Melioidosis is caused by B. pseudomallei, a Gram-negative bacterium that resides in soil and surface water in endemic regions. The disease has a varied clinical presentation from asymptomatic disease to chronic abscesses or pneumonia, to fulminant bacteremic disease. Outbreaks of melioidosis have been described following heavy rainfall and floods from Australia and Thailand.^1,2 

We report a cluster of cases of melioidosis over two months following heavy rainfall and floods in southeast India. The Coromandel Coast in southeast India includes the coastal areas of Tamil Nadu, Puduchery and Andhra Pradesh, as shown in the map in Figure 1. Important cities lying in this area are Chennai, Neyveli and Nellore. This region receives annual rainfall as a result of the northeast monsoon during the months of November- December. In the year 2015, the Coromandel Coast received heavy rainfall of 483 mm on November 9-10 and 272 mm on December 1, resulting in flooding of these regions. Unprecedented rainfall has probably resulted in this case clustering.

Figure 1. Coromandel coast in South India highlighting the flood affected area

Case series

Following heavy rainfall and flooding of the southeast coast of India, the department of Infectious Diseases, Apollo Hospitals and Apollo Specialty Hospital, Chennai, witnessed an increase in the number of cases of melioidosis from November 2015 until 8 January 2016, in hospitalized patients as well as in outpatient clinics. A retrospective chart review of patients with melioidosis from two tertiary care hospitals in Chennai was carried out. All patients with blood or any other sterile site specimen growing B. pseudomallei between 9 November 2015 and 8 January 2016 (a month after the floods) were studied and compared with an identical time period in the following year. Blood cultures were performed with the BacT-Alert automated blood culture system (bioMérieux, Marcy-l’ Étoile, France) using brain heart infusion broth and incubated at 37°C. Positive cultures were subcultured in appropriate medium; identification was done on Vitek 2 system (bioMérieux) and drug susceptibility testing was performed manually by the Kirby Bauer disc diffusion susceptibility method. All other specimens were cultured on the following media: sheep blood agar, chocolate agar, MacConkey agar, with isolate identification through standard microbiological techniques.

Patient characteristics and epidemiological links with flood exposure are depicted in Table 1 (Appendix). Twelve patients with blood culture growing B. pseudomallei were identified during this 2 month period. All were from the flood affected area. Eleven cases (91%) had definite flood water exposure in the form of wading through water, mainly walking barefoot in flooded water, in some cases up to chest deep. None of them had aspirated flood water. Among the 12, only 3 (cases 3, 7 and 11) had occupational risk in the form of farming. All presented with fever. The time period between exposure to flood or rain and onset of symptoms ranged between 5 to 24 days. Eight of the 12 patients with a history of exposure to flood water had an incubation period less than 14 days. The patients with pneumonia and sepsis syndrome (cases 1, 2 and 8) had a brief incubation period of 5 days and symptom duration of 5-8 days. The initial case (case 1) was reported on 15 November 2015 with a short incubation period of 5 days and presented with severe community-acquired pneumonia and sepsis ending fatally. Similar presentations were noted in cases 2 and 8, who presented with a brief duration of symptoms, had sepsis syndrome, and both eventually succumbed to the illness. Those with localized abscesses and bacteremia had longer incubation periods of 6-19 days and also had longer duration of symptoms. Seven had septic foci elsewhere and 4 had one other site which was culture positive. Five patients had soft tissue or joint infections. Cases 3 and 5 presented with fever, elbow joint pain and had features of septic arthritis with synovial fluid growing B. pseudomallei. Cases 6 and 12 had forearm abscesses and case 10 had diabetic foot infection. Three had visceral organ involvement in the form of prostatic abscess, pyelonephritis, epididymitis and parotitis. The majority (83%) had diabetes, 4 suffered from alcoholism, one had chronic kidney disease, 2 had pre-existing skin lesions (eczema and diabetic foot infection). Three patients (cases 1, 6 and 8) succumbed to sepsis after a brief period (hours) of hospitalization and in these cases the diagnosis was confirmed after the patients’ death, when cultures grew B. pseudomallei. Carbapenems were used for empiric therapy in patients with sepsis, while definitive initial therapy for melioidosis was performed with ceftazidime, followed by oral eradication treatment with co-trimoxazole with or without doxycycline. Overall the mortality was 33%. Cases that presented with acute severe community-acquired pneumonia and sepsis had 100% mortality.

