- Email: [email protected]
Melioidosis in a rural community of Western Province, Papua New Guinea
Transactions of the Royal Society of Tropical Medicine and Hygiene (2007) 101, 809—813
available at www.sciencedirect.com
journal homepage: www.elsevierhealth.com/journals/trst
Melioidosis in a rural community of Western Province, Papua New Guinea J.M. Warner a,∗, D.B. Pelowa b, B.J. Currie c, R.G. Hirst a a
Microbiology and Immunology, School of Veterinary and Biomedical Sciences, James Cook University, Townsville, Qld 4811, Australia b Balimo Rural Hospital, Balimo WP, Papua New Guinea c Menzies School of Health Research, Charles Darwin University, Darwin, NT 0811, Australia Received 19 July 2006; received in revised form 22 February 2007; accepted 23 February 2007 Available online 11 May 2007
KEYWORDS Melioidosis; Burkholderia pseudomallei; Chloramphenicol; Co-trimoxazole; Doxycycline; Papua New Guinea
Summary A prospective study was conducted to determine the significance of melioidosis in the Balimo district of Western Province, Papua New Guinea. During 1998, after the establishment of laboratory procedures and increasing local clinical awareness, the disease was found in 1.8% (95% CI 0.37—5.1%) of individuals presenting with fever refractory to standard treatment. The clinical incidence was 20.0 per 100 000 population (95% CI 12.2—30.9). The median age of culture-confirmed cases was 9.5 years (interquartile range 8.3—14.8 years). The seroprevalence of 747 community children in the region tested was 8.2% (95% CI 6.2—10.4%). Most individuals presented during the rainy season with a febrile disease refractory to standard treatment, sometimes mimicking tuberculosis. Some family clustering was apparent. All patients with bacteraemic melioidosis died, but treatment with the available conventional therapies of chloramphenicol, co-trimoxazole or doxycycline resulted in survival and cure in six patients with subacute/localised melioidosis. Further studies are needed to ascertain the local epidemiology and why children appear particularly at risk, as well as to establish the true extent of melioidosis in Papua New Guinea. © 2007 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved.
1. Introduction Melioidosis is an infection caused by the saprophytic Gramnegative bacillus Burkholderia pseudomallei. Melioidosis
∗
Corresponding author. Tel.: +61 7 4779 6375; fax: +61 7 4779 1526. E-mail address: [email protected] (J.M. Warner).
is emerging as a significant cause of community-acquired sepsis in the tropics, but the great variety of clinical presentations and difficulties with laboratory diagnosis make accurate determination of the distribution problematic (Cheng and Currie, 2005; White, 2003). Whilst there have been sporadic reports of melioidosis from Papua New Guinea, seroprevalence studies from the capital city of Port Moresby indicated that limited or no exposure to B. pseudomallei occurs in that region of the country (Barnes et al., 1991; Currie, 1993; De Buse et al., 1975; Kingston,
0035-9203/$ — see front matter © 2007 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.trstmh.2007.02.024
810 1971; Lee and Naraqi, 1980; Newland, 1969; Rowlands and Curtis, 1965). However, an unpublished report suggested that melioidosis may be endemic in the rural Balimo region of the Western Province of Papua New Guinea (Reece, 1984). We have therefore undertaken a prospective study to determine the extent of autochthonous melioidosis within this rural community.
2. Materials and methods Balimo is the rural district centre of the Gogodala language group who number approximately 30 000. Less than 10% live in the district township and associated villages, with most being subsistence gardeners, fisherman and hunters scattered along the Aramia River and associated flood plain (Figure 1). The study was conducted at Balimo Health Centre with the support of staff and the community. Burkholderia pseudomallei solid and liquid selective and differential media were prepared as described by Ashdown (1979) with 8 mg/l gentamicin. Organisms were presumptively identified as B. pseudomallei based on typical colonial appearance, oxidase reaction (positive) and gentamicin resistance inferred from vigorous growth on ASH agar (Dance et al., 1989). Identification of each isolate was corroborated with API 20NE (bioM´ erieux, Baulkham Hills, NSW, Australia) or Microbact 24E (MedVet, Adelaide, SA, Australia) and confirmed using two PCR protocols (Dharakul et al., 1999; Winstanley and Hart, 2000). The indirect haemagglutination assay (IHA) was used for serology as per the methods of Ashdown (1987). A ‘maybe melioidosis’ campaign was established at the health centre, which included increasing clinical awareness and establishing laboratory diagnostics. Consideration of melioidosis included any patient presenting with community-acquired pneumonia or sepsis and any patient with persistent fever refractory to standard treatment (typically antimalarials and short course tuberculosis treatment). These patients were investigated by serology plus selective culture of sputum and/or gastric aspirate and throat swabs. Blood cultures were taken on all patients with persistent fever and reactive IHA serology.
Figure 1 The rural Balimo district of Western Province, Papua New Guinea, showing the Aramia flood plain at the height of the wet season and the location of village communities, demonstrating their very close association with water.
