean
}
type TEMPORAL_CONSTRAINT = {
idList: string | Array<number>,
constraintType: "sameEncounter" | "differentEncounter" | "directChronologicalOrdering" | "sameEpisodeOfCare",
occurrenceChoiceList?: Array<{_id: number, occurrenceChoice: string}>,
timeRelationMinDuration?: TEMPORAL_CONSTRAINT_DURATION,
timeRelationMaxDuration?: TEMPORAL_CONSTRAINT_DURATION,
datePreferenceList?: Array<{_id: number, occurrenceChoice: string}>,
filteredCriteriaIdList?: string | Array<number>,
dateIsNotNullList?: boolean | {_id: number, dateIsNotNull: boolean}
}
type TEMPORAL_CONSTRAINT_DURATION = {
years?: number,
months?: number,
days?: number,
hours?: number,
minutes?: number,
seconds?: number
}
type OCCURENCE_CONSTRAINT = {
n: number,
operator: string,
sameEncounter?: boolean,
sameDay?: boolean,
timeDelayMin?: boolean,
timeDelayMax?: boolean
}