Indmedica Home | About Indmedica | Medical Jobs | Advertise On Indmedica
Search Indmedica Web
Indmedica - India's premier medical portal

Indian Journal of Community Medicine

Profile of Dengue Fever Outbreaks in Maharashtra

Author(s): Prakash Doke*, Satish Pawar**

Vol. 25, No. 4 (2000-10 - 2000-12)

*Addl. Director, Directorate of Health Services, Mumbai **Additional District Health Officer, Zilla Parishad, Pune


Research questions: 1. What is the profile of dengue fever outbreaks in Maharashtra? 2. How the epidemics are investigated and controlled?

Objectives: 1. To study the epidemic profile of dengue fever outbreaks in Maharashtra state. 2. To study the correctness of outbreak investigations and control measures carried out by public health department of the State. 3. To develop experience based guidelines for field officers regarding dengue fever epidemic investigations and control measures.

Study period: Jan 1997 to Dec. 1997.

Study design: Record based retrospective study.

Setting: Public health department, Maharashtra state.

Sample size: All the outbreaks occurred in Maharashtra state during 1997.

Results and conclusions: Dengue fever outbreaks are reported late and time period between reporting of epidemic to health department and response by PHC is also long. Outbreaks are more common in middle sized, scarcity affected and piped water supplied villages. DF outbreaks can be predicted (81%) with the help of indicators like sudden rise in fever rate, presence of Aedes aegypti breeding and negative PBS in fever cases of affected village.

Keywords : Dengue fever, Outbreaks, PBS, Fevers, Health institutions


Fever is the commonest reason for hospital attendance in rural India. In recent past, Malaria was the main reason for fever. Even today, probably malaria may rank first. Outbreaks of fever are regularly reported from health workers and health institutions. Now-a-days after malaria, dengue fever has become important cause of fever outbreaks.

In Maharashtra state, dengue outbreaks are being reported since 1965. Previously dengue outbreaks were normally confined to urban areas like Nagpur, Barshi, Pune, Amalner, Miraj etc. Presently, sporadic cases are reported from urban areas and all the epidemics are reported from rural areas. The review of last few years data indicates that the frequency of dengue fever outbreaks is increasing.

National Institute of Virology (NIV) Pune has investigated number of dengue fever outbreaks from rural Maharashtra. These investigations mainly dealt with virological studies. Systematic outbreak reporting, investigations and control measures by state and district health authorities started recently.

Though number of studies are available; experience-based guidelines and information is not available for control of dengue fever for field officers. The information and control measures presently adopted are based on the book picture of the dengue fever and guidelines issued by Govt. of India. The importance of local guidelines became more evident when it was revealed that, quite a few epidemiological factors were different in outbreaks occurring in rural areas of Maharashtra state than the picture of dengue fever elsewhere. The concern became more serious after 1995 Delhi DHF outbreak. Thus, a systematic study of dengue fever outbreaks for accurate information was thought to be necessary. Further, study in local context would be more relevant in local application of the knowledge and utilization of this knowledge in developing strategies for controlling dengue in rural areas of the state by the field officers.

Material and Methods:

This study was undertaken by public health department in all districts of Maharashtra. The study relates to 31 districts of the state comprising 1,695 primary health centers and 895 million population.

All the fever outbreaks reported in the rural areas of the state during 1997 were included in the study. As the previous precise data on fever rate of each locality in the state was not known, the definition of the outbreak prepared by State Epidemic Control Bureau based on population, number of outbreaks reported and under-reporting, was utilized. By this definition, outbreak meant 5 or more cases of fever/1,000 population in one week.

