Evidence for health policy in India: do we have enough data?

Recently, two reports were published in leading journals. The first report in Lancet (Vol.379: May12, 2012) by Rajesh Dikshit et al., infers that the most common fatal cancer in women aged 30–69 years is cervical with burden of 17.1% ¹ . The latest paper in the Journal of the Royal Society of Medicine states the highest age-adjusted mortality rate of 7.7 per 100,000 as being for cervical cancer ² . Earlier evidence suggests it is around 65.5 in a rural area ³ . Thus the range of estimates for the disease burden varies from a low of 7.7 to a high of 65.5. For policy makers, this poses as a significant problem, as to which estimate to trust?

All the population-based registries from India and other data sources, as in the recent article, refer to data mostly from cancers reported from registries. Those depend mainly data from urban conurbations of the country ² . The data from registries in India cover less than five percent of the total population of the country. Conclusions drawn from these registries cannot be viewed as representative of the total population, given that rural areas are mostly missed out, and cervical cancer rates might be higher in rural areas. The new cases of cancer detected by registries underrepresent the total number of cases, and may overrepresent the less severe cases or cases from upper socioeconomic strata who are able to afford healthcare ⁴ .

At best, none of these studies can provide causal interpretations. Methodologically, none of these papers have sufficient information to tackle one putative question: whether cancer of the cervix is highly prevalent or not in India. Hence, authors of the paper did not have reliable data to either support or reject the idea that HPV vaccine should be put to trial ⁵ .