Measuring recall of medical information in non-English-speaking people with cancer : a methodology

Background: Many patients who require an interpreter have difficulty remembering information from their medical consultations. Memory aids such as consultation audio-recordings may be of benefit to these patients. However, there is no established means of measuring patients' memory of medical information. Objectives: This study aimed to develop a method for eliciting and coding recall of medical information in non-English-speaking patients. Design: This method, called Patient-Interpreter-Clinician coding (PICcode), was developed in the context of a phase II trial conducted in two outpatient oncology clinics in Melbourne, Australia, and was refined iteratively through consultation with an expert panel and piloting. Between-coder differences in early versions of the coding system were resolved through discussion and consensus resulting in refinements to PICcode. Results: The final version of PICcode involved transcribing, translating and coding of audio-recorded consultations and semi-structured interviews (SSI). The SSIs were designed to elicit patients' free-recall of medical information. Every unit of medical information in the consultation was identified and categorized in a coding tree. SSIs were coded to identify the extent to which information was recalled from the consultation. Discussion: The iterative changes involved in developing PICcode assisted in clarifying precise details of the process and produced a widely applicable coding system. PICcode is the most comprehensively described method of determining the amount of information that patients who use an interpreter recall from their medical consultations. PICcode can be adapted for English-speaking patients and other healthcare populations.