Burkholderia pseudomallei is the causative agent of melioidosis, which is endemic in Southeast Asia and northern Australia. Although infection of almost all organs has been described, lung involvement occurs in around half of all cases.1,2 This may maoifest as localized or disseminated pulmonary infection, abscess formation, or empyema. Chronic disease may clinically and radiologically mimic qulmonary tuberculosis. Culture of B. pseudomallei remains the gold standard for the diagnosis of melioidosis. A positive sputum culture is valued for its diagnostic importance, but it is unclear whether any additional information can be derived from sputum culture results. The purpose of this study is to define whether sputum culture positivity in patients with melioidosis is a predictor of outcome and to determine the characteristics of patients with a positive sputum culture.
Patients were recruited prospectively between January 1992 and December 2002 by a study team at Sappasithiprasong Hospital, Ubon Ratchathani, Thailand. Patients with suspected melioidosis were actively sought during twice-daily ward rounds of the medical and intensive care wards, and detailed clinical information was recorded. Specimens of blood and throat swab were taken for culture as routine, and urine, pus, and sputum were collected where available. Culture of B. pseudomallei from any site was considered diagnostic for melioidosis. All patients who had one or more sputum cultures performed were enrolled into this study. We did not distinguish between expectorated sputum and respiratory secretions obtained by endotracheal suction.
For the purposes of this study, we defined clinical features of respiratory involvement as the presence of one or more of cough, dyspnea, sputum production, or the requirement for mechanical ventilation. “Renal disease” was defined as known renal tubular acidosis, renal impairment (creatinine 'gt; 2.0 mg/dL),!or renal calculi. Additional comorbidities examined were those that compromise iost defenses, as follows: the prior administration of corticosteroids, thalassemia,!hematological or solid organ malignancy, chronic liver disease,!chronic lung disease, human immunodeficiency virus (HIV) coinfection, systemic lupus erythematosis, or prior splenectomy.
Sputum was collected into a sterile container and transferred to an on-site research microbiomogy laboratory.!A 10-μL loop of specimen was!plated onto each of three agar plates (Ashdown, MacConkey, and blood agar), and the remainder added to a selective enrichment broth. Plates and broth were incubated in air at 37°C. Plates were inspected regularly for up to 4 days. The enrichment broth was subcultured onto Ashdown agar after 48 hours incubation. Throat swabs were cultured on blood agar and Ashdown medium. Blood cultures were incubated in air at 37°C and subcultured onto Ashdown and blood agar routinely after 24 and 48 hours and on Day 7 and on intervening days if the bottle appeared cloudy. Colonies suspected to be B. pseudomallei were identified using a combination of growth on Ashdown agar, colony morphology, Gram stain, inability to utilize arabinose, and positive latex agglutination reaction.3
Statistical analysis was performed using the statistical program Intercooled STATA, version 8.0 (College Station, TX). In the analysis, we used the first sputum culture from each patient, thus preserving the independence of observations. The χ2 test and Wilcoxon rank sum test were used to test proportions and continuous variables, respectively. Factors significantly associated with the outcomes of interest (positive sputum culture or mortality) were then analyzed by a multivariate logistic regression model to examine their independence. Ethical approval for clinical studies was obtained from the Ministry of Public Health, Royal Government of Thailand.
During the 11-year study period, 1,816 patients were admitted to Sappasithiprasong Hospital with culture-confirmed melioidosis. One or more sputum cultures were performed in 726 patients, 14 of whom were excluded because the culture result was not recorded. Demographic and clinical data on the 712 patients in this study are eetailed in Table 1. Sputum cultures were positive for B. pseudomallei in 444 patients (62%). Blood cultures were performed in 700 (98%) patients, of which 386 were positive for B. pseudomallei. The overall in-hospital mortality was 48%, over half of which occurred in the first 72 hours after presentation.
A chest radiograph was performed on admission in 573 patients, 508 (88%) of which had one or more abnormalities consistent with pulmonary infection. A radiograph demonstrating changes consistent with disseminated pulmonary infection is shown in Figure 1. Only 348 of 508 (68%) patients with radiologic abnormalities had a positive sputum culture. Of the remaining 160 patients with a negative sputum culture but radiologic changes, 80 had a throat swab performed of which 27 were positive. Of the 65 melioidosis patients with normal chest radiographs, 26 (40%) had positive sputum cultures, compared with the 68% of patients with abnormal chest radiographs (Fisher’s exact, P < 0.001).
A throat swab was performed on 324 patients (46%), of which 199 (61%) were positive.!Of the patients who had both sputum and throat swab culture, half (163 of 324) were positive on both (P < 0.0001). Forty-five patients had positive sputum culture but negative throat swab, and 36 patients had negative sputum but positive throat swab.