All the fever outbreaks were investigated by district/state epidemiologist. The findings were recorded in pretested format of investigation report. After receiving information regarding outbreak, Medical officer of PHC visited the village of outbreak and examined the patients clinically and arrived at a tentative diagnosis. During the investigation of all the epidemics house to house survey was carried out, for determining the extent of epidemic. The peripheral blood smears were collected and examined for presence of malarial parasites. Search for breeding of Aedes aegypti was carried out with the help of trained personnel. Blood samples were collected from patients in acute phase, by venupuncture. The sera samples were immediately sent to National Institute of Virology, Pune in a vaccine carrier maintaining temperature around 4° C. The adult Aedes aegypti mosquitoes when collected were sent to National Institute of Virology, Pune for virus isolation. If IgM antibodies (indication of recent infection) were detected or dengue virus isolated from any of the patient's sera or virus isolated from mosquitoes collected from outbreak site, the outbreak was considered as dengue fever outbreak. Simultaneously, control measures which included symptomatic treatment, thermal fogging with pyrethrum/technical malathion, emptying and cleaning of all domestic and peridomestic water containers, indoor residual spraying etc. were initiated.


All the outbreaks observed during the study were from rural areas of Maharashtra state. A total of 246 fever outbreaks were reported in the year 1997, out of which 128 were due to Malaria. Ninety outbreaks were suspected as dengue outbreaks, of which 46 were confirmed by laboratory investigations.

These 46 confirmed dengue fever outbreaks were reported from 11 districts, affecting 7661 people and without any death. All the outbreaks were reported from rural areas. It was seen that, Marathwada region was most affected.

Fifty seven percent outbreaks were reported to medical officer of the primary health centre within 24 hours. In only 10% of outbreaks, reporting was after 72 hours. The reporting was mainly through health staff. After receiving the information Medical Officer of the concerned primary health center visited the outbreak village within 24 hours in 17 outbreaks (36.96%), within 72 hours in 71.74% outbreaks and started control measures. Only 2 outbreaks were not visited within 5 days. Maximum outbreaks were reported in the month of June whereas maximum cases were seen in July. Another peak was also observed in September. It clearly indicated the bimodal distribution.

Table I: Water supply of affected villages.

Type of water supply Affected villages State source
No. (%)
Piped water 21 (53.85) 31.3
Well water 10 (25.64) 45.9
Hand pump/Bore well 7 (17.95) 19.2
Other 1 (2.56) 3.6
Total 39 (100) 100

The water supply of the affected villages is given in Table I. Piped water supply was commonest (53.85%) in dengue affected villages. The classification of villages in the state according to water supply is also given in this Table.

Table II: Dengue fever - Age group wise attack rate.

Age group (in years) Population at risk Cases Attack rate
No. (%)
<1 1525 17 (0.22) 1.11
1-4 5392 304 (3.97) 5.64
5-14 13100 1881 (24.55) 14.36
15-44 29340 3423 (44.68) 11.66
45-59 8046 1437 (18.76) 17.86
60+ 4631 599 (7.82) 12.93
Total 62034 7661 (100) 12.35

The median population of the affected village was 2800. Only one affected village was having population more than 10,000. The median number of persons affected per outbreak was 150. In 6 outbreaks affected persons were less than 50 and in equal number of outbreaks affected persons were more than 300. The attack rate is given in Table II. The overall attack rate was 12.35% and was not uniform for all ages. Minimum attack rate was observed in infants (1.11%) and maximum (17.86%) in the age group 45-59 years. Females were 884 per 1000 male patients. The median duration of the outbreaks was 16 days. Only 7% outbreaks lasted less than 5 days and 28% outbreaks lingered 26 or more days.

Table III: Dengue fever - Clinical manifestations.

Complaints Patients
No. (%)
Fever 7406 (100)
Headache 2858 (38.59)
Muscle and joint pain 799 (10.79)
Nausea/vomiting 437 (5.90)
Retrobulbar pain 112 (1.51)

Clinical data of 255 patients was not available.

The presenting clinical symptoms are given in Table III. Mild to moderate fever was invariably present in all the patients. The onset was acute and duration was less than a week. Along with fever, symptoms like headache and muscular and/or joint pains of varying severity were present in 38.59% and 10.79% cases. Nausea and/or vomiting accompanied fever in 5.90% cases. Few cases had another bout of fever.

Table IV: Serological results of suspected dengue fever cases.

