Anonymised

Overview

The Recruit Anonymised dataset provides fully anonymised clinical and operational data from healthcare organisations participating in recruit studies. All patient and user identifiers are hashed for privacy protection while maintaining referential integrity for data analysis. The dataset is designed for research studies requiring de-identified participant information across consultations, medications, observations, and healthcare operations.

Data is exposed through structured models covering patient demographics, clinical events, medications, organisational context, and administrative entities - all scoped to study participants and organisations.

---

Available Models

Core Clinical Models

Model	Description
Patient	Demographics, registration status, and consent flags
Observation	Clinical observations, test results, and coded entries
Consultation	Patient consultations with practitioners
Problem	Active and past clinical problems
Referral	Inbound and outbound referrals
Immunisation	Immunisation and vaccination records
Allergy	Patient allergies and adverse reactions
Diary Entry	Diary entries and recall events
Slot	Appointment slots and booking status

Medication Models

Model	Description
Drug Record	Medication courses (repeat and acute prescriptions)
Issue Record	Individual prescription issues from a drug record
Drug Record Problem Link	Links between drug records and clinical problems
Issue Record Problem Link	Links between issue records and clinical problems

Code & Reference Models

Model	Description
Clinical Code	SNOMED CT and Read code mappings
Codeable Concept	Extended code metadata including fully specified names
Drug Code	Drug code mappings and pharmaceutical identifiers
MKB Mapping Attributes	Language and attribute code mappings
MKB Mapping DMD Preparation	dm+d preparation code mappings
MKB Mapping Ethnicity	Ethnicity SNOMED code mappings
MKB Version	MKB coding database version and metadata

Organisational & Administrative Models

Model	Description
Organisation	Healthcare organisations participating in recruit studies
Organisation Location	Locations and branches within organisations
User In Role	Healthcare practitioners and their roles within organisations
Session	User sessions and system access records
Session User	User associations with sessions
Sharing Organisation	Organisations with data sharing agreements

---

Key Concepts

Anonymisation

All patient and user identifiers are pseudonymised using deterministic hashing:

• Patient identifiers: pseudo_patient_id, pseudo_nhs_no, pseudo_registration_guid
• User identifiers: Hashed user GUIDs in user_in_role and session models
• Organisational identifiers: EMIS organisation GUIDs (not pseudonymised)

Pseudonymisation is stable within your user context, enabling longitudinal analysis of patient episodes across consultations and medications.

Study Scoping

All clinical records are scoped to study participants via the study_id foreign key. Records are included in the dataset only if the patient is registered for the specified study.

Data Freshness

Data availability and update frequency depends on the source systems and study requirements. Consult your study protocol and data governance documentation for specific refresh schedules.

Organisational Context

Each clinical record includes a reference to the healthcare organisation where the event occurred (emis_registration_organisation_guid or similar). This enables analysis across multiple organisations and locations.

---

Common Patterns

Frequently Asked Questions

For practical guidance on common modelling questions with anonymised data, refer to your study-specific documentation or data dictionary.

Joining Clinical Events to Patient

Join clinical event models to patient using pseudo_registration_guid and study_id to access demographic information:

SELECT
  o.emis_observation_guid,
  o.observation_datetime,
  o.emis_code_id,
  p.gender,
  p.date_of_birth,
  p.ethnic_category_snomed_concept_id
FROM explorer_recruit_anon.observation AS o
INNER JOIN explorer_recruit_anon.patient AS p
  ON o.pseudo_registration_guid = p.pseudo_registration_guid
  AND o.study_id = p.study_id
WHERE o.study_id = 'STUDY001'

Resolving Clinical Codes

Use the clinical_code model to look up human-readable terms for clinical events:

SELECT
  o.emis_observation_guid,
  o.observation_datetime,
  c.term,
  c.snomed_concept_id
FROM explorer_recruit_anon.observation AS o
INNER JOIN explorer_recruit_anon.clinical_code AS c
  ON o.emis_code_id = c.emis_code_id
  AND o.study_id = c.study_id
WHERE o.study_id = 'STUDY001'
LIMIT 100

Filtering by Organisation

Scope queries to specific healthcare organisations using the organisation GUID:

SELECT
  p.pseudo_patient_id,
  p.date_of_birth,
  COUNT(DISTINCT o.emis_observation_guid) AS observation_count
FROM explorer_recruit_anon.patient AS p
LEFT JOIN explorer_recruit_anon.observation AS o
  ON p.pseudo_registration_guid = o.pseudo_registration_guid
  AND p.study_id = o.study_id
WHERE p.study_id = 'STUDY001'
  AND p.emis_registration_organisation_guid = 'ORG-GUID-001'
GROUP BY p.pseudo_patient_id, p.date_of_birth

Joining Medications to Problems

Link medication records to clinical problems through problem links:

SELECT
  dr.emis_drug_guid,
  dr.drug_name,
  p.problem_description,
  drpl.linkage_type
FROM explorer_recruit_anon.drug_record AS dr
INNER JOIN explorer_recruit_anon.drug_record_problem_link AS drpl
  ON dr.emis_drug_guid = drpl.emis_drug_guid
  AND dr.study_id = drpl.study_id
INNER JOIN explorer_recruit_anon.problem AS p
  ON drpl.emis_problem_guid = p.emis_problem_guid
  AND drpl.study_id = p.study_id
WHERE dr.study_id = 'STUDY001'
LIMIT 100