J.M. Warner et al. Confirmed cases were those with a positive culture for B. pseudomallei; suspected melioidosis was defined as those with clinically compatible illness plus evidence of either direct familial association during a confirmed outbreak or IHA seroreactivity. Treatment regimens were based on recommendations at the time (Sookpranee et al., 1992), although limited availability of antibiotics necessitated modifications. Notably, ceftazidime, which is the drug of choice for melioidosis with sepsis, was not available. For those presenting with acute sepsis, initial treatment included a two-drug combination of i.v. chloramphenicol and/or doxycycline and/or co-trimoxazole for 3 weeks. In survivors, this was followed by oral therapy for 3 months with one or two of the above antibiotics. Those with fever without sepsis or with just localised disease were treated with 3 months of oral therapy with one or two of the above antibiotics. As rainfall in Balimo is not recorded, the average monthly rainfall data for 1994—1995 and 1998 at the provincial capital of Daru were compared with the date of diagnosis of each confirmed melioidosis case, as Daru rainfall generally reflects that seen in Balimo. Five regional community schools were chosen for a seroprevalence survey of children. The schools were all were located within a 10 km radius of the health centre and represented the majority of children of that age group in the immediate Balimo region. Sera were separated and stored at −20 ◦ C until assayed by IHA. Seroprevalence was calculated per school. An IHA titre of 1:40 was chosen as the cut-off value for a positive reactor, consistent with its use for diagnostic purposes. Regional differences in seroprevalence were compared using Fisher’s exact test with 95% CI.
3. Results Culture-confirmed cases and suspected cases of melioidosis as well as their outcomes are presented in Table 1. Case AW was described in the original 1984 report by Reece (1984). During 1994—1998, two periods of melioidosis screening were conducted, totalling 16 months of active screening. Unfortunately, detailed sampling data from the portion of the study conducted in 1994—1995 were lost, with only the results of the culture-confirmed cases remaining (Table 1). However, during the 1998 study (March—July) 245 samples from 170 patients were selectively cultured as per the protocols listed above. Accurate dates of birth are difficult to determine in rural Papua New Guinea, therefore patient age was classified as either ‘child’ or ‘adult’ based on physical maturity. However, when dates of birth were able to be determined through review of medical records (more likely in young children born at the health centre) age was recorded. Of the patients screened in the 1998 study, 104 (61%) were adults and 66 (39%) were children. The samples selectively cultured included 1 ascitic fluid, 1 eye swab, 2 ear swabs, 25 throat swabs, 37 blood cultures, 54 skin lesions and abscess pus and 125 sputa or gastric aspirates. Of the 170 patients sampled in 1998, 3 were culture-positive for B. pseudomallei (Table 1). During the total 16 months of the study, eight cases of culture-confirmed melioidosis autochthonous to the Balimo region were diagnosed. A further two cases of suspected
Melioidosis in a rural community of Western Province, PNG
811
Table 1
Culture-confirmed or suspected melioidosis cases documented in Balimo during 1983, 1994—1995 and 1998
Patient
Familya
Sex
Age (years)
Date
Initial presentation
Culture
IHA
Village
Outcome
1983 AW
*
M
18
Jan. 1983
Fever, respiratory symptoms
Blood
NT
Adiba
Died
1994—1995 KW — ID — SS * GD — DS ** KimS ** KawS **
M F M F F M M
10 9 6 9 14 7 24
Jan. 1994 June 1994 May 1994 Jan. 1995 Jan. 1995 Jan. 1995 Feb. 1995
Fever, respiratory symptoms Fever, groin abscess Persistent fever Fever, respiratory symptoms Fever Fever Fever
Blood Blood NT Sputum Blood NT Blood
NT NT 1024 >5120 NT 320 NT
Adiba Kini Adiba Adiba Kimama Kimama Kimama
Died Died Well Well Died Well Died
1998 NG TG RI
F M M
4 6 17
May 1998 May 1998 July 1998
Fever, respiratory symptoms Fever Fever
Sputum Throat Abscess
40 >5120 <5
Balimo Balimo Togowa
Well Well Well
*** *** —
IHA: indirect haemagglutination assay; NT: culture or serology not available or organism not recovered. a Patients with the same symbol (*, ** or ***) are members of same immediate family.
melioidosis were empirically diagnosed based on strong clinical suspicion, serology and response to melioidosis-directed therapy (Table 1). The median age of the culture-confirmed cases was 9.5 years (interquartile range 8.3—14.8 years). According to medical records, all individuals presented initially with subacute/localised disease diagnosed with or without respiratory symptoms and signs, although several developed severe pneumonia and died. In each case during their admission, before a diagnosis of melioidosis was made, short-course tuberculosis treatment was considered. The annual incidence of culture-confirmed melioidosis in the population was 20.0 per 100 000 population (95% CI 12.2—30.9). This figure is likely an underestimate as it assumes that the total population of the Gogodala language group had access to the Balimo Health Centre. The prevalence of culture-confirmed melioidosis within the patient cohort during March—July 1998 was 1.8% (95% CI 0.37—5.1%). All four bacteraemic melioidosis patients died, with a delay in diagnosis noted for each case. In contrast, all six individuals diagnosed with localised/subacute melioidosis recovered with the use of melioidosis-directed therapy. No recurrent infection has subsequently been reported in these survivors. Figure 2 demonstrates that melioidosis cases occurred during or immediately after the wet season between December and May. A total of 747 children aged 8—13 years participated in the serological study. The seroprevalence of each regional group is presented in Figure 3. The overall seroprevalence was 8.2% (95% CI 6.2—10.4%). The seroprevalences of Adiba and Buila schools were 15.2% and 16.9%, respectively, and these were significantly higher than the other three centres (Fisher’s exact test, P < 0.001).