Age group (in years) Total samples Laboratory findings % Recent dengue % Ever positive
Recent dengue Past flavi Negative
<1 1 0 0 1 0 0
1-4 2 0 0 2 0 0
5-14 62 13 11 38 20.96 38.70
15-44 288 67 95 126 23.26 56.25
45-59 84 16 34 34 1905 59.52
60+ 43 6 11 26 13.95 39.53
Total 480 102 151 227 21.25 52.70

A total of 871 sera samples were sent to National Institute of Virology (NIV). In 108 samples (inclusive of 6 from paired sera sample) evidence of recent infection with dengue virus was certain. Highest positivity or evidence of recent infection (23.26%) was observed among young adults (15-45 years age group). The positivity gradually decreased to 13.95% among persons aged 60 years or more. Overall positivity for recent dengue infection was 21.25%. Ever positivity of flavivirus group increased with age reaching about 60% among 45-59 years age group.

A total of 211 adult Aedes aegypti mosquitoes were collected. Virus isolation was successful from 17 out of these 211 mosquitoes. Applying criteria of isolation of DEN virus in adult aedes mosquito or evidence of recent infection (detection of IgM antibodies in sera), 46 outbreaks were confirmed as dengue fever outbreak.

Table V: Entomological investigations of affected villages.

Container No. of containers examined No. with presence of aedes larvae Container index Breeding preference ratio
Partly buried mud pots 648 187 28.85 1.54
Cement tanks 160 42 26.25 1.4
Drums 407 87 21.37 1.14
Metal pots 190 38 20.00 1.07
Others 919 80 8.7 0.47
Total 2324 434 18.67

X2=77.39; p<0.01

A total of 2,984 houses were surveyed for larval breeding. 343 houses were positive. The containers surveyed were 2,324. The number of positive containers was 434. The number of different containers surveyed including tyres and their Container Index (CI) and Breeding Perference Ratio (BPR) are given in Table V.

Table VI: Prediction of dengue fever epidemic in rural areas of Maharashtra State.

Investigation findings Dengue fever outbreak Total
Yes No
Aedes breeding (+) & PBS MP (-)
Yes 39 9 48
No 3 145 148
Total 42 154 196

Sensitivity: 92.85%; Specificity: 94.15%; Predictive value as dengue fever outbreak: 81.25%

Information regarding evidence of aedes breeding and PBS status of affected population was collected for 192 outbreaks. Positive predictive value was calculated from these findings. The criterion of evidence of aedes mosquito breeding was applied to fever outbreaks for prediction of dengue fever outbreak. With this criterion, the positive predictive value was 48%. When one more indicator, absence of malarial parasite in the peripheral blood smear of the patients was applied, the positive predictive value increased to 81.2%.


In an outbreak, once the etiology becomes evident, all the cases with similar symptomatology are usually diagnosed accordingly and control measures are initiated. The detailed investigation of all cases is carried on afterwards. It takes atleast 6-8 weeks for the availability of results of these investigations. The epidemiological factors strongly associated with the occurrence of disease if identified; will diagnose the outbreak of dengue fever in the field earlier. Then specific control measures may be started promptly and spread of epidemic is halted.

Dengue fever outbreaks reported from rural areas of Maharashtra state are included in this study. The outbreaks are not equally distributed throughout the state. The Marathwada region is severely affected. Konkan and east parts of Vidharbha are always spared. All the affected districts are known for chronic water scarcity problems. The overall incidence in the affected 11 districts is 8.02/100,000 per year.

The health staff has reported majority of the outbreaks and only 16% outbreaks are reported through Grampanchayats. Similar observations are noted for all types of outbreaks since long time. As there is no health worker for each village, it takes time to receive information by health worker and reporting. The reporting time gap can be reduced by making the gram panchayat responsible for reporting of outbreak. It can also be noted that reporting of outbreak is a constitutional responsibility of gram panchayat as local self-government.