The data on incidence of melioidosis in India are lacking, except for some case series from a few south Indian centers. Hence the true disease burden is largely unknown. Although hospital records show that there is usually about one case per month diagnosed in our center in the preceding and succeeding periods, there were 4 cases in November 2015, 7 in December 2015 and 2 in January 2016 (Figure 2).

Figure 2. Monthly distribution (number) of melioidosis cases (2015-2016)

The number of cases during the same time period during the following year were analyzed. The number of bacteremic cases are shown in Figure 3. In December 2016, the Coromandel coast received an average rainfall of 100 mm but there was no flooding. From November 2016 until January 2017, 5 cases of melioidosis were diagnosed in the same hospitals among which 2 were from the Coromandel coast (Chennai and Tirupathi) and 3 cases came from northeast India, and none of them had flood exposure. Among these 5 cases, 3 had bacteremia and 2 had pus growing B. pseudomallei. By applying a two tailed test of significance, the z-score was 2.3324 and p=0.019) indicating that the difference is statistically significant. Thus there was a demonstrable significant increase in cases with bacteremic melioidosis in these 2 centers following heavy rainfall and floods as compared to the same time period the following year.

Figure 3. The number of bacteremic melioidosis cases during the 2-month span in 2015 as compared to 2016

Melioidosis is largely a tropical disease that is endemic in southeast Asia, notably in Thailand and in northern Australia.³The number of cases from south Asia in 2015 was estimated to be 73,000 (44% of the global burden) which is higher than that from East Asia and the Pacific, although far fewer than these were reported.⁴ B. pseudomallei is a soil pathogen and is also present in surface water in endemic regions. In a study by Prakash et al., B. pseudomallei was isolated from the soil samples of paddy fields in the coastal region of Tamil Nadu, South India confirming the presence of the organism in this region.⁵ India is predicted to be environmentally suitable and endemic for B. pseudomalleiand melioidosis is increasingly being reported from India.^6-10

We found a significant increase in cases with bacteremic melioidosis in these 2 centers following heavy rainfall and floods as compared to the same time period the following year. In a previous study from the same center, 32 cases were identified over a period of 6 years, the majority being from southeast India; no clustering or association with rainfall was noted.⁹

Among the 12 patients in this study, 11 had definite flood exposure. There are several studies from South India, northern Australia, and Thailand indicating an association between rainfall and an increase in the occurrence of melioidosis. In previous studies from southwest India the intensity of rainfall correlated with the incidence of melioidosis.^7,8In a 12-year prospective study that included 318 culture confirmed cases from northern Australia there was a correlation between rainfall in the previous 14-day period and subsequently developing fatal and pulmonary melioidosis. Heavy rains and winds are thought to be risk factors for inhalational melioidosis.¹ In their prospective study of 540 cases over 20 years, Currie et al. found that 81% of cases occurred during the rainy season.¹¹In another study from Thailand, it was also noted that there is a significant seasonal increase in incidence of melioidosis with rainfall.¹² It is also interesting to note that an increased incidence of melioidosis occurred among survivors of a tsunami in Indonesia and typhoon in Taiwan.^2,13

The reasons postulated for an increase in incidence of melioidosis following rainfall and winds include aerosolization of bacteria from soil or surface water due to heavy winds and rainfall, leading to a larger bacterial inoculum.¹ Another reason postulated is the transportation of B. pseudomallei from deeper layers of the soil to the surface due to rising water table.¹⁴

The incubation period in our study ranged between 5 and 24 days. Eight out of 12 patients with history of exposure to flood water had an incubation period less than 14 days. In the 12-year prospective study that included 318 cases in the northern territory of Australia, by Currie et al., the median rainfall in the preceding 14 days was an independent risk factor for mortality, patients having 2.5 times the risk of dying from melioidosis if the rainfall in the preceding 14 days was more than 125 mm.¹ Most of our patients had diabetes (83.3%), which is the commonest host risk factor for melioidosis whilst four had alcoholism, and one had chronic kidney disease which are all considered as risk factors for melioidosis.¹¹ Two had skin lesions (eczema and diabetic foot infection) which perhaps served as portals of entry for B. pseudomallei when wading through flood waters.

