Smear-negative pulmonary TB (SNPT) represents 30–60% of all pulmonary TB cases. The mortality of these patients can reach 25% in populations with high prevalence of HIV infection, and 10–20% of TB transmission at the population level are attributable to SNPT cases.
We conducted a retrospective study to evaluate epidemiological, clinical, and radiological characteristics of patients with SNPT and to compare these with patients who were diagnosed as having smear-positive pulmonary TB (SPPT). All adult patients (≥ 18 years old) with a positive culture for Mycobacterium tuberculosis, and a diagnosis of pulmonary TB were included in the study.
198 patients met the inclusion criteria (positive culture for Mycobacterium tuberculosis) and were included in the analysis. Of these patients, 69 (34.8%) were smear positive (SPPT) and 129 (65.2%) were smear negative (SNPT). In univariate analysis, cough, dyspnea, and hemoptysis were less frequent in SNPT patients in comparison with SPPT patients. In a multivariate model, having no cough and no radiographic pattern typical of TB were the characteristics independently associated with a diagnosis of SNPT.
Citation: Campos LC, Rocha MVV, Willers DMC, Silva DR (2016) Characteristics of Patients with Smear-Negative Pulmonary Tuberculosis (TB) in a Region with High TB and HIV Prevalence. PLoS ONE 11(1): e0147933. https://doi.org/10.1371/journal.pone.0147933
Editor: Claire Thorne, UCL Institute of Child Health, University College London, UNITED KINGDOM
Received: September 2, 2015; Accepted: January 11, 2016; Published: January 25, 2016
Copyright: © 2016 Campos et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All files are available from the Dryad database (accession number:10.5061/dryad.fh8hs).
Funding: This work was supported by FIPE-HCPA (Fundo de Incentivo à Pesquisa – Hospital de Clínicas de Porto Alegre).
Competing interests: The authors have declared that no competing interests exist.
Pulmonary tuberculosis (TB) is still one of the most common infectious diseases in the world. It is the second leading cause of death from an infectious disease worldwide, after the human immunodeficiency virus (HIV) . Brazil is ranked 16th among the 22 high-burden countries that collectively account for 80% of TB cases globally, with an incidence of 33.5 cases/100,000 inhabitants/year in 2014. The city of Porto Alegre has the highest incidence of TB in the country (99.3 cases/100,000 inhabitants/year in 2014), and 28.0% of TB patients are coinfected with HIV .
For the diagnosis of TB, the World Health Organization (WHO) recommends initial smear microscopy for the detection of acid-fast bacilli. However, this is not a sensitive technique, and only 57% of new cases of pulmonary TB reported were smear-positive . Sputum culture for mycobacteria has higher diagnostic yield in comparison with smear microscopy. Nevertheless, culture is more time-consuming in its results—the turn-around time is about 2–8 weeks—so it is less useful to guide the clinical decision-making process . Thus the treatment is often carried out empirically, based on clinical criteria, and can result in unnecessary costs and toxicities.
Smear-negative pulmonary TB (SNPT) represents 30–60% of all pulmonary TB cases, according to region . In Brazil, 24–30% of cases of pulmonary TB in adults are SNPT . The mortality of these patients can reach 25% in populations with high prevalence of HIV infection, which may be largely related to delay in diagnosis [6,7]. Furthermore, although smear-positive patients are considered to be more infectious, 10–20% of TB transmission at the population level are attributable to SNPT cases [8–10].
As much as possible, patients should be correctly diagnosed for the early initiation of treatment for SNPT. On the other hand, the treatment of those without the disease should be avoided. In the absence of rapid, simple, and accurate TB diagnostic tools for SNPT, frequently occurring clinical features are significant in supporting the diagnosis of SNPT, especially in areas with high TB and HIV infection. The aim of this study was therefore to identify the prevalence and the epidemiological characteristics, clinical presentation, and radiological findings of patients with SNPT and to compare these with patients who were diagnosed as having smear-positive pulmonary TB (SPPT).
Study Design and Location
We conducted a retrospective study to evaluate epidemiological, clinical, and radiological characteristics of patients with SNPT and to compare these with patients diagnosed with SPPT. The study location was Hospital de Clínicas de Porto Alegre (HCPA), a general, tertiary care, university-affiliated hospital with 750 beds and approximately 29,000 hospitalizations/year. It is a reference center for HIV in south of Brazil. The ethics committee at Hospital de Clínicas de Porto Alegre has approved access to patient records on January 22nd, 2013 (number 13–0024). Patient confidentiality has been maintained.