Figure 2 Mean annual rainfall for 1994—1995 and 1998 in Daru, and number of melioidosis cases. Columns indicate monthly rainfall; points indicate number of melioidosis cases.
rural regions approaching 100 per 1000 births and an average life expectancy of 52 years (Papua New Guinea Ministry of Health, 2000). Pneumonia is the leading cause of death in each province. Signs and symptoms of melioidosis can be extremely variable. Whilst pneumonia is the commonest
4. Discussion Papua New Guineans experience the poorest of health outcomes in the Pacific region, with childhood mortality in
Figure 3 Seroprevalences in the Balimo district community school population using a cut-off indirect haemagglutination assay titre of ≥40. Error bars signify binomial 95% CI.
812 presentation of melioidosis overall, there can be many other manifestations, and diagnosis is dependent upon microbiology that may not be available in regions of possible endemicity (Cheng and Currie, 2005; White, 2003). In this study, improving the diagnostic capacity of a rural health centre has demonstrated autochthonous melioidosis in a rural community in Papua New Guinea for the first time. Furthermore, early diagnosis with the use of selective media enabled prompt directed therapy, resulting in those with localised disease making a full recovery. Although numbers are small, a feature of melioidosis in this region is the childhood predilection. Children were thought to acquire the organism through their robust behaviour within the environment, such as washing and playing in the lagoon and playing in the mud at the beginning of the wet season (all common practices). Clustering of cases within family groups reflects presumptive shared exposure of a common reservoir. Of note, a recent report from rural northeastern Brazil describes a similar newly recognised endemic focus of melioidosis where mostly children have been involved and where a local water source has been implicated (Rolim et al., 2005). Of the nine culture-confirmed cases documented, seven represent three outbreaks from three isolated regions, which also share the highest seroprevalence in their community schoolchild populations. This apparent spatial clustering of seroprevalence and clinical incidence may well reflect differences in the localised persistence of B. pseudomallei in the environment, as occurs in other melioidosis-endemic regions (Cheng and Currie, 2005). It is possible that in these locations the washing practices and environmental conditions may enable acquisition of B. pseudomallei via the respiratory or nasopharyngeal routes, either from water as an endogenous reservoir or from water contaminated with soil after times of heavy rain. Furthermore, anecdotal evidence from Adiba village suggested that the 1983 case and other undocumented suspected melioidosis cases were associated with soil excavation during house building. The accepted co-morbidity factors for acute-onset bacteraemic melioidosis, such as diabetes, liver disease and chronic renal disease (Currie et al., 2000; Suputtamongkol et al., 1999), are uncommon in Balimo. This, together with the heavy environmental exposure in children, may explain why melioidosis was less common in the adult population in this study. Melioidosis sepsis in adults may well become more common in this region if diseases such as type 2 diabetes become endemic, as has occurred in many periurban village communities in Port Moresby and in other Pacific nations as well as in the indigenous Australian population. The overall melioidosis seroprevalence of this district (6.2—10.4%) is similar to the seroprevalence of the Top End of the Northern Territory of Australia (5—13%). The Top End melioidosis annual incidence of between 16.5—34.5 per 100 000 (Currie et al., 2000) is also similar to the estimated annual incidence in the Balimo district of 20.0 per 100 000 population. Further clinical surveillance of this and other regions of Papua New Guinea is required to establish the true burden of melioidosis in the country. Given the extreme limitations on health resources in most of Papua New Guinea, it is possible that melioidosis is more widespread than currently recognised and may in some other locations be a major cause of community-acquired sepsis.
J.M. Warner et al. The relatedness of B. pseudomallei isolates from this study to other isolates from Papua New Guinea and also to those from the Torres Strait and mainland Queensland could provide insights into local melioidosis biogeography. Rainfall has been implicated as a clear factor that predisposes both the extent and intensity of melioidosis in endemic communities (Currie and Jacups, 2003). As in northern Australia and Thailand, the cases in the Balimo region were clustered in the rainy season. In Balimo, cases appeared particularly with the first rain after a prolonged dry season (Figure 2). The onset of the rainy season is when local health authorities should be on heightened alert. In conclusion, the Balimo region of Papua New Guinea is endemic for melioidosis, and melioidosis should be considered in the differential diagnosis of patients presenting with febrile illness from this region, particularly if the fever is refractory to standard treatment. Further studies are needed to ascertain the local epidemiology and why children appear particularly at risk and why family clustering occurs, as well as to establish the true extent of melioidosis in Papua New Guinea. Authors’ contributions: JMW, DBP and RGH designed the study protocol; JMW and DBP undertook the field work; BC provided clinical input; RGH analysed the data; JMW, RGH and BC drafted the manuscript. All authors read and approved the final manuscript. JMW is guarantor of the paper. Acknowledgements: The following are gratefully acknowledged: the staff of Balimo Health Centre for clinical support; Dr David Learoyd for help with clinical assessment and access to medical records during the 1994—1995 portion of the study and, as medical superintendent, the opportunity to conduct this study during the period of employment at Balimo in 1994 and 1995; village and school communities of Buila, Balimo, Adiba, Kimama and Saweta; Dr Kath Donovan for access to Dr Stuart Reece’s report; Joy Koehler and Bruce Samway for access to IHA antigen and help with the serology assay; the School of Public Health and Tropical Medicine, James Cook University, particularly Dr Wayne Melrose, Dr Paul Turner and Dr Jeffrey Hii, for laboratory consumables; and Mr Jimmy Gomoga, Deputy Director of Meteorology, for the rainfall data and advice on the suitability of using Daru rainfall data to reflect seasonal rainfall patterns at Balimo. Funding: None. Conflicts of interest: J.M. Warner was employed at Balimo Health Centre during 1994 and 1995. D.B. Pelowa was employed at Balimo Health Centre during the whole time of the study. J.M. Warner has family links to Papua New Guinea, but not to the Balimo region. D.B. Pelowa has family links to the Gogodala language group in and around Balimo. Although the authors have a conviction to improve the public health of rural Papua New Guineans, this is not considered a conflict of interest, simply a professional interest. Ethical approval: Papua New Guinea Medical Research and Advisory Council.
