Malaria kills an estimated 205,000 people per year in India, not the 15,000 estimated annually by the World Health Organization, researchers have learned.
Using data from a large cohort study of rural deaths in India, Dr. Neeraj Dhingra and Dr. Prabhat Jha of St. Michael's University, Toronto, and the University of Toronto, and their colleagues, determined that 3.6% of unattended febrile deaths of people between 1 month and 70 years of age were attributable to malaria.
Photo credit: CDC
Modeling using known population statistics, an estimated 55,000 early-childhood, 30,000 childhood and 120,000 adult deaths occur each year from malaria in India, the researchers wrote, while acknowledging lower and upper limits of 125,000 and 277,000 malaria deaths annually.
The findings, published online ahead of print Oct. 21 in the Lancet, suggest that the WHO, which relies heavily on India's hospital-based epidemiologic surveillance, has woefully underestimated India's true malaria burden.
"Because the Indian national malaria program cures nearly all the cases it treats, it detects only about 1,000 malaria deaths each year," Dr. Dhingra and Dr. Jha wrote, adding that the WHO estimates, while taking into consideration the likelihood of some undiagnosed cases, nonetheless "[depend] indirectly on the low death rates in diagnosed patients." The malaria death rates did correspond, however, with the Indian national program's reported malaria transmission trends by geographic region.
For their research, Dr. Dhingra and Dr. Jha examined results from verbal autopsies – interviews with household members of the deceased – with data recorded using standardized questionnaire forms. The verbal autopsies were conducted between 2001 and 2003, by trained nonmedical field workers, in 6,671 randomly selected areas of India. Of the 122,291 autopsies conducted, 75,342 were of people between 1 month and 70 years of age.
Of the 2,681 deaths attributable to malaria, 90% were in rural areas and 86% were not in a hospital or clinic, Dr. Dhingra and Dr. Jha noted: "Most deaths in rural India take place at home, without prior intervention by any qualified health care worker."
In an accompanying editorial in the Lancet, Robert W. Snow, Ph.D., of the KEMRI–University of Oxford–Wellcome Trust Research Programme in Nairobi, said the finding that 86% of India’s malaria deaths did not occur in hospitals or clinics suggests that "the health-management information system in India is not fit for purpose for the recording of malaria morbidity and mortality." This, Dr. Snow said, "is particularly surprising for a country that boasts a space program and is an emerging global economic leader."
The verbal autopsy results were used to determine the onset and severity of the fever leading to death, among other clinical characteristics of malaria such as shivering, jaundice, vomiting, breathlessness, decreased urine output, headache, convulsions, or unconsciousness. Blood tests for malaria were rarely reported. Two physicians analyzed each autopsy record, assigning a code for cause of death. Malaria deaths were catalogued separately from other febrile deaths such as those caused by dengue or typhoid.
"The major source of uncertainty in our estimates arises from the possible misclassification of malaria deaths as deaths from other diseases," Dr. Dhingra and Dr. Jha wrote, saying that their lower and upper estimates – of 125,000 and 277,000 annual deaths – were calculated by including only those deaths immediately coded by two physicians as malaria and, for the high end, all deaths with malaria as the initial diagnosis by one coder, a quarter of which were later attributed to other causes.
In his editorial, Dr. Snow praised the researchers' methodology and conclusions. "First, there was a strong geographical correlation with state-reported malaria mortality statistics; second, the malaria mortality data showed credible temporal trends with peaks after the wet season in every district; third, there was striking correspondence with malaria transmission rates calculated independently at the district level; and fourth, this spatial correlation was not seen in three other diseases whose symptoms are often confused with malaria (dengue, typhoid, and meningitis)," Dr. Snow wrote.
Similar disparities in the WHO malaria statistics and disease burden, Dr. Snow wrote, "could exist in other heavily populated, remote regions that are exposed to malaria and have unreliable access to health care, such as Burma, Bangladesh, Pakistan, Afghanistan, and Indonesia."
All those countries, except Pakistan and Afghanistan, occur in the WHO's South-East Asia region, which also includes India. In its 2009 global report on malaria, the World Health Organization said that the region, "received the least money per person at risk for malaria and saw the lowest increase in external financing between 2000 and 2007," adding that, in general, "High levels of external assistance are associated with increased procurement of commodities and decreases in malaria incidence."