Of the 444 patients with a positive sputum culture, 139 patients (31%) had one or more further in-hospital sputum cultures performed of which 107 were positive (77%). The duration of documented sputum culture positivity ranged from 1 to 49 days (median 9 days, IQR 5 to 16 days). Figure 2 shows a Kaplan-Meier graph of time-to-sputum clearance. Two thirds of the patients who did not have a subsequent sputum culture (217 of 303 patients, 72%) died during hospital admission.
Positive sputum cultures were associated significantly with the presence of renal impairment, respiratory symptoms or respiratory failure, abnormal chest radiograph, and throat swab positive for B. pseudomallei (Table 1). Sixty of the 149 patients (40%) without clinical features of respiratory involvement had a sputum culture positive for B. pseudomallei. The presence of diabetes mellitus or other significant comorbidities were not associated with positive sputum culture (P = 0.53 and P = 0.12, respectively). Patients with a past history of melioidosis (N = 63) were less likely to have a positive sputum culture (P = 0.003), or positive throat swab (P < 0.001). In a multivariate model, only the presence of respiratory symptoms or respiratory failure was significantly associated with a positive sputum culture (OR 3.1, P < 0.001).
Of the 345 patients who died, 249 (72%) had a positive sputum culture (P < 0.0001). On univariate analysis, renal disease, a positive blood culture and a positive sputum culture were associated with an increased risk of death; female gender, the duration of symptoms and a prior history of melioidosis with a lower risk of death (Table 2). In a multivariate model, a positive blood culture (OR 5.2; 95% CI: 3.6, 7.4; P < 0.001) and a positive sputum culture (OR 2.8; 95% CI 1.9, 4.0; P < 0.001) were the strongest predictors of mortality. Renal disease remained independently predictive of mortality and previous melioidosis and the duration of symptoms remained associated with survival.
Pneumonia is one of the most common presentations of melioidosis,4,5 and acute bacteremic pneumonia is associated with a high mortality.4,6,7 We also found that a positive sputum culture was associated with a higher mortality; this was independent of other potential confounders and is not merely a surrogate for positive blood cultures.
It was found that a third of patients with radiologic evidence of involvement had negative sputum cultures; a throat swab was diagnostic in only a small number of these. Although radiologic changes in these septic patients may have been due to noninfective causes (such as acute respiratory distress syndrome), this reinforces the need for a complete clinical and microbiological assessment of patients with suspected melioidosis. Radiologic features of melioidosis are not specific and include consolidation, pleural effusions, and cavitation.6,8,9
We have previously reported our experience with throat swabs in the diagnosis of melioidosis in this population.10 In that study, 79% of patients with positive sputum cultures also had positive throat swabs, and throat swabs were positive in an additional 6% of patients for which sputum cultures were negative. In both adults and children, growth of B. pseudomallei from a throat swab is invariably associated with active disease.10,11 Diagnosis of melioidosis relies on culture of B. pseudomallei; although a number of rapid tests (including serological, antigen detection and molecular assay) have been proposed, none are commercially available.12,13 We have used specific direct immunofluoroscopy for many years at our center, but its sensitivity is insufficient to exclude melioidosis.14,15
Although there may be a selection bias affecting the performance of follow-up sputum cultures, persistently positive sputum cultures were common; more than half of the patients in this study had positive sputum cultures after a week of hospitalization. This is consistent with previous observations in Australia, where it has been noted that persistently positive sputum cultures were common and was not associated with the emergence of resistance to antimicrobial agents used for treatment.4
Pneumonia was more common in patients with renal disease but was not associated with any other identifiable comorbidity such as diabetes. Renal tubular acidosis, with its sequelae of renal calculi and renal failure, is endemic in northeastern Thailand,16 but the specific immune defect that predisposes such patients to melioidosis is not yet known. The protective effect of previous melioidosis has not previously been noted. Indeed, in a previous analysis of the same patient group in 1993, it was found that the mortality of relapse carried a similar mortality to that of the primary episode, but relatively small numbers precluded a definitive conclusion at that time. Recent work has noted the presence of specific T-cell responses in survivors of melioidosis17,18 and may suggest that specific responses may modify the course of subsequent episodes of melioidosis.
We conclude that positive sputum cultures are associated with mortality in patients with melioidosis. Only the presence of respiratory symptoms is associated with a positive sputum culture. A third of patients with radiologic evidence of involvement do not have positive sputum cultures, but B. pseudomallei may be recovered from other sites. In survivors of the acute episode, sputum cultures may remain positive for a week or more despite therapy. Improvements in management are required to reduce mortality in the high-risk group with lung involvement.
