How Many Sputum Culture Results Do We Need To Monitor Multidrug-Resistant-Tuberculosis (MDR-TB) Patients during Treatment?

Abstract

Discharge of a hospital patient after a single negative sputum culture may save money when treating multidrug-resistant tuberculosis. However, after initial sputum conversion in 336 South Africans, 11.6% and 5.4% reconverted after 1 and 2 months, respectively. These findings endorse the WHO definitions of 2 negative cultures taken ≥30 days apart after sputum culture conversion.

TEXT

Multidrug-resistant tuberculosis (MDR-TB) is a global problem, with an estimated 310,000 cases worldwide in 2011 (1). With 10,085 notified cases, South Africa ranks high on the global scale (1). Molecular epidemiologic data suggest that both nosocomial transmission and community transmission play an important role, because a significant proportion of patients have primary infection with MDR-TB (2).

The follow-up of MDR-TB cases for treatment efficacy and infectiousness includes monitoring sputum smear conversion and culture results. Expert opinion-based definitions suggest two consecutive negative smears or cultures, taken 30 days apart (3, 4). Recent World Health Organization (WHO) recommendations for MDR-TB patients include the use of culture rather than microscopy alone, with a minimum of monthly sputum culture taken until culture conversion, followed by quarterly culture examination afterwards (4). However, this requires adequate quality of culture performance and limited resources are recognized as an important obstacle (4).

While still infectious, MDR-TB patients need to be isolated, even in the more recently recommended ambulatory care models (4). In cases of drug-susceptible TB, patients are considered infectious until the first 2 weeks of TB treatment are completed and sputum smears are negative on three consecutive days (5). However, recent studies using culture conversion report that the time until conversion may be longer, implying prolonged infectiousness (6–8). The question arose whether substantial transmission may continue to occur during treatment of drug-resistant TB in ambulatory settings (9).

It may be tempting in resource-strapped settings to use the first negative culture result to reevaluate the need for respiratory isolation. Certainly, two consecutive negative samples are better than one, but do we really need them both? Or, notwithstanding limited resources, would three negative cultures not be even better for infection control? This study aimed at evaluating the optimal number of consecutive sputum cultures after culture conversion for decisions on respiratory isolation of MDR-TB patients.

Clinical records from a case series of MDR-TB patients admitted to Sizwe Tropical Disease Hospital in Johannesburg, South Africa, between 1 January 2008 and 31 July 2009 were reviewed. This hospital serves as a referral center for all cases of MDR- and extensively drug-resistant (XDR)-TB for Gauteng province, South Africa. Treatment is initiated in-hospital; after sputum culture conversion, patients are discharged and followed up by a health facility-based ambulatory care model. Ethical approval was obtained from the Ethics Committee of the Faculty of Health Sciences of the University of the Witwatersrand.

Demographic data, Mycobacterium tuberculosis resistance patterns, treatment regimens, and parameters of treatment response (sputum culture, discharge date, adverse events, death, loss to follow-up) were extracted from patient files. Inclusion criteria for the current analysis were having sputum smear-positive MDR-TB and resistance to at least isoniazid and rifampin, possibly in addition to resistance to one or more of the following; ethambutol, ethionamide, pyrazinamide, or streptomycin. Patients with monodrug resistance, or resistance to ofloxacin and kanamycin (XDR-TB), were excluded. All patients were treated according to their susceptibility test results.

For each patient in this case series, progression over consecutive sputum cultures was assessed. Only a single monthly culture following initial sputum culture conversion (from positive to negative) was counted; consecutive sputum cultures were labeled “month 1” if taken between day 7 and day 37, “month 2” if taken between day 37 and day 67, etc. For patients with multiple sputum cultures within 1 month, the latest result was included. The Kaplan-Meier method was used to calculate the cumulative proportion remaining sputum culture negative per month following initial conversion (positive-negative) and to calculate the cumulative proportion remaining sputum culture positive following reconversion (positive-negative-positive). Patients who were deceased, discharged, or lost to follow-up were censored, which means that these patients were excluded from the denominator of cumulative proportion calculation at all time points following their censoring event.

A total of 371 sputum-culture-positive MDR-TB patients were included; 199 (53.6%) were male, and the median age was 36 (interquartile [IQ] range, 29 to 45). On admission, 272 (73.3%) were HIV positive. Of those 272, 158 (58.1%) were already receiving antiretroviral treatment; the median CD4 cell count was 157 (IQ range, 80 to 296). Of the total of 371 patients, 26 patients died and 9 were lost to follow-up before a first negative culture, leaving 336 patients who had a first sputum culture conversion after various intervals.