Melioidosis in a rural community of Western Province, PNG
References Ashdown, L.R., 1979. Identification of Pseudomonas pseudomallei in the clinical laboratory. J. Clin. Pathol. 32, 500—504. Ashdown, L.R., 1987. Indirect haemagglutination test for melioidosis. Med. J. Aust. 147, 364—365. Barnes, D.J., Gottlieb, T., Naraqi, S., Benn, R., 1991. The role of viruses and atypical organisms in the pathogenesis of adult pneumonia in Papua New Guinea. P. N. G. Med. J. 34, 13—16. Cheng, A.C., Currie, B.J., 2005. Melioidosis: epidemiology, pathophysiology, and management. Clin. Microbiol. Rev. 18, 383—416. Currie, B., 1993. Melioidosis in Papua New Guinea: is it less common than in tropical Australia? Trans. R. Soc. Trop. Med. Hyg. 87, 417. Currie, B., Jacups, S.P., 2003. Intensity of rainfall and severity of melioidosis, Australia. Emerg. Infect. Dis. 9, 1538—1542. Currie, B.J., Fisher, D.A., Howard, D.M., Burrow, J.N., Lo, D., SelvaNayagam, S., Anstey, N.M., Huffam, S.E., Snelling, P.L., Marks, P.J., Stephens, D.P., Lum, G.D., Jacups, S.P., Krause, V.L., 2000. Endemic melioidosis in tropical northern Australia: a 10-year prospective study and review of the literature. Clin. Infect. Dis. 31, 981—986. Dance, D.A., Wuthiekanun, V., Naigowit, P., White, N.J., 1989. Identification of Pseudomonas pseudomallei in clinical practice: use of simple screening tests and API 20NE. J. Clin. Pathol. 42, 645—648. De Buse, P.J., Henderson, A., White, M., 1975. Melioidosis in a child in Papua New Guinea successful treatment with kanamycin and trimethoprim—sulphamethoxazole. Med. J. Aust. 2, 476—478. Dharakul, T., Tassaneetrithep, B., Trakulsomboon, S., Songsivilai, S., 1999. Phylogenetic analysis of Ara+ and Ara− Burkholderia pseudomallei isolates and development of a multiplex PCR procedure for rapid discrimination between the two biotypes. J. Clin. Microbiol. 37, 1906—1912.
813 Kingston, C.W., 1971. Chronic or latent melioidosis. Med. J. Aust. 2, 618—621. Lee, L., Naraqi, S., 1980. Primary gram negative pneumonia in adults in Papua New Guinea. P. N. G. Med. J. 23, 174—178. Newland, R.C., 1969. Chronic melioidosis: a case in Sydney. Pathology 1, 149—152. Papua New Guinea Ministry of Health, 2000. National Health Plan 2001—2010, Policy Directions and Priorities. Papua New Guinea Ministry of Health, Port Moresby. Reece, A.S., 1984. A Clinical Report: the experience of the Balimo Health Centre with the clinical syndrome of melioidosis, including a suggested diagnostic, therapeutic and epidemiological program. Balimo Health Centre, Balimo. Rolim, D.B., Vilar, D.C., Sousa, A.Q., Miralles, I.S., de Oliveira, D.C., Harnett, G., O’Reilly, L., Howard, K., Sampson, I., Inglis, T.J., 2005. Melioidosis, northeastern Brazil. Emerg. Infect. Dis. 11, 1458—1460. Rowlands, J.B., Curtis, P.G., 1965. A case of melioidosis in Papua and New Guinea. Med. J. Aust. 2, 494—496. Sookpranee, M., Boonma, P., Susaengrat, W., Bhuripanyo, K., Punyagupta, S., 1992. Multicenter prospective randomized trial comparing ceftazidime plus co-trimoxazole with chloramphenicol plus doxycycline and co-trimoxazole for treatment of severe melioidosis. Antimicrob. Agents Chemother. 36, 158—162. Suputtamongkol, Y., Chaowagul, W., Chetchotisakd, P., Lertpatanasuwun, N., Intaranongpai, S., Ruchutrakool, T., Budhsarawong, D., Mootsikapun, P., Wuthiekanun, V., Teerawatasook, N., Lulitanond, A., 1999. Risk factors for melioidosis and bacteremic melioidosis. Clin. Infect. Dis. 29, 408—413. White, N.J., 2003. Melioidosis. Lancet 361, 1715—1722. Winstanley, C., Hart, C.A., 2000. Presence of type III secretion genes in Burkholderia pseudomallei correlates with Ara(−) phenotypes. J. Clin. Microbiol. 38, 883—885.