Patients and Data Collection
The records of all adult patients (≥ 18 years old) with a positive culture for Mycobacterium tuberculosis during the study period (January 2013—June 2015) were analyzed. We included only patients with a diagnosis of pulmonary TB, according to the Brazilian Guidelines for Tuberculosis , and a positive culture in spontaneous sputum, induced sputum, or bronchoalveolar lavage. Patients with extrapulmonary TB were excluded from the study. Patients with a positive culture and a negative smear microscopy were classified as SNPT, and patients with a positive culture and a positive smear microscopy were classified as SPPT .
The following data were collected from patient records using a standardized data extraction tool: demographic data (sex, age, ethnicity, years of schooling), behavioral data (smoking status, alcohol abuse, injection drug use), and medical history (clinical form of TB, symptoms at admission, diagnostic methods, presence of comorbidities, prior TB treatment, drug regimen, interval from hospital admission until diagnosis, length of hospital stay, intensive care unit [ICU] admission, duration of mechanical ventilation, and hospitalization outcome [death or discharge]). A current smoker was defined as reporting smoking at least 100 cigarettes in their lifetime, and at the time of the survey were smoking at least one day a week. A former smoker was defined as reporting smoking at least 100 cigarettes in their lifetime but who, at the time of the survey, did not smoke at all. Never smoked reported having smoked < 100 cigarettes in their lifetime . Alcohol abuse was defined as daily consumption of at least 30 grams (equivalent to a pint and a half of 4% beer) for men and 24 grams (equivalent to a 175 ml glass of wine) for women. An independent physician analyzed the chest X-rays (CXRs) and classified them as typical of TB (the presence of nodular, alveolar, or interstitial infiltrates predominantly affecting the zones above the clavicles or upper zones; the presence of cavitation affecting the upper zones or the apical segment of the lower lobe) or compatible with TB (enlarged hilar nodes, pneumonic lesion, atelectasis, mass lesion, miliary), according to previously described guidelines .
Data analysis was performed using SPSS 18.0 (Statistical Package for the Social Sciences, Chicago, Illinois). Data were presented as number of cases, mean ± standard deviation (SD), or median with interquartile range (IQR). Categorical comparisons were performed by chi-square test using Yates’s correction if indicated or by Fisher’s exact test. Continuous variables were compared using the t-test or Wilcoxon test. Multivariate logistic regression analysis was used to evaluate factors associated with a diagnosis of SNPT, using selection of factors associated (p ≤ 0.20) with this diagnosis in univariate analysis or those known to have clinical significance. Stepwise regression models were also examined to evaluate the possible collinearity among the predictors. The predictors selected in the final model were based on both statistical and clinical significance. The goodness of fit of the multiple logistic regression models was assessed using the Hosmer-Lemeshow test. Odds ratios (ORs) and nominal 95% confidence intervals (CIs) were presented. A two-sided p value < 0.05 was considered significant for all analyses. Sample size calculation was based on the prevalence of SNPT in Brazil (24–30%) . Considering the amplitude of the confidence interval of 0.15 and a 95% confidence level, we estimated a sample size of 128 patients.
During the study period, 198 patients met the inclusion criteria [positive culture for Mycobacterium tuberculosis in spontaneous sputum (n = 49), induced sputum (n = 54), or bronchoalveolar lavage (n = 95)] and were included in the analysis. Out of these patients, 69 (34.8%) were smear positive (SPPT) and 129 (65.2%) were smear negative (SNPT). The characteristics of the study population are shown in Table 1.
The study subjects had a mean age of 46.6 ± 17.1 years in SNPT group and 46.0 ± 16.9 years in SPPT group (p = 0.833). There were no statistically significant differences between study groups according to gender, ethnicity, smoking, alcohol abuse, and drug use. Among SNPT there were 53 (41.1%) HIV positive patients, and among SPPT there were 25 (36.2%) HIV positive patients (p = 0.608).
