Journal of Antimicrobial Chemotherapy (27) 59, 794 798 doi:.93/jac/dkm25 Sputum conversion among patients with pulmonary tuberculosis: are there implications for removal of respiratory isolation? Jesús Fortún *, Pilar Martín-Dávila, Auxiliadora Molina 2, Enrique Navas, José Manuel Hermida, Javier Cobo, Enrique Gómez-Mampaso 2 and Santiago Moreno Infectious Diseases Department, Ramon y Cajal Hospital, Madrid, Spain; 2 Microbiology Department, Ramon y Cajal Hospital, Madrid, Spain Received 2 September 26; returned 24 November 26; revised January 27; accepted 2 January 27 Background: Limited data are available to predict the length of time required for a patient to achieve sputum culture conversion after starting therapy for pulmonary tuberculosis. Methods: Rates of sputum smear and culture conversion were determined at weeks 2, 4, 8 and 6 after initiating therapy in patients admitted to our Respiratory Isolation Unit from January 997 to December 23. Results: For the 84 patients included in the analysis, the mean time from the initiation of appropriate therapy to sputum culture and smear conversion were 34 + 26 and 38 + 32 days (mean + SD) respectively. Only 53% of patients obtained negative sputum cultures within the first 4 weeks of therapy. Multivariate analysis showed that the persistence of positive cultures during the first 4 weeks of therapy was associated with high bacillary counts in sputum smears at diagnosis [OR: 2.86; 95% confidence interval (95% CI):.2 6.66], lung cavitations (OR: 4.; 95% CI:.63 9.9) and a prolonged period of symptoms (OR: 3.57; 95% CI:.43 3.57). The only factor associated with the persistence of positive cultures after more than 6 weeks of therapy was infection with a multidrug-resistant strain. Conclusions: High initial sputum bacillary counts and drug resistance result in delayed culture conversion. This should be taken into account when decisions regarding the potential discontinuation of isolation are made. The early identification of drug resistance is important for effective infection control in hospitals. Keywords: Mycobacterium tuberculosis, drug-resistant tuberculosis, resistance, infection control Introduction Data regarding the point at which respiratory isolation can safely be discontinued for hospital patients undergoing treatment for active tuberculosis (TB) are scare. The available evidence regarding the infectiousness of patients on treatment comes from a comparison of the prevalence of positive tuberculin skin tests among household contacts, experimental models 2 and the quantification of viable cough-generated aerosols. 3 The available data suggest that once patients with TB begin receiving effective chemotherapy, they rapidly become less infectious. 4 Different studies have demonstrated that specific circumstances, including lung cavitation, larynx involvement, a prolonged course of symptoms and incorrect treatment, increase the time to sputum conversion. 5 In the USA, guidelines from the Centers for Diseases Control and Prevention (CDC) recommend that isolation for hospitalized patients can only be discontinued once patients are receiving effective chemotherapy, their condition is improving clinically and when three consecutive sputum samples, collected on different days, are negative for acid-fast bacilli (AFB). 5 This last recommendation is mandatory in patients returning to a congregate setting (e.g. a homeless shelter or detention facility). However, the reduced number of available isolation beds in hospitals, hospital policies aimed at avoiding prolonged admissions and the rapid improvement of patients after starting antituberculous therapy make it difficult to apply these guidelines in all circumstances. To examine this area further, a prospective study was performed to estimate the duration of respiratory isolation required... *Corresponding author. Tel: þ34-9-336-87-; Fax: þ34-9-336-87-92; E-mail: fortun@mi.madritel.es... 794 # The Author 27. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org
Sputum conversion and consequences for isolation in tuberculosis for patients with TB on the basis of weekly sputum AFB stains and cultures. A secondary analysis investigated predictors for time to sputum culture conversion among patients with active TB. In addition, we sought to determine whether respiratory isolation could reasonably be discontinued for certain patients. For the purposes of analysis, factors associated with sputum conversion were determined for two different periods: an early period (weeks 2 and 4) and late period (weeks 8 and 6). Multiple logistic regression analysis was used in both cases to determine the independent risk factors associated with a persistence of positive sputum cultures. All variables with a P,. observed in the univariate analysis were entered in the multivariate model. Patients and methods Setting Ramón y Cajal Hospital is a University teaching hospital where more than 8 patients with TB attend every year. Patients with TB requiring hospitalization may be admitted to different departments in the hospital, including Infectious Disease, Respiratory, Paediatric and General Internal Medicine Departments. Since 995, a Respiratory Isolation Unit