On New York University Week: Receiving a diagnosis can be done from behind a computer screen, but is it as reliable as an in-person visit?
Daphna Harel, associate professor of applied statistics, explores this question.
Daphna Harel is an Associate Professor of Applied Statistics at the Steinhardt School of Culture, Education, and Human Development at New York University. She is an applied statistician who studies issues of measurement and modeling in the applied health sciences. Her research focuses on modeling challenges for data arising from self-reported questionnaires and other surveys. Her recent methodological work investigates shortening and scoring methods for patient reported outcome measures.
Depression Assessments Reliable Across Technology, Setting, and Type
With the rise of telehealth, many patients are able to receive diagnoses and treatments without an office visit to a doctor or therapist. This evolution in healthcare has expanded access for various populations who were previously unable to see doctors for reasons including disability, lack of transportation, and time. With telehealth accompanying an increase in online forms, we sought to compare whether electronic forms and assessments by phone are as reliable as those completed in person.
Our study, which was conducted in collaboration with the DEPRESSD project and McGill University, compares responses to several commonly used depression assessments that were obtained through different modes of administration. These assessments pose questions related to mood and other depression symptoms, including questions about feelings of hopelessness, as well as levels of sleep, appetite, and concentration.
We were particularly interested in whether there were differences in how people answered questions about their levels of depression when asked through different technologies, specifically pen-and-paper, electronic forms, or by phone. We tested responses in different settings, comparing answers given in a medical or research setting versus private settings like their homes. We also evaluated whether responses differed if the questions were read aloud to respondents in an interview, or whether the assessment was self-administered.
We found that scores obtained from different modes of administration were reliable, regardless of how the participants were reached. The results show that patients and clinicians can choose the mode of administration for a depression assessment tool that works best for them without worrying that choice is influencing the results.
This is important because everyone deserves to receive the best possible healthcare in ways that are convenient to them and their clinicians. Being able to assess depression symptoms in ways that fit each person allows for more individualized healthcare.
