Patterns and distribution of tobacco consumption in India
http://www.100md.com
《英国医生杂志》
EDITOR—Singh raises the issue of under-reporting due to social stigma and the lack of adult male members being interviewed directly in the second national family health survey dataset. However, as he correctly acknowledged, the second national family health survey is the best dataset available on the prevalence of consumption of paan masala, tobacco, smoking, and alcohol in India. Besides, these data come from the household questionnaire, in which one respondent was interviewed about their household.
In 40% of cases these respondents were men and 60% were women. Male respondents reported that 33% of men in their households smoked, whereas female respondents reported that 34% of men in their households smoked. Some reporting bias may exist on the tobacco consumption of men but this is likely to be modest. However, readers should be aware that under-reporting, based on sex and perhaps standard of living, may well be present in these data and we thank Singh for raising this issue. The focus of our paper was not on prevalence of tobacco consumption but on disparities in consumption, and it is possible that the extent of inequalities may be somewhat attenuated as a result of reporting bias.
Sehmi makes an interesting point about religion and tobacco consumption. Unfortunately, the dataset did not permit (analytically) a further breakdown of the religious grouping. However, we fully acknowledge the important role that religion may have in reducing tobacco consumption and influencing other health related behaviours.(Authors' reply)
In 40% of cases these respondents were men and 60% were women. Male respondents reported that 33% of men in their households smoked, whereas female respondents reported that 34% of men in their households smoked. Some reporting bias may exist on the tobacco consumption of men but this is likely to be modest. However, readers should be aware that under-reporting, based on sex and perhaps standard of living, may well be present in these data and we thank Singh for raising this issue. The focus of our paper was not on prevalence of tobacco consumption but on disparities in consumption, and it is possible that the extent of inequalities may be somewhat attenuated as a result of reporting bias.
Sehmi makes an interesting point about religion and tobacco consumption. Unfortunately, the dataset did not permit (analytically) a further breakdown of the religious grouping. However, we fully acknowledge the important role that religion may have in reducing tobacco consumption and influencing other health related behaviours.(Authors' reply)