All patients in the study were bacteremic and 2 had pneumonia. Four of 12 patients (33.3%) all of whom presented with a clinical picture of sepsis, hypotension and metabolic acidosis, had a fatal outcome. Two cases in our study, both of whom presented after a brief duration of symptoms as sepsis syndrome, died a few hours after admission with cultures becoming positive afterwards. Two patients had forearm abscesses, two had septic arthritis, one patient each had epididymitis, parotitis, and one had both pyelonephritis and prostatic abscess, all of which are well recognized manifestations of melioidosis.

Most of our patients were treated either with a carbapenem or with ceftazidime; patients initially started on empiric carbapenem were later switched to ceftazidime after confirmation of etiology and antibiotic susceptibility report. Co-trimoxazole and/or doxycycline were continued in the eradication phase for 3 months with good response.

A significant increase in melioidosis cases, especially bacteremic melioidosis, after flood exposure was noted in our center. Diabetes was the most common comorbidity found in our study and a third of patients had fatal outcome. After floods and heavy rainfall, a high index of suspicion is warranted for melioidosis in patients, especially diabetics, presenting with sepsis syndrome, pneumonia or abscesses, and empiric therapy may be warranted until the disease is excluded. 

1.CurrieBJ, Jacups SP. Intensity of rainfall and severity of melioidosis, Australia. Emerg Infect Dis 2003;9:1538-42. [Crossref]

2.Athan E,Allworth AM,Engler C, Bastian I,Cheng AC. Melioidosis in tsunami survivors. Emerg Infect Dis 2005;11:1638-9. [Crossref]

3.Leelarasamee A. Melioidosis in Southeast Asia. Acta Trop2000;74:129-32. [Crossref]

4.Limmathurotsakul D,Golding N, Dance DA,et al. Predicted global distribution ofBurkholderia pseudomalleiand burden of melioidosis. Nat Microbiol2016;1 pii: 15008. [Crossref]

5.Prakash A,Thavaselvam D,Kumar A,et al. Isolation, identification and characterization ofBurkholderia pseudomalleifrom soil of coastal region of India. Springerplus2014;3:438. [Crossref]

6.Jesudason MV,Anbarasu A,John TJ. Septicaemic melioidosis in a tertiary care hospital in south India. Indian J Med Res2003;117:119-21.

7.Antony B,Pinto H,Dias M,et al. Spectrum of melioidosis in the suburbs of Mangalore, S West Coast of India. Southeast Asian J Trop Med Public Health2010;41:169-74.

8.Vidyalakshmi K,Lipika S,Vishal S,Damodar S,Chakrapani M. Emerging clinico-epidemiological trends in melioidosis: analysis of 95 cases from western coastal India. Int J Infect Dis2012;16:e491-7. [Crossref]

9.Gopalakrishnan R, Sureshkumar D , Thirunarayan MA , Ramasubramanian V. Melioidosis: an emerging infection in India. J Assoc Physicians India2013;61:612-4.

10.Ray U, Dutta S, Ramasubban S, Sen D, Tiwary IK. Melioidosis: series of eight cases. J Assoc Physicians India2016;64:42-6.

11.Currie BJ,Ward L,Cheng AC. The epidemiology and clinical spectrum of melioidosis: 540 cases from the 20 year Darwin prospective study. PLoS Negl Trop Dis2010;4:e900. [Crossref]

12.Suputtamongkol Y,Hall AJ,Dance DA, et al. The epidemiology of melioidosis in Ubon Ratchatani, northeast Thailand. Int J Epidemiol1994;23:1082-90. [Crossref]

13.Ko WC,Cheung BM,Tang HJ,et al. Melioidosis outbreak after typhoon, Southern Taiwan. Emerg Infect Dis 2007;13:896-8. [Crossref]

14.Thomas AD,Forbes-Faulkner J,Parker M. Isolation of Pseudomonas pseudomallei from clay layers at defined depths. Am J Epidemiol1979;110:515-21. [Crossref]