Cough was the most common symptom in both groups, although it was more frequent in SPPT (54 [78.3%]) than in SNPT (71 [55.0%]) (p = 0.002). Dyspnea and hemoptysis were also more frequent in patients with SPPT (32 [46.4%] and 9 [13.0%], respectively) than in patients with SNPT (35 [27.1%] and 4 [3.1%], respectively) (p = 0.010 and p = 0.013, respectively). The radiographic patterns typical of TB were the most common in both study groups. Patients with SNPT showed a pattern typical of TB less frequently (75 [58.1%]) in comparison with patients with SPPT (50 [72.5%]). However, this difference was not statistically significant in univariate analysis (p = 0.066). As expected, the median length of hospital stay was higher in SNPT patients (20 days [10–49 days]) than in SPPT patients (14.5 days [7–26 days]) (p = 0.010).
Logistic regression analysis estimating the ORs of factors associated with a diagnosis of SNPT was conducted. Age, male sex, former smoker, cough, weight loss, dyspnea, hemoptysis, radiographic pattern typical of TB and compatible with TB were the variables included in the multivariate model. The Hosmer-Lemeshow test showed that the multivariate model demonstrated a good fit (p = 0.684), and there was no evidence for collinearity among the independent variables. The following variables remained significant in the final multivariate model: no coughing (OR 3.10, 95%CI 1.55–6.19, p = 0.001) and no radiographic pattern typical of TB (OR 1.95, 95%CI 1.02–3.76, p = 0.045) (Table 2).
In this retrospective study, we demonstrated a high prevalence of SNPT (65.2%) among patients with TB in a tertiary care hospital in a region with high TB and HIV prevalence. In univariate analysis, cough, dyspnea, and hemoptysis were less frequent in SNPT patients in comparison with SPPT patients. In addition, the median length of hospital stay was higher in SNPT patients than in SPPT patients. In a multivariate model, having no cough and no radiographic pattern typical of TB were the characteristics associated with a diagnosis of SNPT.
SNPT is a relevant clinical and public health problem [6,13,14], especially in high HIV prevalence countries, where the number of patients with SNPT is increasing rapidly . We found a very high prevalence of SNPT in our study. Previous investigations showed that the proportion of cases of SNPT ranged from 10% to 61% of all pulmonary TB cases [8,16–19]. Many referral hospitals are confronted with relatively more SNPT cases than the lower level health facilities , and this is an important problem in daily medical practice in hospitals. Treating a patient inappropriately may cause side effects such as drug-induced hepatitis . On the other hand, prompt initiation of treatment is vital considering that SNPT patients have worse outcomes and excessive early mortality compared with SPPT patients [22–26]. Therefore, in order to avoid harmful delays in the establishment of a diagnosis, most cases of SNPT are diagnosed on the basis of the clinical presentation and radiological findings .
Regarding signs and symptoms, cough, dyspnea, and hemoptysis were less frequent in SNPT patients in comparison with SPPT patients in our study in univariate analysis. In a multivariate analysis, having no cough remained independently associated with a diagnosis of SNPT. SNPT patients are especially likely to show no or mild respiratory symptoms and systemic manifestations . In a retrospective study  based in a chart review of 146 HIV-infected patients, SNPT patients were significantly less likely to present with dyspnea, similarly to our study. Hemoptysis is also less common in these patients and some studies have reported a decreased proportion of patients with cough among SNPT patients [29,30] as we have demonstrated. In a study developed to validate a clinical-radiological score to assess the probability of pulmonary TB in patients suspected of SNPT, productive cough was less frequent in SNPT patients . In another study conducted in Brazil, 551 patients with clinical-radiological suspicion of SNPT were enrolled, and the presence of spontaneous sputum was negatively associated with SNPT .
In relation to radiological signs, we found in a multivariate model that having no radiographic pattern typical of TB was statistically associated with a diagnosis of SNPT. Atypical chest x-ray patterns or even normal findings were more frequent in SNPT cases as demonstrated previously by other authors [32,33]. In a study  that included all culture-confirmed, pulmonary TB cases reported to the US National TB Surveillance System from 1993 to 2008, SNPT cases were more likely to present atypical radiographic findings. Typical radiological patterns of pulmonary TB reactivation are less often observed in SNPT patients due to the smaller burden of mycobacteria , especially in areas with high prevalence of HIV infection. In these settings, clinicians are less likely to find typical radiological signs of pulmonary TB, and the diagnosis of TB is therefore more challenging .
