Purpose: Clinical quality registries and other systems that conduct routine post-discharge surveillance of patient outcomes following surgery may have difficulty surveying patients who have limited proficiency in the language of the healthcare provider. Interpreter proxies (family and carers) are often used due to limited access to certified healthcare interpreters (due to cost or availability). The aim of this study was to assess the reliability of engaging interpreter proxies compared with

more »

... tified healthcare interpreters for the administration of patient-reported health-related surveys for people with limited English proficiency (LEP). Methods: People with LEP and due for a routine 6-month telephone follow-up post knee or hip arthroplasty were invited to participate. Participants were randomly allocated to having their first interview with an interpreter proxy or a certified healthcare interpreter followed by the second (crossover) interview within 2 weeks (range: 4 to 12 days) after the first interview using the alternative method. Agreement between the two methods was assessed using quadratic weighted Cohen's kappa, intraclass correlation and concordance correlation co-efficient where appropriate for EQ-5D health domains, total Oxford hip and knee scores, patient satisfaction, operation success, readmission, reoperation, and postsurgical complication responses. The mean of the differences between the same data items collected by each of the two methods was also calculated. Results: 85 participants (96%) completed the study. There was substantial to excellent inter-rater agreement (kappa=0.69-0.87 and ICCs above 0.74) for all but one measure. The mean differences between family proxy and healthcare interpreter scores for each participant were small, ranging from 0.01 (score range of 1-5) to 0.72 (score range of 0-100). Conclusion: Using interpreter proxies is a reliable alternative to certified healthcare 3 interpreters in conducting patient-reported health surveys, potentially making this process easier and cost effective for researchers and registries.