available at www.sciencedirect.com
journal homepage: www.elsevierhealth.com/journals/trst
Melioidosis in a rural community of Western Province, Papua New Guinea J.M. Warner a,∗, D.B. Pelowa b, B.J. Currie c, R.G. Hirst a a
Microbiology and Immunology, School of Veterinary and Biomedical Sciences, James Cook University, Townsville, Qld 4811, Australia b Balimo Rural Hospital, Balimo WP, Papua New Guinea c Menzies School of Health Research, Charles Darwin University, Darwin, NT 0811, Australia Received 19 July 2006; received in revised form 22 February 2007; accepted 23 February 2007 Available online 11 May 2007
KEYWORDS Melioidosis; Burkholderia pseudomallei; Chloramphenicol; Co-trimoxazole; Doxycycline; Papua New Guinea
Summary A prospective study was conducted to determine the significance of melioidosis in the Balimo district of Western Province, Papua New Guinea. During 1998, after the establishment of laboratory procedures and increasing local clinical awareness, the disease was found in 1.8% (95% CI 0.37—5.1%) of individuals presenting with fever refractory to standard treatment. The clinical incidence was 20.0 per 100 000 population (95% CI 12.2—30.9). The median age of culture-confirmed cases was 9.5 years (interquartile range 8.3—14.8 years). The seroprevalence of 747 community children in the region tested was 8.2% (95% CI 6.2—10.4%). Most individuals presented during the rainy season with a febrile disease refractory to standard treatment, sometimes mimicking tuberculosis. Some family clustering was apparent. All patients with bacteraemic melioidosis died, but treatment with the available conventional therapies of chloramphenicol, co-trimoxazole or doxycycline resulted in survival and cure in six patients with subacute/localised melioidosis. Further studies are needed to ascertain the local epidemiology and why children appear particularly at risk, as well as to establish the true extent of melioidosis in Papua New Guinea. © 2007 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved.
1. Introduction Melioidosis is an infection caused by the saprophytic Gramnegative bacillus Burkholderia pseudomallei. Melioidosis
∗
Corresponding author. Tel.: +61 7 4779 6375; fax: +61 7 4779 1526. E-mail address: [email protected] (J.M. Warner).
is emerging as a significant cause of community-acquired sepsis in the tropics, but the great variety of clinical presentations and difficulties with laboratory diagnosis make accurate determination of the distribution problematic (Cheng and Currie, 2005; White, 2003). Whilst there have been sporadic reports of melioidosis from Papua New Guinea, seroprevalence studies from the capital city of Port Moresby indicated that limited or no exposure to B. pseudomallei occurs in that region of the country (Barnes et al., 1991; Currie, 1993; De Buse et al., 1975; Kingston,
0035-9203/$ — see front matter © 2007 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.trstmh.2007.02.024
810 1971; Lee and Naraqi, 1980; Newland, 1969; Rowlands and Curtis, 1965). However, an unpublished report suggested that melioidosis may be endemic in the rural Balimo region of the Western Province of Papua New Guinea (Reece, 1984). We have therefore undertaken a prospective study to determine the extent of autochthonous melioidosis within this rural community.
2. Materials and methods Balimo is the rural district centre of the Gogodala language group who number approximately 30 000. Less than 10% live in the district township and associated villages, with most being subsistence gardeners, fisherman and hunters scattered along the Aramia River and associated flood plain (Figure 1). The study was conducted at Balimo Health Centre with the support of staff and the community. Burkholderia pseudomallei solid and liquid selective and differential media were prepared as described by Ashdown (1979) with 8 mg/l gentamicin. Organisms were presumptively identified as B. pseudomallei based on typical colonial appearance, oxidase reaction (positive) and gentamicin resistance inferred from vigorous growth on ASH agar (Dance et al., 1989). Identification of each isolate was corroborated with API 20NE (bioM´ erieux, Baulkham Hills, NSW, Australia) or Microbact 24E (MedVet, Adelaide, SA, Australia) and confirmed using two PCR protocols (Dharakul et al., 1999; Winstanley and Hart, 2000). The indirect haemagglutination assay (IHA) was used for serology as per the methods of Ashdown (1987). A ‘maybe melioidosis’ campaign was established at the health centre, which included increasing clinical awareness and establishing laboratory diagnostics. Consideration of melioidosis included any patient presenting with community-acquired pneumonia or sepsis and any patient with persistent fever refractory to standard treatment (typically antimalarials and short course tuberculosis treatment). These patients were investigated by serology plus selective culture of sputum and/or gastric aspirate and throat swabs. Blood cultures were taken on all patients with persistent fever and reactive IHA serology.
Figure 1 The rural Balimo district of Western Province, Papua New Guinea, showing the Aramia flood plain at the height of the wet season and the location of village communities, demonstrating their very close association with water.