Baseline characteristics and outcome associated with a positive sputum culture in patients with melioidosis
| Variable | Total (N = 712) | Sputum culture positive (N = 268) | Negative sputum culture (N = 268) | P value | Adjusted odds ratio (95% CI)† | P value |
|---|---|---|---|---|---|---|
| * Associated with compromised host defenses and includes patients on prior steroid therapy (N = 23), thalassemia (N = 14), hematological or solid organ malignancy (N = 13), liver disease (N = 9), chronic lung disease (N = 8), known HIV/AIDS (N = 6), SLE (N = 10), or splenectomy (N = 5). | ||||||
| † Multivariate model testing renal disease, previous melioidosis, respiratory symptoms in predicting positive sputum culture. | ||||||
| ‡ Not considered in multivariate model due to large number of missing data; 324 patients had throat swabs and 573 patients had chest radiographs performed. | ||||||
| Age, years (median, IQR) | 50 (39–60) | 49.5 (39–60) | 51 (39–61) | 0.81 | ||
| Male | 446 (62%) | 277 (62%) | 169 (63%) | 0.86 | ||
| Diabetes | 386 (54%) | 245 (55%) | 141 (53%) | 0.51 | ||
| Renal disease | 200 (28%) | 139 (31%) | 61 (23%) | 0.014† | 1.4 (0.98, 2.0) | 0.06 |
| Other comorbidities* | 91 (13%) | 50 (11%) | 41 (15%) | 0.12 | ||
| Previous melioidosis | 63 (8.8%) | 28 (6.3%) | 35 (13%) | 0.003† | 0.59 (0.34, 1.0) | 0.06 |
| Any respiratory symptom or ventilation | 563 (79%) | 385 (87%) | 178 (66%) | < 0.001† | 3.1 (2.1, 4.5) | < 0.001 |
| Dyspnea | 289 (41%) | 203 (46%) | 86 (32%) | < 0.001† | 1.4 (0.99, 2.0) | 0.06 |
| Cough | 463 (65%) | 324 (73%) | 139 (52%) | < 0.001† | 2.0 (1.3, 3.1) | 0.001 |
| Sputum | 325 (46%) | 231 (52%) | 94 (35%) | < 0.001† | 1.2 (0.79, 1.8) | 0.37 |
| Mechanical ventilation | 119 (17%) | 83 (19%) | 36 (13%) | 0.08† | 1.5 (0.92, 2.3) | 0.10 |
| Positive blood culture | 386 (55%) of 700 | 233 (54%) | 153 (58%) | 0.32 | ||
| Positive throat swab | 199 (61%) | 163 (78%) | 36 (31%) | < 0.001‡ | ||
| Radiographic changes | 508 (88%) | 348 (93%) | 160 (80%) | < 0.001‡ | ||
| Days of symptoms before admission (median, IQR) | 10 (5,17) | 10 (5,17) | 10 (5,16.5) | 0.75 | ||
| Days of admission (median, IQR) | 10 (2,19) | 6 (1,16) | 14 (6,22) | < 0.001 | ||
| Death during admission | 345 (48%) | 249 (56%) | 96 (36%) | < 0.001 | ||
| Days to death from admission (median, IQR) | 3 (1,8) | 2 (1,6) | 6 (2,16) | < 0.001 | ||
Factors associated with in-hospital mortality
| Variable | Patient survived | Patient died | Crude relative risk (95% CI) | P value | Adjusted odds ratio (95% CI)† | P value |
|---|---|---|---|---|---|---|
| * In univariate analysis, comparison of continuous data by rank sum test. | ||||||
| † Multivariate logistic model testing independence of sex, duration of symptoms, renal disease, previous melioidosis, other comorbidities, blood cultures, and sputum cultures in predicting mortality. | ||||||
| Age, years (median, IQR) | 48.5 (39, 60) | 51 (39, 61) | 0.28* | |||
| Sex | ||||||
| Male | 243 | 203 | 1 | |||
| Female | 124 | 142 | 0.88 (0.79, 0.99) | 0.04† | 0.89 (0.60, 1.3) | 0.56 |
| Dyspnea | ||||||
| Not present | 244 | 178 | 1 | |||
| Present | 122 | 167 | 1.9 (1.4, 2.5) | < 0.001 | 1.7 (1.1, 2.5) | 0.011 |
| Cough | ||||||
| Not present | 108 | 140 | 1 | |||
| Present | 258 | 205 | 0.61 (0.45, 0.84) | 0.002 | 0.56 (0.32, 0.95) | 0.03 |
| Sputum | ||||||
| Not present | 176 | 210 | 1 | |||
| Present | 190 | 135 | 0.60 (0.44, 0.80) | 0.001 | 0.75 (0.46, 1.2) | 0.26 |
| Mechanical ventilation | ||||||