Similarly, the response gap which is most important factor for initiating containment measures needs to be reduced. Guidelines issued by the health department reiterate visit by MO of PHC within 24 hours after receiving information of the outbreak. Only 17 (36.96%) outbreaks were attended within the indicated time period. The response time for fever outbreaks is almost twice as compared to response time for the diarrhoea outbreaks. This indicates that field staff gives more priority to diarrhoeal outbreaks probably because of possibility of high preventable mortality. Medical officers need to be oriented about the adverse effects of fever outbreaks if the outbreak is unattended for long time.

Maximum outbreaks as well as cases have been reported in the months of June and July. The second and smaller peak is observed in the month of September. In 1997 both peaks were delayed. Generally in the state the first peak is observed in the month of May and the second smaller peak in the month of August. These two peaks clearly indicate that, there are two patterns of transmission. During the first peak period i.e. in summer, the breeding places of Aedes are domestic containers like buried earthen pots, small cement tanks etc in which water is stored. In summer due to water scarcity emptying and cleaning of these containers is avoided. Field experience shows that, scarcity usually starts in the month of March. Aedes larvae start appearing in these containers from mid April. After critical increase in density, if the virus is introduced in the village, the outbreak is started at the juncture of summer and monsoon. The Marathwada circle, part of central Maharashtra, represents this pattern. These districts have high incidence rates.

The second pattern is upsurge of cases at the end of mansoon. The showering results in collection of water in sundry peridomestic containers and thereby breeding of Aedes. The rainfall that is not very heavy certainly creates environment favorable for breeding. The districts receiving medium rainfall represented the second pattern.

The districts that always escape from the dengue fever outbreak are the districts of western shore and tribal/forest areas of eastern Maharashtra. These areas receive very heavy rainfall.

The water storing habits of the people of rural areas in this state are product of quantity of available water, continuity of supply, rainfall, availability of containers etc. It is observed that tap water is the commonest supply of water (53.85%) among affected villages followed by well water (25.64%), hand pump/bore well (17.95%) and tanker (2.56%). This distribution when compared with state figures for rural Maharashtra, it was observed that, tap water was at highest risk (X2=10.81; p<0.02). The reason is twofold; electric motors are used for lifting the water from source for tap water supply scheme. Many a times water cannot be lifted due to irregular electric supplies. At the same time water supply scheme or the source is not sufficient to cater the village. Due to these reasons, the water supply of the village is irregular and insufficient. In such situations, almost every household has water storage capacity of one to two days requirement. As there is no guarantee that when the water will be available next time, water containers are never emptied. This is ideal situation for aedes mosquito breeding.

When the median population of the affected villages was compared with the median population of the villages in the state (859) it was observed that the bigger villages were more affected. Only 28.27% affected villages were having population less than 1000 (X2=17.41; p<0.01). It indicated that larger the village, greater was the risk. The towns and cities, however, were spared making the problem essentially of rural areas. In smaller villages, the water supply is usually through hand pumps or wells. Water from these places is available for 24 hours independent of electricity. So the practice of villagers is to store only the required quantity of water. Cleaning of utensils, bathing etc. activities are performed at water source only. Thus only drinking water is stored at house. The other reason for less reporting from smaller villages could be sometimes outbreaks in smaller villages may go unnoticed, as the absolute number of persons affected is very small.

Although all ages and both sexes are susceptible, affected females per 1000 affected males were only 884 (RD=4.17; p<0.01). The Aedes mosquito is a day biter. Clothes like sari always cover females during the day even while working. Due to this there are less number of mosquito bites to females. However, it is also believed that such observation in rural area is due to reporting bias in sex. In a village in Tamil Nadu females were more (53.5%) than males1. Sex preponderance was not observed in some studies2,3.

When the outbreak occurs in a village, high attack rate is observed. In reported outbreaks, the overall attack rate varies from 10 to 47%2-6. The attack rate was minimum among infants (1.11%) and highest among the age group of 45-59 years (17.86%). It was more or less similar in adults. Young adults are reported to suffer maximum7. Overall age groups affected are similar to observations recorded in earlier outbreaks that have occurred elsewhere1-3,7. In the outbreak reported from Tamil Nadu, children were more affected1. In Maharashtra State, the usual practice of community is to cover the infant completely for avoiding cold or disease. This is naturally useful to avoid mosquito bite.