J.M. Warner et al. Confirmed cases were those with a positive culture for B. pseudomallei; suspected melioidosis was defined as those with clinically compatible illness plus evidence of either direct familial association during a confirmed outbreak or IHA seroreactivity. Treatment regimens were based on recommendations at the time (Sookpranee et al., 1992), although limited availability of antibiotics necessitated modifications. Notably, ceftazidime, which is the drug of choice for melioidosis with sepsis, was not available. For those presenting with acute sepsis, initial treatment included a two-drug combination of i.v. chloramphenicol and/or doxycycline and/or co-trimoxazole for 3 weeks. In survivors, this was followed by oral therapy for 3 months with one or two of the above antibiotics. Those with fever without sepsis or with just localised disease were treated with 3 months of oral therapy with one or two of the above antibiotics. As rainfall in Balimo is not recorded, the average monthly rainfall data for 1994—1995 and 1998 at the provincial capital of Daru were compared with the date of diagnosis of each confirmed melioidosis case, as Daru rainfall generally reflects that seen in Balimo. Five regional community schools were chosen for a seroprevalence survey of children. The schools were all were located within a 10 km radius of the health centre and represented the majority of children of that age group in the immediate Balimo region. Sera were separated and stored at −20 ◦ C until assayed by IHA. Seroprevalence was calculated per school. An IHA titre of 1:40 was chosen as the cut-off value for a positive reactor, consistent with its use for diagnostic purposes. Regional differences in seroprevalence were compared using Fisher’s exact test with 95% CI.
3. Results Culture-confirmed cases and suspected cases of melioidosis as well as their outcomes are presented in Table 1. Case AW was described in the original 1984 report by Reece (1984). During 1994—1998, two periods of melioidosis screening were conducted, totalling 16 months of active screening. Unfortunately, detailed sampling data from the portion of the study conducted in 1994—1995 were lost, with only the results of the culture-confirmed cases remaining (Table 1). However, during the 1998 study (March—July) 245 samples from 170 patients were selectively cultured as per the protocols listed above. Accurate dates of birth are difficult to determine in rural Papua New Guinea, therefore patient age was classified as either ‘child’ or ‘adult’ based on physical maturity. However, when dates of birth were able to be determined through review of medical records (more likely in young children born at the health centre) age was recorded. Of the patients screened in the 1998 study, 104 (61%) were adults and 66 (39%) were children. The samples selectively cultured included 1 ascitic fluid, 1 eye swab, 2 ear swabs, 25 throat swabs, 37 blood cultures, 54 skin lesions and abscess pus and 125 sputa or gastric aspirates. Of the 170 patients sampled in 1998, 3 were culture-positive for B. pseudomallei (Table 1). During the total 16 months of the study, eight cases of culture-confirmed melioidosis autochthonous to the Balimo region were diagnosed. A further two cases of suspected
Melioidosis in a rural community of Western Province, PNG
811
Table 1
Culture-confirmed or suspected melioidosis cases documented in Balimo during 1983, 1994—1995 and 1998
Patient
Familya
Sex
Age (years)
Date
Initial presentation
Culture
IHA
Village
Outcome
1983 AW
*
M
18
Jan. 1983
Fever, respiratory symptoms
Blood
NT
Adiba
Died
1994—1995 KW — ID — SS * GD — DS ** KimS ** KawS **
M F M F F M M
10 9 6 9 14 7 24
Jan. 1994 June 1994 May 1994 Jan. 1995 Jan. 1995 Jan. 1995 Feb. 1995
Fever, respiratory symptoms Fever, groin abscess Persistent fever Fever, respiratory symptoms Fever Fever Fever
Blood Blood NT Sputum Blood NT Blood
NT NT 1024 >5120 NT 320 NT
Adiba Kini Adiba Adiba Kimama Kimama Kimama
Died Died Well Well Died Well Died
1998 NG TG RI
F M M
4 6 17
May 1998 May 1998 July 1998
Fever, respiratory symptoms Fever Fever
Sputum Throat Abscess
40 >5120 <5
Balimo Balimo Togowa
Well Well Well
*** *** —
IHA: indirect haemagglutination assay; NT: culture or serology not available or organism not recovered. a Patients with the same symbol (*, ** or ***) are members of same immediate family.
melioidosis were empirically diagnosed based on strong clinical suspicion, serology and response to melioidosis-directed therapy (Table 1). The median age of the culture-confirmed cases was 9.5 years (interquartile range 8.3—14.8 years). According to medical records, all individuals presented initially with subacute/localised disease diagnosed with or without respiratory symptoms and signs, although several developed severe pneumonia and died. In each case during their admission, before a diagnosis of melioidosis was made, short-course tuberculosis treatment was considered. The annual incidence of culture-confirmed melioidosis in the population was 20.0 per 100 000 population (95% CI 12.2—30.9). This figure is likely an underestimate as it assumes that the total population of the Gogodala language group had access to the Balimo Health Centre. The prevalence of culture-confirmed melioidosis within the patient cohort during March—July 1998 was 1.8% (95% CI 0.37—5.1%). All four bacteraemic melioidosis patients died, with a delay in diagnosis noted for each case. In contrast, all six individuals diagnosed with localised/subacute melioidosis recovered with the use of melioidosis-directed therapy. No recurrent infection has subsequently been reported in these survivors. Figure 2 demonstrates that melioidosis cases occurred during or immediately after the wet season between December and May. A total of 747 children aged 8—13 years participated in the serological study. The seroprevalence of each regional group is presented in Figure 3. The overall seroprevalence was 8.2% (95% CI 6.2—10.4%). The seroprevalences of Adiba and Buila schools were 15.2% and 16.9%, respectively, and these were significantly higher than the other three centres (Fisher’s exact test, P < 0.001).