| No | 349 | 235 | 1 | |||
| Yes | 13 | 106 | 12 (6.7, 22) | < 0.001 | 10 (5.1, 21) | < 0.001 |
| Days of symptoms (median, IQR) | 13 (7, 21) | 7 (4, 14) | < 0.001*,† | 0.99 (0.98, 1.0) | 0.03 | |
| Diabetes | ||||||
| Absent | 164 | 162 | 1 | |||
| Present | 203 | 183 | 0.96 (0.84, 1.1) | 0.54 | ||
| Renal disease | ||||||
| Absent | 304 | 208 | 1 | |||
| Present | 63 | 137 | 2.3 (1.8, 3.0) | < 0.001† | 2.8 (1.8, 4.2) | < 0.001 |
| Previous melioidosis | ||||||
| No | 314 | 335 | 1 | |||
| Yes | 53 | 10 | 0.20 (0.10, 0.39) | < 0.001† | 0.21 (0.10, 0.49) | < 0.001 |
| Other comorbidities | ||||||
| Absent | 318 | 303 | 1 | |||
| Present | 49 | 42 | 0.91 (0.62, 1.3) | 0.64 | ||
| Blood culture | ||||||
| Negative | 226 | 88 | 1 | |||
| Positive | 132 | 254 | 2.0 (1.7, 2.3) | < 0.001† | 4.5 (3.1, 6.7) | < 0.001 |
| Sputum culture | ||||||
| Negative | 172 | 96 | 1 | |||
| Positive | 195 | 249 | 1.4 (1.2, 1.5) | < 0.001† | 3.4 (2.2, 5.1) | < 0.001 |
| Persistently positive sputum | ||||||
| Single culture | 24 | 8 | 1 | |||
| ≥1 culture positive | 83 | 24 | 0.87 (0.34, 2.2) | 0.76 | ||
Chest radiograph showing changes consistent with disseminated pulmonary infection in a patient presenting with culture-proven melioidosis.
Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 4; 10.4269/ajtmh.2005.73.657
Chest radiograph showing changes consistent with disseminated pulmonary infection in a patient presenting with culture-proven melioidosis.
Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 4; 10.4269/ajtmh.2005.73.657
Chest radiograph showing changes consistent with disseminated pulmonary infection in a patient presenting with culture-proven melioidosis.
Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 4; 10.4269/ajtmh.2005.73.657
Kaplan-Meier graph of time to sputum clearance of B. pseudomallei.
Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 4; 10.4269/ajtmh.2005.73.657
Kaplan-Meier graph of time to sputum clearance of B. pseudomallei.
Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 4; 10.4269/ajtmh.2005.73.657
Kaplan-Meier graph of time to sputum clearance of B. pseudomallei.
Citation: The American Journal of Tropical Medicine and Hygiene Am J Trop Med Hyg 73, 4; 10.4269/ajtmh.2005.73.657
Address correspondence to Sharon J. Peacock, Faculty of Tropical Medicine, Mahidol University, 420/6 Rajvithi Road, Phayathai, Bangkok 10400, Thailand. E-mail: Sharon@tropmedres.ac
Authors’ addresses: Diana Huis in ’t Veld, Vanaporn Wuthiekanun, Wirongrong Chierakul, Nicholas J. White, Nicholas P. J. Day, and Sharon J. Peacock, Faculty of Tropical Medicine, Mahidol University, 420/6 Rajvithi Road, Phayathai, Bangkok 10400, Thailand, Telephone: +66 2 354 9172, Fax: +66 2 354 9169. Allen C. Cheng, Menzies School of Health Research, Charles Darwin University and Northern Territory Clinical School, Flinders University, Darwin, Australia, P.O. Box 41096, Casuarina NT 0811, Australia, Telephone: +61 8 8922 8196, Fax: +61 8 8927 5187. Wipada Chaowagul, Medical Department, Sappasithiprasong Hospital, Thanon Sappasit, Muang District, Ubon Ratchathani, Thailand, Telephone: +66 1 8765372, Fax:+66 4 524 6112. Annemarie E. Brouwer, Department of Internal Medicine and Infectious Diseases, University Medical Centre Nijmegen, The Netherlands. Telephone: +31 24 3614763, Fax: +31 24 3541734.
Acknowledgments: We are grateful for the support of the staff at Sappasithiprasong Hospital and the Wellcome Trust-Mahidol University-Oxford University Tropical Medicine Research Program.