Median duration of the outbreaks was 16 days, which reflects promptness and effectivity of the containment measures carried out by the health department. Most of the outbreaks reported elsewhere had lasted longer1-3,8,9.

Clinical features observed were similar to earlier outbreaks reported from Maharashtra and elsewhere2-4,8,9. A small town in Maharashtra reported rash in 20% patients4. Patients of dengue fever in Ajmer City also experienced rash and haemorrhages8. During the study very few patients reported retrobulbar pain, typical in dengue fever. It was not first complaint in any of the patients in all outbreaks. Similarly, the tourniquet test was performed on many of the patients. This was also not positive in any of the patients. It is noticed that there is great difficulty in looking for petechiae in dark skinned patients. Symptoms and signs suggestive of dengue haemorrhegic fever and shock syndrome were not reported from outbreaks in Maharashtra State.

A total of 102 patients out of 480 had IgM antibodies against dengue virus, alone or together with other flaviviruses. This suggested recent infection with dengue virus. The overall positivity was 21.25% for the recent infection. There was no statistically significant difference in various age groups for recent infection (X2=1.63; p>0.05). The seropositivity ranges from 6.25% to 75%. Ever positivity (evidence of dengue virus infection, recent or old) increased with age similar to the finding elsewhere11. The ever positivity increased with age reaching 59.52% in the age group of 45-59 years and declined afterwards. This finding indicates the wide circulation of dengue virus in the Maharashtra State.

In this study, the larval indices were calculated after occurrences of the outbreak, hence, are ought to be higher. The House Index (HI) was 11.49% and Bretau Index (BI) 14.54% in affected villages. The HI reported from various studies varies from 17.5% to 75%4,6,10. Variation of BI is much more than HI. In various outbreaks the range is 6 to 121%. An extensive study regarding Aedes aegypti in Maharashtra state revealed BI as high as 600 in Mumbai City in wet season. Generally a HI greater than 5% and BI greater than 20% for any village/locality is an indication that the area is dengue sensitive. In Maharashtra State, the breeding of aedes is generally observed more frequently in domestic water storage containers than artificial water collections around the house. There are not more than 2-3 water storage containers per house. So the BI is below critical level (<20%) in affected villages. The container index observed in various outbreaks ranged from 2.22% to 19%5,6,8-10. Container preference reflected as BPR was highest for partly buried earthen pots. Similar observations were recorded that too particularly in Marathwada region12, which had highest incidence. In most of the studies earthen pots, tubs, cement tanks showed highest container index1,5,8,10,12. Other containers were metal pots, tyres and other sundry containers like coconut shells, plastic wares etc. which also showed breeding of Aedes in different proportions1,5,8-11.

All containers of water potable/unpotable, inside the house/surrounding the house were considered together. Rubber tyres were grouped under others. As outbreaks are reported from rural areas, trade of old tyres and vulcanizing shops are not common in rural areas of the state. Similarly, coconut shells are almost exclusively seen on coastal areas, which are spared from dengue, hence, as expected breeding in coconut shells was not observed. The information about container indices was repeatedly stressed in the health education activities. The containers in affected villages in that priority are examined and cleaned by health personnel with the help of community members. Although containers are not classified as outdoor or indoor, in one study in Maharashtra, in all the seasons container index was found higher for outdoor containers12.

The isolation of DEN virus was possible only in 8.06% of Aedes aegypti caught while investigating the outbreaks. The evidence of presence of virus either through isolation or indirectly by immunoflorencent antibody test is usually obtained in same outbreaks. It ranges from 0% to maximum 55%3,5,9,10. Similarly, percentage of virus isolation from acute sera, in different studies reported, varied from 6.67% to 50%2-5,8,9.

