How does the classification of area-level socioeconomic status in a cohort of out-of-hospital cardiac arrest patients vary depending on the size of the geographical area?
摘要
Lower socio-economic status (SES) is associated with poorer survival following out-of-hospital cardiac arrest (OHCA), yet methods for assigning SES vary across studies. Postal areas (POA) are frequently used as proxies for SES classification instead of smaller geographical units and may be derived from either the patient’s residential or arrest address. This study examined the agreement between SES classifications based on different sized geographical areas. All OHCAs attended by St John Western Australia between 2015 and 2024 with a valid residential address were included. Each case was assigned two geographical area codes: smallest census area (SA1; ~400 residents) and POA (~ 7,000 residents). SES, defined using the Index of Relative Socio-Economic Disadvantage (IRSD), was assigned at both levels according to the census year corresponding to each arrest. Concordance between SES quintiles derived from SA1- and POA-based IRSD classifications was assessed using percent agreement and weighted kappa. For arrests occurring outside the patient’s residence, SES quintiles based on residential and arrest addresses were compared separately at the SA1 and POA levels. Of 25,874 eligible OHCAs, agreement between POA-and SA1-derived SES quintiles was 38% (n = 9880) using residential address data. Among OHCAs occurring outside the residence (n = 3,41), SES classifications had agreement between residential and arrest addresses of 47% at the SA1 (n = 1470) and 59% at the POA level (n = 1,63). SES attribution is influenced by both the area size and address source used, highlighting the importance of methodological transparency in health outcomes research. This study demonstrates POA-derived SES classification, results in substantial discordance compared with SA1-derived classification.