Although approximately 40% of our study sample is HIV positive, no statistically significant association was observed between SNPT and HIV infection. In HIV high-prevalence countries, the increase of SNPT is thought to contribute significantly to delays in TB diagnosis and treatment [34–36]. Additionally, SNPT was originally thought to have a better prognosis compared with SPPT, however the HIV epidemic has led to changes in the presentation of SNPT disease from a slowly progressive disease to one with high mortality rate [37,38]. Previous studies demonstrated that HIV positive patients are more likely to have SNPT than HIV negative ones [6,17,23,39], and this probability increases as immunosuppression increases . Nonetheless, other authors [14,31] have also found no association between HIV infection and SNPT. One of these studies  was also conducted in Brazil, including patients with clinical-radiological suspicion of SNPT to develop a prediction model for SNPT for outpatients. HIV infections or AIDS were not among the variables significantly associated with a diagnosis of SNPT.
One of the limitations of this study is that we recruited patients from a single hospital, and the results may thus not apply to other settings. Furthermore, the information obtained retrospectively from chart review may be less complete than data collected prospectively. Also, it would be interesting to build a predictive model and evaluate the utility of the identified clinical features. In addition, we could not demonstrate excess early mortality in SNPT vs SPPT because our sample size was not calculated to identify this difference. Probably a larger sample would be needed to demonstrate it. However, the study was not designed with this purpose. In spite of these concerns, the knowledge of the epidemiological characteristics, clinical presentation, and radiological findings of patients with SNPT is important to define the characteristics that could be used to identify individuals with confirmed SNPT.
In conclusion, we found a very high prevalence of SNPT among patients with TB in a setting with high TB and HIV prevalence. The absence of cough in the presence of other symptoms suggestive of TB, and having no radiographic pattern typical of TB where independent predictors of SNPT. Future prospective studies are warranted in this setting to create clinical scoring algorithms to help to identify and correctly to treat patients with SNPT.
We would like to acknowledge the support from the International Clinical Operational Health Services Research Training Award (ICOHRTA/ Fogarty International Center/National Institutes for Health -NIH) and Johns Hopkins University (Johns Hopkins Bloomberg School of Public Health).
Conceived and designed the experiments: LCC MVVR DMCW DRS. Performed the experiments: LCC MVVR DMCW. Analyzed the data: DRS. Wrote the paper: LCC MVVR DMCW DRS.
- 1. World Health Organization. Global tuberculosis control: WHO report 2014. Available: www.who.int.
- 2. Secretaria de Vigilância em Saúde—Ministério da Saúde. Boletim Epidemiológico 2015. Available: www.saude.gov.br
- 3. Conde MB, Melo FA, Marques AM, Cardoso NC, Pinheiro VG, Dalcin PT, et al. III Brazilian Thoracic Association Guidelines on tuberculosis. J Bras Pneumol 2009, 35: 1018–1048. pmid:19918635
- 4. World Health Organization. Improving the diagnosis and treatment of smear-negative pulmonary and extrapulmonary tuberculosis among adults and adolescents. Recommendations for HIV prevalent and resource-constrained settings. 2007. Available: www.who.int
- 5. Ministério da Saúde. Programa Nacional de Controle da Tuberculose 2011. 2011. Available: http://portal.saude.gov.br/portal/arquivos/pdf/2site_31_05_2011.pdf.
- 6. Getahun H, Harrington M, O'Brien R, Nunn P. Diagnosis of smear-negative pulmonary tuberculosis in people with HIV infection or AIDS in resource-constrained settings: informing urgent policy changes. Lancet 2007, 369: 2042–2049. pmid:17574096
- 7. Macpherson P, Dimairo M, Bandason T, Zezai A, Munyati SS, Butterworth AE, et al. Risk factors for mortality in smear-negative tuberculosis suspects: a cohort study in Harare, Zimbabwe. Int J Tuberc Lung Dis 2011, 15: 1390–1396. pmid:22283900
- 8. Behr MA, Warren SA, Salamon H, Hopewell PC, Ponce de LA, Daley CL, et al. Transmission of Mycobacterium tuberculosis from patients smear-negative for acid-fast bacilli. Lancet 1999, 353: 444–449. pmid:9989714
- 9. Hernandez-Garduno E, Cook V, Kunimoto D, Elwood RK, Black WA, FitzGerald JM. Transmission of tuberculosis from smear negative patients: a molecular epidemiology study. Thorax 2004, 59: 286–290. pmid:15047946
- 10. Tostmann A, Kik SV, Kalisvaart NA, Sebek MM, Verver S, Boeree MJ, et al. Tuberculosis transmission by patients with smear-negative pulmonary tuberculosis in a large cohort in the Netherlands. Clin Infect Dis 2008, 47: 1135–1142. pmid:18823268
- 11. World Health Organization. Guidelines for controlling and monitoring the tobacco epidemic: World Health Organization; 1998. Available: www.who.int.