was created in the hospital to admit all hospitalized patients from different departments requiring respiratory isolation. The Respiratory Isolation Unit includes single rooms, all of which are equipped with negative air pressure in relation to the corridor. Subjects Patients admitted to the Respiratory Isolation Unit between January 997 and December 23, who were managed by the staff of the Infectious Disease Department, were studied prospectively. Treatment regimens and duration were adjusted to CDC guidelines. With the exception of cases with a known resistant infection, all patients received treatment with isoniazid, rifampicin and pyrazinamide, with or without ethambutol. To estimate the sterilization rate, only patients with sputum cultures, performed weekly during and after hospitalization until a sterile culture was obtained, were selected. Culture conversion rates after 2, 4, 8 and 6 weeks of therapy were calculated. As the study had an observational design, patient consent and the approval of Internal Review Boards were not required. Microbiological methods The auramine stain was used during the study. The diagnosis of TB was confirmed microbiologically by a positive mycobacterial culture in clinical samples. Mycobacterial species and resistant strains were identified using DNA probes (AccuProbe, Gen Probe, USA) for Mycobacterium tuberculosis complex. Biochemical, growth test and DNA typing (including RFLP and spoligotyping in multiresistant strains) were used for final identification. In vitro susceptibilities to isoniazid, rifampicin, ethambutol, pyrazinamide and second-line drugs were determined as recommended by the CLSI (formerly NCCLS) using the standard proportion method on Middlebrook 7H agar medium (Bio Medics SL, Madrid, Spain). Multiresistant TB was recorded when resistance to isoniazid and rifampicin (with or without resistance to other drugs) was present. Statistical analysis Analysis focused on risk factors associated with the length of time to AFB smear and culture conversion. Student s t-test was used for continuous data and the Mantel Haenszel extended x 2 test for categorical data. Fisher s exact test was used when the expected number of cases per cell was below five. Odds ratios, 95% confidence intervals and P values were calculated. Results Population of study From January 997 to December 23, 36 patients were diagnosed with pulmonary TB in our centre. Of these patients, 84 (6%) fulfilled all the inclusion criteria and are the basis of this study. The remaining 22 patients were not included because they were managed as outpatients (36 patients), they were hospitalized but not cared for by the staff of the Infectious Disease Department (45 patients) or the weekly microbiological studies could not be completed (4 patients: 5 died during hospitalization, moved to other cities during the first part of treatment and 26 who were lost for follow-up after discharge from the hospital). The baseline characteristics of the 84 patients studied are shown in Table. The patients who were not included in the study had similar demographic, epidemiological and clinical characteristics (data not shown). The therapy for the current episode of pulmonary TB was modified during the first 8 weeks in 34 (8%) of the 84 patients included in the study. Reasons for changes in therapy were the development of drug toxicity (2 patients) and suspicion of drug resistance due to a poor response (4 patients). Timing of sputum smear and culture conversion: associated factors The mean times (+SD) to sputum culture and AFB stain (smear) conversion were 4.8 + 3.7 and 5.5 + 4.6 weeks, respectively. Only 22% and 53% of patients had negative sputum cultures after 2 and 4 weeks of therapy, respectively (Figure a). The mean duration of isolation in hospital was 3.28 + 2.2 weeks or over a week less than the mean length of time required to achieve sputum culture conversion. Factors known to be associated with a high baseline bacillary burden were significantly associated with the persistence of positive cultures in the early period of therapy (until week 4) in both the univariate and multivariate analyses. These factors included a high baseline bacillary count in sputum (. AFB per fields, magnification), lung cavitations and a prolonged period of symptoms (.3 months) (Figure b and c and Table 2). HIV-infected patients showed a significantly higher culture conversion rate by week 4 than HIV-negative patients in the univariate analysis (75% versus 47%; OR: 3.33; 95% CI:.22 9.; P ¼.2). Negative cultures at this point were more common in patients with CD4 counts lower than 2 cells/mm 3 (82%) than in patients with more than 2 CD4 cells/mm 3 (64%). However, in the multivariate analysis, the HIV status did not emerge as being significantly associated with culture conversion during the early period (Table 2). 795
