Prevalence and clinical impact of the label ‘beta-lactam allergy’ in APHP hospitals: a cross-sectional study based on health data from a single day BETALL

Study type

• Study topic:
  Other
• Study topic, other:
  Beta-lactam drug hypersensitivity labelling
• Study type:
  Non-interventional study
• Scope of the study:
  Disease epidemiology
• Data collection methods:
  Secondary use of data

Non-interventional study

• Non-interventional study design:
  Cross-sectional

Population studied

• Short description of the study population:
  The study population includes all adult patients (≥18 years old) who had a consultation or hospitalization within AP-HP on the index date of December 14, 2022, within hospital departments using the ORBIS information system. This includes outpatient visits, day hospitalizations (HDJ), week hospitalizations (HDS), and conventional inpatient stays (MCO). This population corresponds to the full denominator used to estimate the prevalence of β-lactam allergy labelling within AP-HP. Among these patients, those carrying a β-lactam allergy label will form the subpopulation used for secondary analyses.; Inclusion criteria; • Patients aged 18 years or older.; • Patients who consulted or were hospitalized in AP-HP services using ORBIS on 14/12/2022 (consultations, HDJ, HDS, or conventional hospitalization).; Exclusion criteria :; • Patients managed in Emergency Departments (SAU) or Short-Stay Units (UHCD).; • Patients in intensive care or critical care units.; • Patients in post-acute/rehabilitation units (SSR).; • Patients hospitalized in psychiatry departments.; • Patients with major neurocognitive disorders.; • Patients under legal protection (safeguard of justice).; • Patients covered by CMU or AME.; Scope : All eligible patients from all AP-HP hospitals present in the EDS are included. Linked stays around the index date are considered to ensure complete extraction of relevant clinical documents.;
• Age groups:
  • Adult and elderly population (≥18 years)

    • Adults (18 to < 65 years)

      • Adults (18 to < 46 years)
      • Adults (46 to < 65 years)
    • Elderly (≥ 65 years)

      • Adults (65 to < 75 years)
      • Adults (75 to < 85 years)
      • Adults (85 years and over)
• Estimated number of subjects:
  33150

Study design details

• Study design:
  Retrospective cross-sectional study using AP-HP EDS data to estimate the prevalence of β-lactam allergy labelling on 14/12/2022 in adults receiving care. Inclusion of ORBIS-managed patients; exclusion of emergency, ICU, psychiatry, SSR, and protected adults.
• Main study objective:
  Estimate the prevalence of β-lactam allergy labelling among adults receiving care at AP-HP on 14/12/2022 using EDS data
• Setting:
  The study is conducted within the AP-HP hospital network, using data from the Entrepôt de Données de Santé (EDS AP-HP). The source population includes all patients aged 18 years or older who had at least one encounter (inpatient or outpatient) within any AP-HP hospital on 14 December 2022, regardless of medical specialty. Care settings include general medicine, surgery, emergency departments, and psychiatry.; From this source population, the study identifies patients with at least one mention of suspected or confirmed β-lactam allergy (“β-lactam allergy labelling”) extracted from structured and unstructured EHR components (diagnosis codes, medical notes, prescriptions, documented drug allergy lists).; Selection criteria follow the protocol specifications:; • Inclusion criteria:; - Adults ≥18 years old;; - At least one contact with an AP-HP facility on 14/12/2022;; - Availability of EHR data in the EDS;; - For analyses restricted to labelled patients: presence of a β-lactam allergy mention.; • Exclusion criteria:; - Patients who opted out of secondary use of their health data;; - Patients with invalid pseudonymisation keys;; - Records lacking essential demographic data.; The study period is fixed and cross-sectional: exposure and outcomes are assessed at a single index date (14 December 2022), with retrospective retrieval of historical EHR information (e.g., prior hospitalisations, prior prescriptions) when relevant for descriptive stratifications. No intervention, treatment arm, or comparator group is defined, as the study aims solely to characterise β-lactam allergy labelling prevalence and the epidemiological profile of labelled patients.; This setting ensures exhaustive coverage of all AP-HP patients on a predefined day, enabling a robust estimation of β-lactam allergy labelling prevalence at the hospital-network level.;
• Outcomes:
  Primary Outcome; The primary outcome is the prevalence of β-lactam allergy labelling among all adult patients (≥18 years) who had at least one encounter within an AP-HP care unit (outpatient visit, day hospital, emergency care, short-stay or conventional hospitalization) on the index date 14 December 2022.; A patient will be considered labelled as “β-lactam allergic” if at least one explicit mention of suspected or confirmed β-lactam allergy is identified in the electronic health record, using:; - structured data (Orbis allergy list, diagnosis codes),; - unstructured data processed by Natural Language Processing (NLP),; - PMSI (ICD-10 allergy-related codes or mentions).; The prevalence will be computed as:; Prevalence = (Number of patients labelled as β-lactam allergic)/(Total number of adult patients seen at AP-HP on 14/12/2022); This prevalence estimation includes all labelled patients, regardless of whether confirmatory allergy testing has been performed.;
• Data analysis plan:
  The primary analysis will consist of estimating the prevalence of β-lactam allergy labelling, expressed as a percentage with 95% bootstrap confidence intervals (resampling at the patient level).; The allergy status will be classified using all available EHR information:; 1. Confirmed allergy: at least one positive β-lactam allergy test documented in the medical record.; 2. Ruled-out allergy: at least one negative allergy test, indicating the label should no longer apply.; 3. Label without testing: mention of β-lactam allergy with no confirmatory test found.; For the primary prevalence analysis, all patients with any allergy label (with or without testing) will contribute to the numerator. Confirmatory test results will be used in secondary descriptive analyses to characterise mislabelling (e.g., proportion of ruled-out allergies still labelled in the EHR).; Stratified analyses will describe prevalence by sex, age group, hospital type, and care sector. Sensitivity analyses may exclude patients with contradictory or ambiguous information.; No modelling is required, as the study aims at descriptive epidemiology. Internal validity will rely on reproducible extraction rules, NLP quality checks, and robustness analyses across data sources.;