We established KFD surveillance in 3 large public health facilities in Thirthahalli Taluk: Thirthahalli Taluk Hospital, Kannangi Community Health Center, and Konandur Primary Health Center. Patients from neighboring Hosanagara Taluk also seek care at these facilities. A suspected case was defined as sudden onset of fever, headache, and myalgia in patients attending these facilities during the last week of December 2013 through the first week of April 2014. Medical officers at surveillance facilities collected information regarding each patient’s age, sex, place of residence, and clinical and vaccination history. Serum samples from all suspected case-patients were tested for KFDV at the National Institute of Virology by using reverse transcription PCR (RT-PCR) (5). RT-PCR–negative samples were tested by ELISA for KFDV IgM (5). Samples were also tested for dengue virus and Leptospira spp. We analyzed the data to describe the disease by time, location, and person. We used the population of affected villages and applied the age- and sex-distributions of Karnataka State’s population (2011 census) to the population of the affected villages to calculate attack rates.

To calculate vaccine coverage, we obtained 2013 KFD vaccination data from district health officials. To estimate VE, we conducted a retrospective cohort study in 4 villages within the Kannangi Community Health Center catchment area: Garaga-Kikkeri (466 residents in 106 households), where KFD vaccination was conducted in 2013, and the neighboring villages of Kannangi, Kombinakai, and Avalagere (total of 528 residents in 146 households), where vaccination was not conducted. We systematically sampled 60 households from the vaccinated village and 110 from the nonvaccinated villages and collected information from persons 8–66 years of age about the number of KFD vaccine doses received in 2013. Information about laboratory-confirmed cases in vaccinated and unvaccinated persons was obtained from surveillance data. We estimated the relative risk (RR) of acquiring KFD associated with vaccination and then calculated VE as follows: VE = 1 − RR.

During December 31, 2013–April 7, 2014, facility-based surveillance identified 246 suspected cases of KFD; 106 (43.1%) patients were positive for KFDV (78 by RT-PCR, 28 by IgM ELISA); 1 case-patient was also positive for dengue-specific IgM. The laboratory results of suspected case-patients screened up to March 2014 are available elsewhere (6).

Figure 2

Figure 2. Distribution of 106 laboratory-confirmed Kyasanur Forest disease cases by date of symptom onset, Shimoga District, Karnataka State, India, December 2013–April 2014.

Of the 106 case-patients, 102 were from 41 villages in Thirthahalli Taluk (cumulative population 7,317), and 4 were from 3 villages (cumulative population 559) in Hosanagara Taluk (also in Shimoga District) (Figure 1). The overall attack rate was 13.5 cases/1,000 persons. In affected villages, cases were reported from all age groups; rates were highest among persons >15 years of age and among male residents (Table). Eighteen (16.9%) case-patients were in age groups not targeted for vaccination: 8 were <7 and 10 were >65 years of age. The cases began occurring during the last week of December 2013, peaked during February–March 2014, and then declined gradually (Figure 2).

Of the 106 case-patients, 95 (89.6%) reported that they had not been vaccinated and 11 (10.4%) reported being vaccinated (1 received 2 primary and 1 booster dose, 5 received 1 dose, and 5 received 2 doses). Ninety-one case-patients were from villages beyond a 5-km radius of an area with KFD activity in 2013 (i.e., outside an area targeted for vaccination in 2013).

During 2013 (before the outbreak), a total of 19,854 persons 7–65 years of age from Thirthahalli and Hosanagara Taluks had been targeted for KFD vaccination. The coverage of first, second, and booster doses was 23.4%, 15.4%, and 27.3%, respectively.

We included 176 KFD-vaccinated persons (26 received 1 dose, 150 received 2 doses) and 350 unvaccinated persons in the retrospective cohort study. The vaccinated and unvaccinated persons did not differ with respect to age, sex, or occupation (data not shown). Eight laboratory-confirmed KFD case-patients were reported from this cohort (7 were unvaccinated, 1 had received 2 doses of vaccine). The relative risk associated with 1 and 2 doses of vaccine was 0.96 (95% CI 0.06%–16.5%) and 0.33 (95% CI 0.04%–2.69%), respectively. The VE of 1 and 2 doses of vaccine was 4% (0%–96%) and 67% (0%–96%), respectively.

The findings of our investigation highlighted 4 concerns regarding KFD vaccination strategy practiced in the region. First, vaccine coverage in villages selected for vaccination in 2013 was low. Earlier studies also have shown that nearly half of the eligible population in the targeted villages was not vaccinated (4). These findings indicate low acceptance of KFD vaccine, possibly because of vaccine-associated adverse effects and the need for multiple doses. Second, the observed VE was lower than that reported in a previous study (1 dose, 79%; 2 doses, 94%) (7). However, other recent reports also indicate lower VE (2,4,8). Third, the occurrence of cases in areas >5 km away from villages vaccinated in the previous year suggests that targeting vaccination to areas within a 5-km radius of reported KFD activity may not be effective in preventing KFDV transmission outside the vaccinated areas. KFDV is primarily transmitted by the bite of infected ticks, and it is spread by the movement of monkeys that carry infected ticks; thus, vaccinating around zones with reported KFD activity is unlikely to prevent spread of the virus. Fourth, ≈17% of the patients in our study with laboratory-confirmed KFDV infection were <7 or >65 years of age, and persons of these ages are not administered KFD vaccine, probably because lower attack rates were observed among these age groups in earlier outbreaks (7,9,10).

Our study had 2 limitations. First, the facility-based surveillance relied on passive detection of case-patients seeking care from the selected health facilities. The number of case-patients detected is influenced by the health-seeking behavior of the community and by the severity of illness. Second, although VE was found to be low, the cohort study had low power (22.3%); hence, the findings of low effectiveness must be interpreted with caution.

To control KFDV, systematic efforts are needed to improve the current vaccine and vaccine coverage. Current vaccination strategies should be reviewed and the reasons for low VE should be evaluated.