Measurement of population mental health: evidence from a mobile phone survey in India. Issue 5 (9th March 2021)

• Record Type: Journal Article
• Authors: Coffey, Diane; Hathi, Payal; Khalid, Nazar; Thorat, Amit
• Is Part Of: Health policy and planning. Volume 36:Issue 5(2021)
• Page Start: 606
• Page End: 619
• Publication Date: 2021-03-09
• Subjects: Population mental health -- India -- mobile phone survey -- Kessler-6 -- SRQ; Medical policy -- Developing countries -- Periodicals; Public health -- Developing countries -- Periodicals; Health planning -- Developing countries -- Periodicals; 362.1091724
• DOI: 10.1093/heapol/czab023
• Languages: English
• ISSNs: 0268-1080

Abstract

Abstract: In high-income countries, population health surveys often measure mental health. This is less common in low- and middle-income countries (LMICs), including in India, where mental health is under-researched relative to its disease burden. The objective of this study is to assess the performance of two questionnaires for measuring population mental health in a mobile phone survey. We adapt the Kessler-6 screening questionnaire and the World Health Organization's Self-Reporting Questionnaire (SRQ) for a mobile phone survey in the Indian states of Bihar, Jharkhand and Maharashtra. The questionnaires differ in the symptoms they measure and in the number of response options offered. Questionnaires are randomly assigned to respondents. We consider a questionnaire to perform well if it identifies geographic and demographic disparities in mental health that are consistent with the literature and does not suffer from selective non-response. Both questionnaires measured less mental distress in Maharashtra than in Bihar and Jharkhand, which is consistent with Maharashtra's higher human development indicators. The adapted SRQ, but not the adapted Kessler-6, identified women as having worse mental health than men in all three states. Conclusions about population mental health based on the adapted Kessler-6 are likely to be influenced by low response rates (about 82% across the three samples). Respondents were different from non-respondents: non-respondents were less educated and more likely to be female. The SRQ's higher response rate (about 94% across the three states) may reflect the fact that it was developed for use in LMICs and that it focuses on physical, rather than emotional, symptoms, which may be less stigmatized.