Assessing the Quality of Published Surveys in Ophthalmology

Abstract

Purpose : Surveys are an important research modality in ophthalmology, but their quality has not been rigorously assessed. This study evaluated the quality of published ophthalmic surveys.

Methods : Three survey methodologists, three senior ophthalmologists, and two research assistants developed a survey evaluation instrument focused on survey development and testing; sampling frame; response bias; results reporting; and ethics. Two investigators used the instrument to assess the quality of all ophthalmic surveys that were published between January 1, 2018 and December 31, 2018; indexed in MEDLINE/PubMed, Embase, and/or Web of Science; contained the search terms “ophthalmology” and “survey” or “questionnaire” in the title and/or abstract; and were available in English.

Results : The search identified 626 articles; 60 met the eligibility criteria and were assessed with the survey evaluation instrument. Most surveys (93%; 56/60) defined the study population; 48% (29/60) described how question items were chosen; 30% (18/60) provided the survey for review or described the questions in sufficient detail; 30% (18/60) were pre-tested or piloted; 25% (15/60) reported validity/clinical sensibility testing; 15% (9/60) described techniques used to assess non-response bias; and 63% (38/60) documented review by an institutional review board (IRB).

Conclusions : The quality of published ophthalmic surveys can be improved by focusing on survey development, pilot testing, non-response bias, and IRB review. The survey evaluation instrument can help guide researchers in conducting quality ophthalmic surveys and assist journal editors in evaluating surveys submitted for publication.

This is a 2020 ARVO Annual Meeting abstract.