- 12. Diagnostic Standards and Classification of Tuberculosis in Adults and Children. This official statement of the American Thoracic Society and the Centers for Disease Control and Prevention was adopted by the ATS Board of Directors, July 1999. This statement was endorsed by the Council of the Infectious Disease Society of America, September 1999. Am J Respir Crit Care Med 2000, 161: 1376–1395. pmid:10764337
- 13. Harries AD, Maher D, Nunn P. An approach to the problems of diagnosing and treating adult smear-negative pulmonary tuberculosis in high-HIV-prevalence settings in sub-Saharan Africa. Bull World Health Organ 1998, 76: 651–662. pmid:10191561
- 14. Mello FC, Bastos LG, Soares SL, Rezende VM, Conde MB, Chaisson RE, et al. Predicting smear negative pulmonary tuberculosis with classification trees and logistic regression: a cross-sectional study. BMC Public Health 2006, 6: 43. pmid:16504086
- 15. Foulds J, O'Brien R. New tools for the diagnosis of tuberculosis: the perspective of developing countries. Int J Tuberc Lung Dis 1998, 2: 778–783. pmid:9783521
- 16. Affolabi D, Akpona R, Odoun M, Alidjinou K, Wachinou P, Anagonou S, et al. Smear-negative, culture-positive pulmonary tuberculosis among patients with chronic cough in Cotonou, Benin. Int J Tuberc Lung Dis 2011, 15: 67–70. pmid:21276299
- 17. Elliott AM, Namaambo K, Allen BW, Luo N, Hayes RJ, Pobee JO, et al. Negative sputum smear results in HIV-positive patients with pulmonary tuberculosis in Lusaka, Zambia. Tuber Lung Dis 1993, 74: 191–194. pmid:8369514
- 18. Harries AD, Nyangulu DS, Kangombe C, Ndalama D, Wirima JJ, Salaniponi FM, et al. The scourge of HIV-related tuberculosis: a cohort study in a district general hospital in Malawi. Ann Trop Med Parasitol 1997, 91: 771–776. pmid:9625933
- 19. Long R, Scalcini M, Manfreda J, Jean-Baptiste M, Hershfield E. The impact of HIV on the usefulness of sputum smears for the diagnosis of tuberculosis. Am J Public Health 1991, 81: 1326–1328. pmid:1928536
- 20. Basinga P, Moreira J, Bisoffi Z, Bisig B, Van den Ende J. Why are clinicians reluctant to treat smear-negative tuberculosis? An inquiry about treatment thresholds in Rwanda. Med Decis Making 2007, 27: 53–60. pmid:17237453
- 21. Loh LC, Abdul Samah SZ, Zainudin A, Wong GL, Gan WH, Yusuf WS, et al. Pulmonary disease empirically treated as tuberculosis-a retrospective study of 107 cases. Med J Malaysia 2005, 60: 62–70. pmid:16250282
- 22. Hargreaves NJ, Kadzakumanja O, Phiri S, Nyangulu DS, Salaniponi FM, Harries AD, et al. What causes smear-negative pulmonary tuberculosis in Malawi, an area of high HIV seroprevalence? Int J Tuberc Lung Dis 2001, 5: 113–122. pmid:11258504
- 23. Harries AD, Nyirenda TE, Banerjee A, Boeree MJ, Salaniponi FM. Treatment outcome of patients with smear-negative and smear-positive pulmonary tuberculosis in the National Tuberculosis Control Programme, Malawi. Trans R Soc Trop Med Hyg 1999, 93: 443–446. pmid:10674100
- 24. Harries AD, Dye C. Tuberculosis. Ann Trop Med Parasitol 2006, 100: 415–431. pmid:16899146
- 25. Henegar C, Behets F, Vanden Driessche K, Tabala M, Bahati E, Bola V, et al. Mortality among tuberculosis patients in the Democratic Republic of Congo. Int J Tuberc Lung Dis 2012, 16: 1199–1204. pmid:22871326
