Table 1 Entities description. *Processing of sensitive data as ethnicity is subject to legal restrictions under general data protection regulation (GDPR), and its systematic collection or use for secondary research purposes is often not permitted, which restricts its structured availability. The country of birth is then often used as a substitute.

From: Improving social determinants of health documentation in French electronic health records using large language models

Entity type

Definition

Examples

Examples translated into English

Living_Alone

The patient is described as living alone

Vit seul

Living alone

Living_WithOthers

The patient is described as living with other people. This doesn’t include animals but includes nursing home

Vit avec sa soeur; Vit avec ses parents; Vit avec son épouse et ses enfants ; Vit en EHPAD

Lives with his/her sister; Lives with his/her parents; Lives with his wife and children; Lives in nursing home

MaritalStatus_Single

The patient described as being single (i.e. no regular partner)

Célibataire sans enfant; Pas de conjoint;

Single without children; No partner

MaritalStatus_InRelationship

The patient is described as married, registered partnership or with a regular partner

Mariée; Vit en concubinage; Vit avec sa compagne; Pacsé ;

Married ; Lives with partner; Lives with his girlfriend; Civil union

MaritalStatus_Divorced

The patient is described as divorced or separated from is regular partner (except from non-regular partnership)

Divorcé; Plus de contact avec le père de ses enfants ;

Divorced; No more contact with the father of her children

MaritalStatus_Widowed

The patient is described as widowed

Veuf; Veuve;

Widower, Widow

Descendants_Yes

The patient has descendants

Deux filles; 1 fils adopté; Trois petits-enfants; Courses faites par sa fille

Two daughters; 1 adopted son; Three grandchildren; Shopping done by daughter

Descendants_No

The patient has no descendants

Pas d’enfant; Sans enfant; Enceinte de son premier enfant ;

No children; Childless; Pregnant with first child ;

Job

All the patient’s jobs, previous and actual

Courtier dans le textile ; Travaille dans le batiment; Soudeur

Textile broker; Construction worker; Welder

Last_job

The patient’s latest job

Employment_Working

Any documentation that the patient is currently working. It includes short/finite period of time off from work as sick leave, sabbatical, maternity leave or only information about job

Travaille toujours; En reconversion professionnelle ; Agent de ménage ; Maître d’œuvre ;

Still working; Retraining; Housekeeper; Prime contractor ;

Employment_Unemployed

The patient is described as unemployed

Vient de terminer son CDD ; Il est au chômage

Has just finished his fixed-term contract; He is unemployed

Employment_Student

The patient is described as a student

Actuellement lycéen; … en première année de BTS ; Patient étudiant

Currently in high school; … in first year of associated degree ; Student patient

Employment_Pensioner

The patient is described as pensioner. It also includes mentions such as “former” + job references, without explicit mention of pensioner

Retraité; A la retraite; Ancien magasinier

Retired; Retired; Former warehouseman

Employment_Other

The patient is not described with previous employment labels nor job. This entity describes other employment situations: unclear work status without temporality, disability, long sick leave, irregular situation, housewife, volunteer work, etc

Il a été courtier dans le textile ; Elle a travaillé comme dentiste ; Travailleur handicapé et ne travaille plus depuis 2011 ; En invalidité ; Mère au foyer ;

He was a textile broker; She worked as a dentist; Disabled worker and hasn’t worked since 2011; On disability; Housewife ;

Entity type

Definition

Examples

Examples translated into English

Alcohol

The alcohol use is annotated as an event. Only the trigger anchor is annotated here. The status of alcohol use (current, past, none) is annotated through a relationship with the StatusTime entity

Alcool; Ethylisme; CAD; OH; Intoxication ethylique

Alcohol; Ethylism; CAD; OH; Ethyl intoxication

Tobacco

The tobacco use is annotated as an event. Only the trigger anchor is annotated here. The status of tobacco use (current, past, none) is annotated through a relationship with the StatusTime entity

Tabac; Tabagisme ; Consommation tabagique

Tobacco; Smoking; Tobacco consumption

Drug

The drug use is annotated as an event. Only the trigger anchor is annotated here. The status of drug use (current, past, none) is annotated through a relationship with the StatusTime entity

Héroïne en fumette; Cannabis; drogue

Heroin smoking; Cannabis; Drugs

Housing_Yes

The patient is described as having a housing stability

Habite à Nantes; Pas d’aides à domicile ; Vit dans une maison

Lives in Nantes; No home help; Lives in a house

Housing_No

The patient is described as having a housing instability

Sans domicile fixe;

Hébergé chez des tiers

Homeless;

Living with a third party

PhysicalActivity_Yes

The patient is described as exercising regularly

Fait du vélo; Marche 30 min par jour

Cycling; Walking 30 min a day

PhysicalActivity_No

The patient is described as not exercising regularly

Marche avec une canne, sort peu ; Pas d’activité sportive ;

Sédentaire ; Performance status 3

Walks with a cane, goes out little; No sports activity;

Sedentary; Performance status 3

Income

Any documentation about patient’s financial resources. It includes social aids

RSA; AAH;

Active solidarity income; Disabled adult allowance

Education

Any documentation about patient’s education

Certificat d’étude; BTS Informatique; Licence en droit

Certificate of Study; Associate degree in Computer Science; Bachelor’s Degree in Law

Ethnicity*

The patient’s country of birth

Originaire du Maroc; Originaire d’Érythrée

From Morocco ; From Eritrea

StatusTime

The status of substance consumption (alcohol, tobacco and drug). An attribute is associated to this entity with possibles values: current, past, none

Ne consomme pas d’alcool

No alcohol

History

Any date of occurrence of an event

Il y a 8 ans; Il a arrêté depuis 3 ans

8 years ago; He stopped 3 years ago

Duration

Any exposure duration (to substance use, house instability, etc.)

Il a fumé pendant 20 ans.

He smoked for 20 years.

Amount

Any amount related to SDoH, such as Descendants or substance use

20 paquets de cigarettes; 3 verres de vin; 2 enfants;

20 packs of cigarettes; 3 glasses of wine; 2 children;

Frequency

Any frequency related to SDoH, such as physical activity or substance use

Par jour; par mois; occasionnellement

Daily; monthly; occasionally

Type

Specify certain broad entities to get fine-grained details, such as the type of descendants, the type of drugs or alcohol used

Enfants; petits-enfants; vin; bière ; cigarette ; cocaine ; cannabis

Children; grandchildren; wine; beer; cigarettes; cocaine; cannabis