It takes long time to receive laboratory results of the human and mosquito samples sent for investigation. The Public Health Department can not wait for the results for starting control measures. At the same time unnecessary introduction of hazardous and costly insecticide by spraying/fogging has also to be avoided. The authors have promoted following simple and feasible strategy. Firstly, evidence of aedes breeding is considered for prediction of dengue fever epidemic. The positive predictive value of this indicator is 48%, which is unsatisfactory. When second indicator, absence of malarial parasite in peripheral blood smear of the patients was added for prediction, the positive predictive value of duo increased to 81.2%. This means during the investigation of any epidemic, if breeding of aedes aegypti is seen and the peripheral blood smear of the patients is negative then the possibility of the dengue fever epidemic in that village is 81.2%.

Conclusions and Recommendations:

  1. It was observed during study that, fever epidemics are reported late. Rise in fever rate is not picked up early by health workers. Reorientation of health workers regarding the fever surveillance is recommended.
  2. The gap between occurrence of epidemic and receipt of information to medical officer of PHC is unacceptable. Considering that health worker is not posted in each village, the response gap can only be minimized by educating the Gram Panchayat about their constitutional responsibility of reporting any event in the village.
  3. Villages of more than 1,000 population with tap water supply and scarcity villages are at highest risk of getting Dengue Fever outbreaks. Health education of community in these villages regarding emptying, brushing and drying of water containers once a week should be started on priority. All the containers in village should be emptied on same day for maximum impact. The day should be suitable to community. This could be a day of holiday. On that day water supply of the village should be assured and more in quantity.
  4. It is extremely important to conduct entomological study of fever outbreak village. At present, the entomologists are available only at regional level. In such situation it is recommended that, atleast one health supervisor from each block should be adequately trained in recognizing the different types of larvae, collection and dispatching the mosquitoes and calculation of different entomological indices.
  5. All the medical officers should be reoriented regarding the investigation of fever outbreaks, basic entomology and keeping DF/DHF/DSS in mind during the investigation of outbreak.
  6. Prediction of outbreak as dengue fever outbreak was tried in this study with the help of two indicators, evidence of aedes breeding and presence of malarial parasite in PBS of fever patients. The positive predictive value is satisfactory (81.2%). It is recommended that, this finding should be checked in further studies and if consistent, utilized during the control measures of fever outbreaks.


  1. Abdul Koder MS et al: Outbreak and control of dengue in a village in Dharmapuri, Tamil Nadu. J Common Dis 1997; 29(1): 69-71.
  2. Mehendale SM et al: Outbreak of dengue fever in rural areas of Parbhani district of Maharashtra (India) Indian J Med Res 1991; (a) 93: 6-11.
  3. Doke PP et al: Investigation report of an epidemic of dengue fever. Indian Journal of Community Medicine 1991; XVI(3): 119-25.
  4. Prasad Rao GLN et al: Investigation of the 1975 dengue epidemic in Amalner town of Maharashtra. Indian J Med Res 1981; 74: 156-63.
  5. Mahadeo PVM et al: Dengue in Gujarat State, India during 1988 and 1989. Indian J Med Res 1993; (a) 97: 135-44.
  6. Jamaluddin M, Saxena VK: First outbreak of Dengue fever in typical rural area of Haryana state in northern India.
  7. Mohan Rao CVR: Dengue fever in India. Indian J Pediatric 1987; 54: 11-4.
  8. Ghosh SN et al: Investigations on the outbreak of dengue fever in Ajmer city. Rajasthan state in 1969. Part I Epidemiological, clinical and virological study of the epidemic. Indian J Med Res 1974; 62: 511-22.
  9. Padbidri VS et al: The 1993 epidemic of dengue fever in Mangalore, Karnataka state, India, Southeast Asian J Trop Ed Public Health, 1995; 26(4): 699-704.
  10. Mahadeo PVM et al: Aedes aegypti (L.) in Ahmedabad City, Gujarat: Distribution, Dengue Virus Detection and susceptibility to insecticides. J Common Dis 1993; 25(4): 169-83.
Access free medical resources from Wiley-Blackwell now!

About Indmedica - Conditions of Usage - Advertise On Indmedica - Contact Us

Copyright © 2005 Indmedica