Background: EU-ADR project exploits data from eight longitudinal healthcare databases (DBs) from four countries (Italy, Netherlands, UK and Denmark) for drug safety signal detection, using an event-based approach. Due to differences in nature (i.e. general practice vs. claims DB), coding system, language and available information (i.e. discharge diagnoses, lab values, free text) across heterogeneous DBs participating in EU-ADR, harmonization of the queries for the identification of events deemed to be important in pharmacovigilance is necessary. Objectives: To describe the harmonization process for the data extraction in the EU-ADR network concerning four events: Acute Liver Injury (ALI), Neutropenia (NEUTROP), Pancytopenia (PANCYT) and Cardiac Valve Fibrosis (CARDFIB). Methods: Unified Medical Language System (UMLS) concepts corresponding to each event were selected by expert and projected into different coding systems (ICD9-CM, ICD10, READ code, ICPC). Based on these codes and available information coding algorithms were built and queries for event data extraction were run locally by each DB. Using custom-built software (Jerboa), age and gender standardized incidence rates (IRs) were calculated for each event and compared across different DBs. Based on benchmarking of IRs and revision of the queries, harmonized queries were eventually re-ran in DBs and IRs were re-calculated. Results: Codes were mainly searched in general practitioners’ and primary discharge diagnoses. Most DBs could also search lab values measurements for ALI, NEUTROP and PANCYT. Estimated age and gender standardized IRs varied across DBs as follows: 0.7–10/10,000PY (ALI), 0.5–2.5/10,000PY (NEUTROP), 0.3–2.1/100,000PY (PANCYTOP), 0.7–8.6/10,000PY (CARDFIB). Highest IRs for ALI, NEUTROP, and PANCYT were found in DBs with available data about lab values. Conclusions: Harmonization of event data extraction from heterogeneous DBs in a distributed network is possible and represents the base for drug safety monitoring in EU-ADR project. Observed differences in IRs for the events across DBs may depend on different underlying populations, coding system and available information.

Trifiro', G., Avillach, P., Gini, R., Coloma, P., Mazzaglia, G., Picelli, G., et al. (2011). Harmonization of the Event Identification Process in Eight European Healthcare Databases: The Experience from the EU-ADR Project. Intervento presentato a: International Conference on Pharmacoepidemiology & Therapeutic Risk Management, Chicago, Illinois, USA.

Harmonization of the Event Identification Process in Eight European Healthcare Databases: The Experience from the EU-ADR Project

Mazzaglia, G;Scotti, L;
2011

Abstract

Background: EU-ADR project exploits data from eight longitudinal healthcare databases (DBs) from four countries (Italy, Netherlands, UK and Denmark) for drug safety signal detection, using an event-based approach. Due to differences in nature (i.e. general practice vs. claims DB), coding system, language and available information (i.e. discharge diagnoses, lab values, free text) across heterogeneous DBs participating in EU-ADR, harmonization of the queries for the identification of events deemed to be important in pharmacovigilance is necessary. Objectives: To describe the harmonization process for the data extraction in the EU-ADR network concerning four events: Acute Liver Injury (ALI), Neutropenia (NEUTROP), Pancytopenia (PANCYT) and Cardiac Valve Fibrosis (CARDFIB). Methods: Unified Medical Language System (UMLS) concepts corresponding to each event were selected by expert and projected into different coding systems (ICD9-CM, ICD10, READ code, ICPC). Based on these codes and available information coding algorithms were built and queries for event data extraction were run locally by each DB. Using custom-built software (Jerboa), age and gender standardized incidence rates (IRs) were calculated for each event and compared across different DBs. Based on benchmarking of IRs and revision of the queries, harmonized queries were eventually re-ran in DBs and IRs were re-calculated. Results: Codes were mainly searched in general practitioners’ and primary discharge diagnoses. Most DBs could also search lab values measurements for ALI, NEUTROP and PANCYT. Estimated age and gender standardized IRs varied across DBs as follows: 0.7–10/10,000PY (ALI), 0.5–2.5/10,000PY (NEUTROP), 0.3–2.1/100,000PY (PANCYTOP), 0.7–8.6/10,000PY (CARDFIB). Highest IRs for ALI, NEUTROP, and PANCYT were found in DBs with available data about lab values. Conclusions: Harmonization of event data extraction from heterogeneous DBs in a distributed network is possible and represents the base for drug safety monitoring in EU-ADR project. Observed differences in IRs for the events across DBs may depend on different underlying populations, coding system and available information.
abstract
adverse drug-reactions; safety; pharmacovigilance; databases; validation; project; performance; experience; events; design
English
International Conference on Pharmacoepidemiology & Therapeutic Risk Management
2011
2011
20
suppl1
S4
S4
reserved
Trifiro', G., Avillach, P., Gini, R., Coloma, P., Mazzaglia, G., Picelli, G., et al. (2011). Harmonization of the Event Identification Process in Eight European Healthcare Databases: The Experience from the EU-ADR Project. Intervento presentato a: International Conference on Pharmacoepidemiology & Therapeutic Risk Management, Chicago, Illinois, USA.
File in questo prodotto:
File Dimensione Formato  
ICPE 2011.pdf

Solo gestori archivio

Dimensione 3.1 MB
Formato Adobe PDF
3.1 MB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10281/226766
Citazioni
  • Scopus ND
  • ???jsp.display-item.citation.isi??? 0
Social impact