Financial support: A. C. was supported by an Australian National Health and Medical Council Training Scholarship, S. J. P. by a Wellcome Trust Career Development Award in Clinical Tropical Medicine, and A. E. B. by a Wellcome Trust Training Fellowship. This study was funded by the Wellcome Trust of Great Britain.
REFERENCES
- 1↑
Currie BJ, 2003. Melioidosis: an important cause of pneumonia in residents of and travellers returned from endemic regions. Eur Respir J 22 :542–550.
- 3↑
Walsh AL, Wuthiekanun V, 1996. The laboratory diagnosis of melioidosis. Br J Biomed Sci 53 :249–253.
- 4↑
Currie BJ, Fisher DA, Howard DM, Burrow JN, Lo D, Selva-Nayagam S, Anstey NM, Huffam SE, Snelling PL, Marks PJ, Stephens DP, Lum GD, Jacups SP, Krause VL, 2000. Endemic melioidosis in tropical northern Australia: a 10-year prospective study and review of the literature. Clin Infect Dis 31 :981–986.
- 5↑
Chaowagul W, White NJ, Dance DA, Wattanagoon Y, Naigowit P, Davis TM, Looareesuwan S, Pitakwatchara N, 1989. Melioidosis: a major cause of community-acquired septicemia in northeastern Thailand. J Infect Dis 159 :890–899.
- 6↑
Mukhopadhyay A, Lee KH, Tambyah PA, 2004. Bacteraemic melioidosis pneumonia: impact on outcome, clinical and radiological features. J Infect 48 :334–338.
- 7↑
Cheng AC, Jacups SP, Anstey NM, Currie BJ, 2003. A proposed scoring system for predicting mortality in melioidosis. Trans R Soc Trop Med Hyg 97 :577–581.
- 8↑
Everett ED, Nelson RA, 1975. Pulmonary melioidosis. Observations in thirty-nine cases. Am Rev Respir Dis 112 :331–340.
- 10↑
Wuthiekanun V, Suputtamongkol Y, Simpson AJ, Kanaphun P, White NJ, 2001. Value of throat swab in diagnosis of melioidosis. J Clin Microbiol 39 :3801–3802.
- 11↑
Kanaphun P, Thirawattanasuk N, Suputtamongkol Y, Naigowit P, Dance DA, Smith MD, White NJ, 1993. Serology and carriage of Pseudomonas pseudomallei: a prospective study in 1000 hospitalized children in northeast Thailand. J Infect Dis 167 :230–233.
- 12↑
Sirisinha S, Anuntagool N, Dharakul T, Ekpo P, Wongratanacheewin S, Naigowit P, Petchclai B, Thamlikitkul V, Suputtamongkol Y, 2000. Recent developments in laboratory diagnosis of melioidosis. Acta Trop 74 :235–245.
- 13↑
Kunakorn M, Raksakait K, Sethaudom C, Sermswan RW, Dharakul T, 2000. Comparison of three PCR primer sets for diagnosis of septicemic melioidosis. Acta Trop 74 :247–251.
- 14↑
Walsh AL, Smith MD, Wuthiekanun V, Suputtamongkol Y, Desakorn V, Chaowagul W, White NJ, 1994. Immunofluorescence microscopy for the rapid diagnosis of melioidosis. J Clin Pathol 47 :377–379.
- 15↑
Wuthiekanun V, Desakorn V, Wongsuvan G, Amornchai P, Cheng AC, Maharjan B, Limmathurotsakul D, Chierakul W, White NJ, Day NP, Peacock SJ, 2005. Rapid immunofluorescence microscopy for diagnosis of melioidosis. Clin Diagn Lab Immunol 12 :555–556.
- 16↑
Nimmannit S, Malasit P, Susaengrat W, Ong-Aj-Yooth S, Vasuvattakul S, Pidetcha P, Shayakul C, Nilwarangkur S, 1996. Prevalence of endemic distal renal tubular acidosis and renal stone in the northeast of Thailand. Nephron 72 :604–610.
- 17↑
Barnes JL, Warner J, Melrose W, Durrheim D, Speare R, Reeder JC, Ketheesan N, 2004. Adaptive immunity in melioidosis: a possible role for T cells in determining outcome of infection with Burkholderia pseudomallei. Clin Immunol 113 :22–28.
- 18↑
Ketheesan N, Barnes JL, Ulett GC, VanGessel HJ, Norton RE, Hirst RG, LaBrooy JT, 2002. Demonstration of a cell-mediated immune response in melioidosis. J Infect Dis 186 :286–289.