Quantification of NSW Ambulance Record Linkages with Multiple External Datasets.

Area of Science:

• Health Informatics
• Public Health Data Linkage
• Emergency Medical Services Research

Background:

• Accurate linkage of ambulance data with other health records is crucial for comprehensive patient care analysis.
• Previous studies have highlighted challenges in establishing robust data linkages across different healthcare datasets.
• Understanding linkage rates and potential biases is essential for the validity of research using linked health data.

Purpose of the Study:

• To quantify linkage rates between New South Wales (NSW) Ambulance records and four external health datasets.
• To identify and report systematic issues affecting data linkage accuracy for patient and episode-level data.
• To assess how patient and operational variables may introduce bias in subsequent studies using linked ambulance data.

Main Methods:

• Utilized data from NSW Ambulance for July 2006 to June 2009, linked with Emergency Department, Admitted Patient, NSW Registry of Births, Deaths and Marriages, and ABS mortality data.
• Employed the Centre for Health Record Linkage (CHeReL) in NSW for the linkage process.
• Analyzed linkage success based on 'expected' and 'not expected' links within 24 hours of paramedic attendance, and over a 3-year period for mortality.

Main Results:

• Achieved an overall episode-specific hospital linkage rate of 97.2% for 2,041,728 NSW Ambulance care episodes.
• 8.6% of episodes where patients were not transported to hospital resulted in an emergency department attendance within 24 hours.
• 5.2% of all care episodes linked to a death record within 3 years; 2.4% of death episodes occurred within 7 days of a paramedic encounter.
• Nonlinkage to hospital admission records decreased with increasing patient age; other linkage issues were less discriminatory.

Conclusions:

• The study provides a robust quantification of linkage rates and limitations for a large NSW Ambulance linked dataset.
• Identified age as a factor influencing linkage success, highlighting a potential source of bias.
• These findings are critical for the accurate interpretation of future research and the development of evidence-based clinical and operational policies for NSW Ambulance.