I thank Basu & Arya for their kind words about our paper and for their reaffirmation of the importance of addressing adherence to treatment. However, although they note that, ‘In countries having lesser mental healthcare resources, such coordinated provision of treatment is lacking’, our results (online Table DS2) show that coordinated treatment is typically lacking even in higher-income countries. Indeed, the median number of visits in the past 12 months among patients receiving treatment for mental disorders in general medical services is no different in high-income (1.5) than in low-/lower-middle-income (1.4) countries and only slightly higher in upper-middle-income countries (2.1). We also found that the proportion of patients prematurely terminating primary care treatment of mental disorders is quite high in high-income countries (35.4%) as well as in lower-income countries (52.5% for both groups).

Although Basu & Arya consider the World Mental Health question on stopping treatment irrelevant to relationships with spiritual or religious healers, great care was taken in crafting the question sequence in which this question was embedded to be broadly applicable across treatment sectors and countries. The sequence began by asking respondents whether they ever in their life saw any of the professionals on a long country-specific customised list, for problems with their emotions, nerves, or use of alcohol or drugs. Respondents who reported having done so were asked whether they saw each type of professional for such problems in the past 12 months and, if so, number of visits, perceived helpfulness and whether or not they were still seeing the professional for these problems. Only those who said they had stopped seeing the professional were then asked, ‘Did you complete the full recommended course of treatment? Or did you quit before the [provider] wanted you to stop?’ I agree with Basu & Arya that the framing of this question and of the response options may not have been the most natural way to describe an on-going relationship with a spiritual or religious healer, and I agree that customisation might well yield important new information. However, we would expect reports of having ‘stopped’ to be lower-bound estimates of the extent to which care for on-going emotional problems lacked continuity, so the high proportions of patients in lower-income countries who gave such reports are cause for deep concern. Basu & Arya also note correctly that data on reasons for terminating treatment, including stigma, were not reported in the paper. Such data exist in the World Mental Health Surveys and will be presented in future reports.