- 26. Raviglione MC, Harries AD, Msiska R, Wilkinson D, Nunn P. Tuberculosis and HIV: current status in Africa. AIDS 1997, 11 Suppl B: S115–S123. pmid:9416373
- 27. Tozkoparan E, Deniz O, Ciftci F, Bozkanat E, Bicak M, Mutlu H, et al. The roles of HRCT and clinical parameters in assessing activity of suspected smear negative pulmonary tuberculosis. Arch Med Res 2005, 36: 166–170. pmid:15847951
- 28. Palmieri F, Girardi E, Pellicelli AM, Rianda A, Bordi E, Rizzi EB, et al. Pulmonary tuberculosis in HIV-infected patients presenting with normal chest radiograph and negative sputum smear. Infection 2002, 30: 68–74. pmid:12018472
- 29. Kassu A, Mengistu G, Ayele B, Diro E, Mekonnen F, Ketema D, et al. Coinfection and clinical manifestations of tuberculosis in human immunodeficiency virus-infected and -uninfected adults at a teaching hospital, northwest Ethiopia. J Microbiol Immunol Infect 2007, 40: 116–122. pmid:17446959
- 30. Selwyn PA, Pumerantz AS, Durante A, Alcabes PG, Gourevitch MN, Boiselle PM, et al. Clinical predictors of Pneumocystis carinii pneumonia, bacterial pneumonia and tuberculosis in HIV-infected patients. AIDS 1998, 12: 885–893. pmid:9631142
- 31. Soto A, Solari L, Diaz J, Mantilla A, Matthys F, van der Stuyft P. Validation of a clinical-radiographic score to assess the probability of pulmonary tuberculosis in suspect patients with negative sputum smears. PLoS One 2011, 6: e18486. pmid:21483690
- 32. Kanaya AM, Glidden DV, Chambers HF. Identifying pulmonary tuberculosis in patients with negative sputum smear results. Chest 2001, 120: 349–355. pmid:11502628
- 33. Shah NS, Cavanaugh JS, Pratt R, Cain KP, Wells C, Laserson K, et al. Epidemiology of smear-negative pulmonary tuberculosis in the United States, 1993–2008. Int J Tuberc Lung Dis 2012, 16: 1234–1240. pmid:22748057
- 34. Colebunders R, Bastian I. A review of the diagnosis and treatment of smear-negative pulmonary tuberculosis. Int J Tuberc Lung Dis 2000, 4: 97–107. pmid:10694086
- 35. Dimairo M, Macpherson P, Bandason T, Zezai A, Munyati SS, Butterworth AE, et al. The risk and timing of tuberculosis diagnosed in smear-negative TB suspects: a 12 month cohort study in Harare, Zimbabwe. PLoS One 2010, 5: e11849. pmid:20676374
- 36. Martinson NA, Karstaedt A, Venter WD, Omar T, King P, Mbengo T, et al. Causes of death in hospitalized adults with a premortem diagnosis of tuberculosis: an autopsy study. AIDS 2007, 21: 2043–2050. pmid:17885294
- 37. Hargreaves NJ, Kadzakumanja O, Whitty CJ, Salaniponi FM, Harries AD, Squire SB. 'Smear-negative' pulmonary tuberculosis in a DOTS programme: poor outcomes in an area of high HIV seroprevalence. Int J Tuberc Lung Dis 2001, 5: 847–854. pmid:11573897
- 38. Mukadi YD, Maher D, Harries A. Tuberculosis case fatality rates in high HIV prevalence populations in sub-Saharan Africa. AIDS 2001, 15: 143–152. pmid:11216921
- 39. Apers L, Wijarajah C, Mutsvangwa J, Chigara N, Mason P, van der Stuyft P. Accuracy of routine diagnosis of pulmonary tuberculosis in an area of high HIV prevalence. Int J Tuberc Lung Dis 2004, 8: 945–951. pmid:15305475