Figure 2 Mean annual rainfall for 1994—1995 and 1998 in Daru, and number of melioidosis cases. Columns indicate monthly rainfall; points indicate number of melioidosis cases.
rural regions approaching 100 per 1000 births and an average life expectancy of 52 years (Papua New Guinea Ministry of Health, 2000). Pneumonia is the leading cause of death in each province. Signs and symptoms of melioidosis can be extremely variable. Whilst pneumonia is the commonest
4. Discussion Papua New Guineans experience the poorest of health outcomes in the Pacific region, with childhood mortality in
Figure 3 Seroprevalences in the Balimo district community school population using a cut-off indirect haemagglutination assay titre of ≥40. Error bars signify binomial 95% CI.
812 presentation of melioidosis overall, there can be many other manifestations, and diagnosis is dependent upon microbiology that may not be available in regions of possible endemicity (Cheng and Currie, 2005; White, 2003). In this study, improving the diagnostic capacity of a rural health centre has demonstrated autochthonous melioidosis in a rural community in Papua New Guinea for the first time. Furthermore, early diagnosis with the use of selective media enabled prompt directed therapy, resulting in those with localised disease making a full recovery. Although numbers are small, a feature of melioidosis in this region is the childhood predilection. Children were thought to acquire the organism through their robust behaviour within the environment, such as washing and playing in the lagoon and playing in the mud at the beginning of the wet season (all common practices). Clustering of cases within family groups reflects presumptive shared exposure of a common reservoir. Of note, a recent report from rural northeastern Brazil describes a similar newly recognised endemic focus of melioidosis where mostly children have been involved and where a local water source has been implicated (Rolim et al., 2005). Of the nine culture-confirmed cases documented, seven represent three outbreaks from three isolated regions, which also share the highest seroprevalence in their community schoolchild populations. This apparent spatial clustering of seroprevalence and clinical incidence may well reflect differences in the localised persistence of B. pseudomallei in the environment, as occurs in other melioidosis-endemic regions (Cheng and Currie, 2005). It is possible that in these locations the washing practices and environmental conditions may enable acquisition of B. pseudomallei via the respiratory or nasopharyngeal routes, either from water as an endogenous reservoir or from water contaminated with soil after times of heavy rain. Furthermore, anecdotal evidence from Adiba village suggested that the 1983 case and other undocumented suspected melioidosis cases were associated with soil excavation during house building. The accepted co-morbidity factors for acute-onset bacteraemic melioidosis, such as diabetes, liver disease and chronic renal disease (Currie et al., 2000; Suputtamongkol et al., 1999), are uncommon in Balimo. This, together with the heavy environmental exposure in children, may explain why melioidosis was less common in the adult population in this study. Melioidosis sepsis in adults may well become more common in this region if diseases such as type 2 diabetes become endemic, as has occurred in many periurban village communities in Port Moresby and in other Pacific nations as well as in the indigenous Australian population. The overall melioidosis seroprevalence of this district (6.2—10.4%) is similar to the seroprevalence of the Top End of the Northern Territory of Australia (5—13%). The Top End melioidosis annual incidence of between 16.5—34.5 per 100 000 (Currie et al., 2000) is also similar to the estimated annual incidence in the Balimo district of 20.0 per 100 000 population. Further clinical surveillance of this and other regions of Papua New Guinea is required to establish the true burden of melioidosis in the country. Given the extreme limitations on health resources in most of Papua New Guinea, it is possible that melioidosis is more widespread than currently recognised and may in some other locations be a major cause of community-acquired sepsis.
J.M. Warner et al. The relatedness of B. pseudomallei isolates from this study to other isolates from Papua New Guinea and also to those from the Torres Strait and mainland Queensland could provide insights into local melioidosis biogeography. Rainfall has been implicated as a clear factor that predisposes both the extent and intensity of melioidosis in endemic communities (Currie and Jacups, 2003). As in northern Australia and Thailand, the cases in the Balimo region were clustered in the rainy season. In Balimo, cases appeared particularly with the first rain after a prolonged dry season (Figure 2). The onset of the rainy season is when local health authorities should be on heightened alert. In conclusion, the Balimo region of Papua New Guinea is endemic for melioidosis, and melioidosis should be considered in the differential diagnosis of patients presenting with febrile illness from this region, particularly if the fever is refractory to standard treatment. Further studies are needed to ascertain the local epidemiology and why children appear particularly at risk and why family clustering occurs, as well as to establish the true extent of melioidosis in Papua New Guinea. Authors’ contributions: JMW, DBP and RGH designed the study protocol; JMW and DBP undertook the field work; BC provided clinical input; RGH analysed the data; JMW, RGH and BC drafted the manuscript. All authors read and approved the final manuscript. JMW is guarantor of the paper. Acknowledgements: The following are gratefully acknowledged: the staff of Balimo Health Centre for clinical support; Dr David Learoyd for help with clinical assessment and access to medical records during the 1994—1995 portion of the study and, as medical superintendent, the opportunity to conduct this study during the period of employment at Balimo in 1994 and 1995; village and school communities of Buila, Balimo, Adiba, Kimama and Saweta; Dr Kath Donovan for access to Dr Stuart Reece’s report; Joy Koehler and Bruce Samway for access to IHA antigen and help with the serology assay; the School of Public Health and Tropical Medicine, James Cook University, particularly Dr Wayne Melrose, Dr Paul Turner and Dr Jeffrey Hii, for laboratory consumables; and Mr Jimmy Gomoga, Deputy Director of Meteorology, for the rainfall data and advice on the suitability of using Daru rainfall data to reflect seasonal rainfall patterns at Balimo. Funding: None. Conflicts of interest: J.M. Warner was employed at Balimo Health Centre during 1994 and 1995. D.B. Pelowa was employed at Balimo Health Centre during the whole time of the study. J.M. Warner has family links to Papua New Guinea, but not to the Balimo region. D.B. Pelowa has family links to the Gogodala language group in and around Balimo. Although the authors have a conviction to improve the public health of rural Papua New Guineans, this is not considered a conflict of interest, simply a professional interest. Ethical approval: Papua New Guinea Medical Research and Advisory Council.
