- Manish Kakkar, senior public health specialist, Public Health Foundation of India—Communicable Diseases ISID Campus, 4 Institutional Area Vasant Kunj, New Delhi, Delhi 110070, India
An estimated 20 000 people in India die each year from rabies,1 but in 2011 only 253 deaths were reported as having this cause.2 An estimated 100 000-200 000 people in India die annually from malaria,3 but in 2011 only 753 such deaths were reported.4
A recent spate of cases of dengue fever and a media outcry have brought the focus back to the widespread problem of under-reporting of cases of disease in India, linked to the ineffectiveness of our public health efforts.5
As of 26 November 37 070 cases of dengue fever had been reported this year in India.6 But a substantially bigger population is at risk, and India reported only an average of 4.2% of the total number of cases reported in the World Health Organization South East Asia region between 2000 and 2010.7
A study estimated that Thailand (population 70 million) had an annual incidence of more than 231 000 cases of symptomatic dengue in 2003-7.8 Given India’s population (1.2 billion) and environment, which is conducive to dengue, we should expect an incidence in India many times that of the Thai estimate. The 37 000 cases reported this year must be gross under-reporting.
The public health system in India records perhaps only about a third of disease events at best. And the proportion is likely to be smaller for dengue, firstly, because it has multiple manifestations (dengue virus has four serotypes); and, secondly, because dengue’s symptoms of rhinorrhoea and cough can make it difficult to distinguish from other febrile illness, especially in children.9 10
Dengue in India typically has an urban distribution. However, it is now also being reported in peri-urban and rural areas.11 Such spread will increase the true incidence. With health resources focused in urban areas, it is likely that its emergence in rural areas will be missed.
Disease surveillance is in its infancy in India. The National Vector Borne Disease Control Programme (NVBDCP) is the central government’s agency for the prevention and control of vector borne diseases. Its guidelines require that clinically suspected cases as well as laboratory confirmed cases are reported, including those from the network of sentinel laboratories—137 hospitals and 13 referral laboratories.12 This reporting is strong in states where dengue is a politically sensitive public health problem and weaker elsewhere.
The Integrated Disease Surveillance Project (IDSP) is a government initiative to detect and respond to disease outbreaks quickly, introduced in 2004 and reaching some states as late as 2007. It has a much larger network of more than 10 000 reporting sites throughout the country, which collect syndromic reports as well as laboratory confirmed cases. The NVBDCP publishes data on mainly laboratory confirmed cases. IDSP figures, although larger, vary widely according to reporter compliance and are also an underestimate.
Even in cases supposedly confirmed in laboratories there can be problems—for example, use of non-standard and non-validated tests and lack of adherence to testing algorithms. Workload, weak diagnostic facilities, and poor quality laboratory testing in primary and community healthcare centres and district hospital laboratories also contribute to under-reporting. Lack of accountability, poor quality testing, and slow laboratory turnaround times discourage clinicians from testing for dengue and encourage them to manage cases on the basis of clinical diagnosis alone.
Some 80% of patient care in India is sought in the private sector,13 which remains a largely untapped source of surveillance data. The official argument has been that this testing is low quality and difficult to monitor, but many private laboratories in city chains have excellent and accredited facilities. Under-reporting also has a political dimension. It allays public worry and reduces pressure on the administration because small numbers of cases indicate that control efforts are working.
Without a transparent disease reporting system we can have only limited understanding of the extent of the problem. Consequently, systems to prevent and respond to outbreaks are not sufficiently strong. We drift from one outbreak to the next. Only with severe outbreaks, as in 1996 and 2003, does dengue get attention, and then the absence of credible data adversely affects the quality of clinical management. Inadequate disease reporting and recording are likely to hamper decision making surrounding the introduction of specific interventions, such as vaccine, and subsequently monitoring.
In the Indian context, best practice in dengue surveillance, to improve the assessment of burden and hence the response to outbreaks, includes simplified case reporting; consistent application of standard case definitions; rapid turnaround of results and data sharing through electronic reporting; reporting of a minimum set of indicators; use of standardised laboratory protocols; and periodic studies to assess under-detection, under-reporting, and quality of surveillance.14
However, without a scientifically designed system of sentinel surveillance, and mechanisms to check the validity of burden estimates, whatever is reported in India will be inaccurate. We need more nationally representative, demographic surveillance sites that also act as disease sentinel sites to better estimate disease burden.
By not recording the true burden of infectious disease, we are ill prepared and allow major outbreaks to happen. We then spend our resources on crisis management, rather than prevention and control, and this leads to more deaths. Indian public health agencies must appraise the current system and make radical improvements.
Cite this as: BMJ 2012;345:e8574
Competing interests: The author has completed the Unified Competing Interest form and declares that no support from any organisation was taken for the submitted work; there are no financial relationships with any organisations that might have an interest in the submitted work in the previous three years, there are no other relationships or activities that could appear to have influenced the submitted work.
Commissioned; not externally peer reviewed.