Fortún et al. Table. Characteristics of patients with pulmonary TB according to the inclusion in the study Patients included (n ¼ 84) (%) Epidemiological data gender (male) 27 (69) age in years (mean + SD) 48 + 2 foreign-born 48 (26) previous TB chemoprophylaxis 8 () previous TB therapy 3 (7) Underlying medical condition previously healthy 8 (59) HIV infection 39 (2) chronic bronchopathy 22 (2) TB clinical form pulmonary TB 5 (82) lung cavitations 97 (53) haemoptysis 8 () symptoms.3 months 57 (3) Microbiological characteristics basal sputum AFB stain. 68 (37) INH resistance 24 (3) multiresistance 9 (5) (a).8.6.4.2 (b).8.6.4.2 % Negative culture % Negative AFB stain % Negative culture if basal AFB > % Negative culture if basal AFB < INH, isoniazid. No differences were observed in relation to the group of patients not included in the analysis, with the exception of isoniazid resistance (3% versus 5%; P ¼.3) and multiresistance (5% versus %; P ¼.5). To evaluate the potential contribution of drug resistance to delay in conversion rates during the first 4 weeks of therapy, the analysis was also repeated, excluding patients with any resistance (isolated isoniazid resistance or multidrug resistance). The results were similar to those obtained with all the patients (data not shown). We also analysed the influence of different drug regimens on the rate of culture sterilization. After excluding patients who had a change of regimen because of drug resistance, no significant difference in culture conversion rate was observed in patients who required a change of therapy because of drug toxicity. The rate of sputum conversion was not different for patients infected with drug-susceptible strains, who did or did not receive ethambutol in addition to isoniazid, rifampicin and pyrazinamide. Multivariate analysis of the factors influencing culture conversion in the late period of therapy (later than week 6) showed that infection with a multiresistant strain was the only factor related to late sputum culture conversion (Figure d and Table 2). Smear conversion during therapy occurred later than culture conversion (Figure a). Factors significantly related to a delay in smear conversion to negative were previous administration of antituberculous drugs (therapy or prophylaxis), a high baseline sputum AFB stain, a duration of symptoms longer than 3 months and infection with a resistant strain (isoniazid resistance or multiresistance). (c).8.6.4.2 (d).8.6.4.2 % Negative culture in patients with cavitary lesions % Negative culture in patients without cavitary lesions % Negative culture in patients with multiresistant strains % Negative culture in patients without multiresistant strains Figure. Smear (AFB) and culture conversion during therapy in patients with pulmonary TB. (a) Smear (AFB) and culture conversion; all patients included. (b) Sputum culture conversion during therapy according to basal AFB stain quantification:. AFB per fields, magnification or, AFB per fields, magnification. (c) Sputum culture conversion during therapy according to the presence or absence of cavitary lesions. (d) Sputum culture conversion during therapy in patients with and without multiresistant strains. 796
Sputum conversion and consequences for isolation in tuberculosis Table 2. Logistic regression analysis of factors associated with sputum culture conversion at different time points Variable OR (95% CI) P After 2 weeks of therapy basal sputum AFB stain.. (2.77 33.3). After 4 weeks of therapy basal sputum AFB stain. 2.86 (.2 6.66).2 symptoms.3 months 3.57 (.43 3.57).7 lung cavitations 4. (.63 9.9).2 After 8 weeks of therapy basal sputum AFB stain. 2.5 (.33.).7 multiresistance 27.8 (.89 4.).36 After 6 weeks of therapy multiresistance out of range.6 Discussion A key conclusion from this study is that sputum cultures remain positive, on average, for a month after the commencement of treatment with standard drugs. Thus, the time required for culture conversion in patients with pulmonary TB is longer than often assumed. This has implications for isolation practice because, despite falling bacillary counts with treatment, patients will potentially be infectious to others while they remain culture positive. In the early treatment period, high sputum bacillary counts at diagnosis, cavitary lung lesions and a prolonged course of symptoms were the factors that we found were associated with slower sputum culture conversion rates. After the eighth week of treatment, only drug resistance was associated with delayed culture conversion. Our results were similar to those obtained by Telzak et al. 6 They analysed factors influencing time to sputum conversion among patients with smear-positive TB and found that the mean number of days to the first of three consecutive negative sputum smears was 33 days. In this study, on stepwise multiple regression analysis, cavitary disease, numerous AFB on the initial smear and no prior history of TB were the factors independently associated with an increased number of days for both smear and culture conversion. In the present study, three-quarters of patients with HIV infection had a negative culture after 4 weeks of therapy versus half of patients without HIV infection. A meta-analysis has demonstrated a reduced frequency of positive tuberculin skin test conversion among housing contacts of HIV tuberculous patients, 7 suggesting a reduced risk of TB transmission. The lower rates of cavitary disease observed in the HIV--infected patient, 8 mainly in patients with,2 CD4 cells/mm 3, may lead