Melioidosis in a rural community of Western Province, PNG
References Ashdown, L.R., 1979. Identification of Pseudomonas pseudomallei in the clinical laboratory. J. Clin. Pathol. 32, 500—504. Ashdown, L.R., 1987. Indirect haemagglutination test for melioidosis. Med. J. Aust. 147, 364—365. Barnes, D.J., Gottlieb, T., Naraqi, S., Benn, R., 1991. The role of viruses and atypical organisms in the pathogenesis of adult pneumonia in Papua New Guinea. P. N. G. Med. J. 34, 13—16. Cheng, A.C., Currie, B.J., 2005. Melioidosis: epidemiology, pathophysiology, and management. Clin. Microbiol. Rev. 18, 383—416. Currie, B., 1993. Melioidosis in Papua New Guinea: is it less common than in tropical Australia? Trans. R. Soc. Trop. Med. Hyg. 87, 417. Currie, B., Jacups, S.P., 2003. Intensity of rainfall and severity of melioidosis, Australia. Emerg. Infect. Dis. 9, 1538—1542. Currie, B.J., Fisher, D.A., Howard, D.M., Burrow, J.N., Lo, D., SelvaNayagam, S., Anstey, N.M., Huffam, S.E., Snelling, P.L., Marks, P.J., Stephens, D.P., Lum, G.D., Jacups, S.P., Krause, V.L., 2000. Endemic melioidosis in tropical northern Australia: a 10-year prospective study and review of the literature. Clin. Infect. Dis. 31, 981—986. Dance, D.A., Wuthiekanun, V., Naigowit, P., White, N.J., 1989. Identification of Pseudomonas pseudomallei in clinical practice: use of simple screening tests and API 20NE. J. Clin. Pathol. 42, 645—648. De Buse, P.J., Henderson, A., White, M., 1975. Melioidosis in a child in Papua New Guinea successful treatment with kanamycin and trimethoprim—sulphamethoxazole. Med. J. Aust. 2, 476—478. Dharakul, T., Tassaneetrithep, B., Trakulsomboon, S., Songsivilai, S., 1999. Phylogenetic analysis of Ara+ and Ara− Burkholderia pseudomallei isolates and development of a multiplex PCR procedure for rapid discrimination between the two biotypes. J. Clin. Microbiol. 37, 1906—1912.
813 Kingston, C.W., 1971. Chronic or latent melioidosis. Med. J. Aust. 2, 618—621. Lee, L., Naraqi, S., 1980. Primary gram negative pneumonia in adults in Papua New Guinea. P. N. G. Med. J. 23, 174—178. Newland, R.C., 1969. Chronic melioidosis: a case in Sydney. Pathology 1, 149—152. Papua New Guinea Ministry of Health, 2000. National Health Plan 2001—2010, Policy Directions and Priorities. Papua New Guinea Ministry of Health, Port Moresby. Reece, A.S., 1984. A Clinical Report: the experience of the Balimo Health Centre with the clinical syndrome of melioidosis, including a suggested diagnostic, therapeutic and epidemiological program. Balimo Health Centre, Balimo. Rolim, D.B., Vilar, D.C., Sousa, A.Q., Miralles, I.S., de Oliveira, D.C., Harnett, G., O’Reilly, L., Howard, K., Sampson, I., Inglis, T.J., 2005. Melioidosis, northeastern Brazil. Emerg. Infect. Dis. 11, 1458—1460. Rowlands, J.B., Curtis, P.G., 1965. A case of melioidosis in Papua and New Guinea. Med. J. Aust. 2, 494—496. Sookpranee, M., Boonma, P., Susaengrat, W., Bhuripanyo, K., Punyagupta, S., 1992. Multicenter prospective randomized trial comparing ceftazidime plus co-trimoxazole with chloramphenicol plus doxycycline and co-trimoxazole for treatment of severe melioidosis. Antimicrob. Agents Chemother. 36, 158—162. Suputtamongkol, Y., Chaowagul, W., Chetchotisakd, P., Lertpatanasuwun, N., Intaranongpai, S., Ruchutrakool, T., Budhsarawong, D., Mootsikapun, P., Wuthiekanun, V., Teerawatasook, N., Lulitanond, A., 1999. Risk factors for melioidosis and bacteremic melioidosis. Clin. Infect. Dis. 29, 408—413. White, N.J., 2003. Melioidosis. Lancet 361, 1715—1722. Winstanley, C., Hart, C.A., 2000. Presence of type III secretion genes in Burkholderia pseudomallei correlates with Ara(−) phenotypes. J. Clin. Microbiol. 38, 883—885.