COVID-19 online surveys need to follow standards and guidelines: Comment on “Does COVID-19 pandemic affect sexual behaviour? A cross-sectional, cross-national online survey” and “Binge watching behavior during COVID 19 pandemic: A cross-sectional, cross-national online survey”

Dear editor,

We recently read two articles (Arafat et al., 2020; Dixit et al., 2020) published in the Psychiatry Research, with great interest. Both being multinational studies and online surveys caught our attention the most. These two articles provide insightful findings regarding sexual and binge-watching behaviors during the COVID-19 pandemic. On a background of the plethora of changes that the pandemic restrictions have brought about in personal and social behaviors, the findings of these articles are really significant.

In one of the articles (Arafat et al., 2020) authors do acknowledge certain limitations, common to both the surveys- use of the non-native language i.e. English alone in the survey questionnaire circulated across Southeast Asian countries, convenience sampling and small sample size. However, many other methodological limitations need to be discussed. Both these online surveys, deviate significantly from the recommended standards for conducting and reporting web-based surveys, i.e. the Checklist for Reporting Results of Internet E-surveys (CHERRIES) (Eysenbach, 2004). We deem that the items that this checklist proposes are essential in enhancing the quality of the online surveys and they cover-up for the inherent limitations i.e. convenience sampling and responder selection bias. These inherent limitations become more prominent for ‘open’ surveys, which both the surveys are in this discussion.

Various items of CHERRIES and point-by-point check, if reported or not, for both the surveys against each of them are shown in supplementary table 1. Here, we comment upon certain key lacunae:

• 1Both these articles do not mention the Institutional Review Board (IRB) or the Institutional Ethics Committee (IEC) clearance and data protection. There is no mention on the confidentiality of data. The Authors needed to mention whether responders needed to share their contact address, such as email ID, in the survey or not. This is very crucial, because personal and sensitive information is being gathered. Although these are difficult times for IRB/IEC to have regular meetings and approve proposals, we believe IRB/IEC approvals and data protection strategies are a must for studies on sensitive topics like sex. Considering the need to not discouraging research because of the delay in IEC approvals, ethics regulatory bodies like the Indian Council for Medical Research (ICMR) has made amendments for fast tracking of reviews by relaxing the terms required for meetings (Indian Council of Medical Research, 2020a).
• 2Both these articles do not describe the development and validation process of the questionnaire used for the survey. It appears that both the surveys have used customized questionnaires, as no references for already validated questionnaires are cited against them. In such cases the whole process of development and validation (at least face and content validity) have to be described. Not only it improves the validity of the survey, but also allows other researchers to use the same questionnaire in other samples.
• 3The articles do not describe the recruitment process and the survey administration in detail. Although we presume that the surveys were ‘open’, the authors do not declare whether the survey is ‘open’ or ‘closed’. The details of the messenger platforms used (email, WhatsApp, etc.) are not mentioned. The geographical distribution of the survey circulation also needs to be mentioned, primarily because only one investigator is associated with one country. For instance, one investigator from Lucknow, a city in India, could possibly send the link only to his contacts and hence cannot generalize the findings as ‘national’. Same being true for investigators from other nations. And also, because the results obtained were not powered sufficiently, terming the surveys as ‘cross-national’ is not suitable.
• 4Both the articles report the use of the “Google form” platform for surveys. While this platform is user friendly, it has a significant limitation of inability to track IP addresses. We assume that the authors did not collect information on the responders contact details such as email ID or used mandatory registration for the sake of confidentiality. If so, the lack of IP address tracking, makes the surveys completely deficient in handling duplicate/multiple entries and therefore questioning the validity of the results.
• 5The surveys do not report response rates i.e. the view, participation and response rates. These are very essential to measure the degree of responder bias, which as mentioned is inherent to all ‘open’ online surveys. While one study (Arafat et al., 2020) does mention about ‘data cleaning’, they did not clarify what exactly they meant. Do they mean dropping incomplete responders?

Another critical issue that we would like to raise here is regarding approvals required for conducting studies in international collaboration. This is especially very relevant to India. We raise this issue because the affiliation of the Indian investigator in both the studies, is a public funded, government medical college. The ICMR, under the direction of the Ministry of Health and Family Welfare, Government of India has established the Health Ministry Screening Committee (HMSC), which reviews and approves all the proposals from government medical organizations with India-international collaboration; this remains valid even for COVID-19 related research (ICMR, 2020b). These studies do not mention such an approval. The investigators and the IRB/IEC that reviews such proposals, needs to be mindful of such regulations.

Nowadays web-based research in the form of online surveys are fast becoming a vogue, especially after the restrictions that the COVID-19 outbreak has forced nations to enforce. We strongly recommend adherence to standard guidelines that will help improve the quality of research and perhaps help maintain its sanctity.