Malaria Data

            The malaria data were compiled from two publications by the National Malaria Eradication Programme (NMEP) part of the Ministry of Health and Family Welfare within the Directorate General of Health Services of the Indian government (NMEP, 1986; NMEP, 1996).  The data contain different measures of malaria that are described below, yearly by state for 1961-1995, and by district for 1975-1984 and 1986-1995.  The data for 1985 were not available at the time the NMEP (1996) book that contains all of the data was published.  Due to the short time period for this study, I was not able to get the missing 1985 data.

            The malaria data are tabulated into the following categories:  the total population under surveillance (Population), the number of Blood Slides Examined (BSE), the number of blood smears found positive for the malaria parasite (Positives), the number blood smears found positive for Plasmodium falciparum (Pf), the percentage of Positives that are caused by Plasmodium falciparum (%Pf), the Annual Blood Exam Rate (ABER), the Annual Parasite Index (API), the Slide Positivity Rate (SPR), and the Slide falciparum Rate (SfR).  The following are the equations used to calculate the derived rates:

ABER = # of Blood Slides collected / Population

API = # of Positives / Population

SPR = # of Positives / BSE

SfR = Pf / BSE

%Pf = Pf / Positives

            The data were collected from the General Health Services, including the Rural Health Services.  In areas where these health services were not considered adequate for collecting data, malaria surveys were performed (NMEP, 1996).  The data include both active and passive case detection as well as data from special surveys.  Passive case detection involves examining blood smears collected from people suspected of having malaria who arrive at dispensaries, clinics, and hospitals (Sharma et al., 1994).  Active case detection involves multi-purpose workers visiting each house every two weeks and questioning about occurrence of fevers.  The worker takes a blood smear from each person who has had a fever since the last visit.  The blood smears are then taken to the local public health center (PHC) for examination.  Active case detection has been continually decreasing.  The data were collected longhand and were then entered into the computer in Delhi from the various sources.  Therefore, the data are of mixed quality (Bradley, 1999, personal communication).  Malaria surveillance assumes that all malaria cases will show a fever.  By taking blood smears of all those with fevers, the malaria rate of the community can be assessed.  The ABER, therefore, should be equal to the fever rate in the community.  The data are not entirely accurate because some malaria patients have a history of fevers but did not happen to have one when the smears are taken and some people who actually do have malaria may be tested but don’t have detectable levels of malaria in their blood (NMEP, 1996).  The population estimates are rather accurate in India because of a relatively good infrastructure for collecting census information.  The formal census that is conducted every ten years is considered to be done well (Bradley, 2000, personal communication). 

            SPR indicates the yearly parasitic load in the population under surveillance.  P. falciparum and P. vivax percentages measure the parasite species distribution in the area.  The number of positives is a combination of those collected through active case detection (malaria surveys) and those collected through passive case detection (data from general and rural health services).  If ABER is low especially during the transmission season, then the number of positive blood slides will not be an accurate measure of the malaria levels in the community, which implies that the API will not give a correct indicator of malaria prevalence in the community.  However, if a lot of blood smears are collected from people who do not have fevers only in order to meet the ABER quota, then the SPR, rather than the API, will be affected and not give an accurate reading of the number of people infected with malaria.  The recommended level of ABER is 10%, and in order to get an accurate measure of the community, it is recommended that ABER not be below 6% (NMEP, 1996).  Therefore, neither API nor SPR is an exact measure of malaria prevalence, and therefore both should be taken into consideration in any study of malaria rates (Bradley, 2000, personal communication). 

           Although the public health infrastructure in India is fairly good, problems can occur with both the active and passive case detection, which affects the reliability of the malaria counts.  The multi-purpose worker (MPW), who collects blood smears, oversees eight health programs and taking blood smears is not the one given the most attention.  In addition, 40-50% of MPW positions are vacant at one time and when a worker is on leave, there is no arrangement to have another staff person take over.  There is no active case detection in urban areas, and malaria cases from private clinics and a few government hospitals are not added into the malaria counts that are published by the NMEP.  Also, drug distribution centers give out drugs to people with malaria without taking blood smears, so those malaria cases are not known or counted.  Malaria cases from defense, paramilitary forces, border road organizations, police organizations, autonomous institutions, tea gardens, coffee plantations, and private hospitals are also not added to NMEP data (Sharma, 1996c).  All of these factors indicate that malaria is underreported in India.

            I also found some discrepancies in the data.  All of the malaria data were the same for Dharwad and Mysore from 1975 to 1984.  It is unlikely that two districts would have the exact same values for 10 years.  However, I was not able to determine which district is correct for this data or if both are.  This should be considered for all references to these districts throughout this report.