to a less effective dissemination of M. tuberculosis to contacts. Infection control policies for TB differ significantly between countries. In the UK, the previous British Thoracic Society guidelines 9 and the more recent National Institute of Clinical Excellence (NICE) guidelines indicate that the isolation of patients with pulmonary TB is generally only required for 2 weeks, unless risk factors for multidrug-resistant TB are present, as patients are considered to usually become non-infectious after this time period on a standard treatment regimen. However, Brindle et al. have demonstrated that patients receiving isoniazid, rifampicin, pyrazinamide and streptomycin required 28 days to get a reduction from 7 to 3.5 logs in cfu/ml sputum, and after the second week, only a 2 log reduction was observed. In the USA, the 994 CDC guidelines established more stringent criteria for removal of patients from isolation: three consecutive negative sputum smears. This represented a significant break from previous practice, which had also identified 2 weeks of TB therapy as a presumptive surrogate for non-infectiousness. 2 The recently updated CDC guidelines for TB infection control in healthcare settings again advise that all patients with suspected TB disease should remain under airborne precautions while they are hospitalized until they have had three consecutive negative AFB sputum smear results. 5 Our study shows that drug resistance is strongly associated with late (.8 weeks) culture conversion. Such patients will require prolonged isolation. The early identification of drug resistance is therefore mandatory for an effective hospital infection control policy. To identify such potentially hazardous cases, it has been recommended that drug resistance is determined in most smear-positive cases at an early stage by the use of modern laboratory techniques. 3 Strict criteria for the assessment of noninfectiousness (three consecutive AFB-negative sputum) should be applied in all cases where such tests demonstrate drug resistance. In conclusion, our study shows that the loss of infectiousness of pulmonary TB during therapy does not occur rapidly in all patients. In the late stages of treatment (.8 weeks), only drug resistance is strongly associated with delayed sputum culture conversion in compliant patients. We suggest that strict criteria for the assessment of non-infectiousness (three consecutive negative smears of sputum specimens) should be applied only in cases with confirmed or probable drug resistance or in patients who are likely to be slow converters (strongly smear positive at diagnosis, cavitary chest X-ray lesions or a prolonged duration of symptoms), with particular caution being exercised if a patient is remaining in hospital or returning to a congregate setting. 797
Fortún et al. Acknowledgements We have had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Transparency declarations None to declare. References. Malone JL, Ijaz K, Lambert L et al. Investigation of healthcareassociated transmission of Mycobacterium tuberculosis among patients with malignancies at three hospitals and at a residential facility. Cancer 24; : 273 2. 2. Kelley CL, Collins FM. Growth of a highly virulent strain of Mycobacterium tuberculosis in mice of differing susceptibility to tuberculous challenge. Tuberc Lung Dis 999; 79: 367 7. 3. Fennelly KP, Martyny JW, Fulton KE et al. Cough-generated aerosols of Mycobacterium tuberculosis: a new method to study infectiousness. Am J Respir Crit Care Med 24; 69: 64 9. 4. Jindani A, Aber VR, Edwards EA et al. The early bactericidal activity of drugs in patients with pulmonary tuberculosis. Am Rev Respir Dis 98; 2: 939 49. 5. Taylor Z, Nolan CM, Blumberg HM. Controlling tuberculosis in the United States. Recommendations from the American Thoracic Society, CDC, and the Infectious Diseases Society of America. MMWR Recomm Rep 25; 54: RR 2: 8. 6. Telzak EE, Fazal BA, Pollard CL et al. Factors influencing time to sputum conversion among patients with smear-positive pulmonary tuberculosis. Clin Infect Dis 997; 25: 666 7. 7. Cruciani M, Malena M, Bosco O et al. The impact of human immunodeficiency virus type on infectiousness of tuberculosis: a meta-analysis. Clin Infect Dis 2; 33: 922 3. 8. Girardi E, Raviglione MC, Antonucci G et al. Impact of the HIV epidemic on the spread of other diseases: the case of tuberculosis. AIDS 2; 4: Suppl 3: S47 56. 9. Anon. Control and prevention of tuberculosis in the United Kingdom: code of practice 2. Joint Tuberculosis Committee of the British Thoracic Society. Thorax 2; 55: 887 9.. Wiselka M. 26: a year for cautious optimism in tuberculosis. J Infect 26; 53: 4.. Brindle R, Odhiambo J, Mitchison D. Serial counts of Mycobacterium tuberculosis in sputum as surrogate markers of the sterilising activity of rifampicin and pyrazinamide in treating pulmonary tuberculosis. BMC Pulm Med 2; : 2. 2. Iseman MD. An unholy trinity-three negative sputum smears and release from tuberculosis isolation. Clin Infect Dis 997; 25: 67 2. 3. Caws M, Drobniewski FA. Molecular techniques in the diagnosis of Mycobacterium tuberculosis and the detection of drug resistance. Ann N Y Acad Sci 2; 953: 38 45. 798