Results of progression over consecutive sputum cultures are shown in Table 1. Sputum cultures of 11.6% (95% confidence interval [CI], 8.0 to 15.1) of patients reconverted to positive within 1 month following the initial negative culture result. An additional 5.4% (95% CI, 1.0 to 9.5) reconverted after two consecutive negative cultures (the formal culture conversion definition) (Table 1). Importantly, in 60.9% (95% CI, 43.9 to 72.7) of patients with a reconversion, the sputum culture returned negative within 1 month. In all of these patients, the sputum culture returned negative within 4 months (Table 1).

Table 1.

Kaplan-Meier survival table presenting the cumulative proportion of the patient population remaining sputum culture negative per month following the initial sputum culture conversion in 336 South African MDR-TB patients and the cumulative proportion remaining sputum culture positive per month in the 54 patients with a reconversion after initial conversion^a

Culture conversion category and time point	No. of patients at risk	No. of reconversion events (p-n-p)	No. of conversion events (p-n-p-n)	Cumulative proportion remaining negative (95% confidence interval)	Cumulative proportion remaining positive (95% confidence interval)	% reconverted to positive	% converted to negative	No. of censored patientsb	No. of patients in indicated censoring category
									Deceased	Lost to follow-up and discharged
Monthly sputum cultures following initial conversion (p-n)
Mo 1	318	37		0.884 (0.849–0.920)		11.6		18c	10	7
Mo 2	232	14		0.830 (0.789–0.874)		5.4		49	6	43
Mo 3	119	1		0.823 (0.780–0.869)		0.7		99	3	96
Mo 4	32	2		0.772 (0.695–0.857)		5.1		86	1	85
Monthly sputum cultures following reconversion after initial conversion (p-n-p)
Mo 1	46		28		0.391 (0.273–0.561)		60.9	8	0	8
Mo 2	17		12		0.115 (0.051–0.261)		27.6	1	0	1
Mo 3	5		3		0.046 (0.012–0.178)		6.9	0	0	0
Mo 4	2		1		0.023 (0.003–0.159)		2.3	0	0	0

^aThe population at risk was calculated based on the previous month's reconversion events and the current month's censored events. Of the 44 patients with a positive-negative-positive-negative (p-n-p-n) conversion event, 3 reconverted a second time (p-n-p-n-p) and 1 reconverted a third time (p-n-p-n-p-n-p) before being discharged with a negative culture result.

^bData corresponding to patients who were deceased, discharged, or lost to follow-up were excluded from the denominator of the cumulative proportion calculation at time points following the censoring event.

^cOne patient was censored because of toxic hepatitis.

Limited data on follow-up cultures were available for discharged patients. Follow-up cultures were available for only 100 (40.2%) of 249 patients discharged after at least one negative culture. Sixteen (16.0%) of these 100 patients had a positive sputum culture postdischarge (reconversion), which is comparable to the proportion observed during inpatient follow-up.

Culture-based monitoring of MDR-TB patients is used to evaluate treatment efficacy and helps to identify those who remain infectious. The internationally agreed-upon definition of culture conversion is two consecutive negative cultures ≥ 30 days apart (3), which^. certainly reduces false-negative results compared to a single negative culture. In fact, in this study, 37 (11.6%) inpatients had a positive culture after a first negative culture result, which supports the definitions for sputum culture conversion used for the current WHO guidelines on the management of MDR-TB (3, 4).

In another setting, namely, New York City, NY, in the United States in the mid-1990s, the yield of continued monthly cultures after culture conversion showed a small percentage (1.3%) of reconversions in drug-susceptible TB, leading the authors of the study to conclude that culture after conversion was unnecessary (10). Culture reconversion has been described in cases of MDR-TB (8), and a recent study, looking into the optimal frequency and type of microbiological monitoring for MDR-TB treatment, reported a culture reconversion rate of 23.7% (11).

Our results imply that three consecutive negative cultures would offer a minimal improvement before reevaluation of the need for respiratory isolation. Yet this would dramatically increase the isolation period, and this would probably not be feasible in nonaffluent settings of high TB endemicity.

In this study, a considerable amount of patients were lost to follow-up. This might have led to a selection of clinically worse patients staying in the hospital, the patients with improved health having been discharged or lost to follow-up. Because of this limitation, generalization of our results to the ambulatory setting must be done with caution. Less close supervision and a less tight management structure than in the hospital setting might lead to a deterioration of adherence and compliance, leading to a less favorable outcome picture for the ambulatory setting. Recommending the management of MDR-TB patients in an ambulatory care model can be justified only when the quality of ambulatory care with regard to close supervision of medication adherence and adequate respiratory isolation can be guaranteed.

In fact, a study from Peru, where an ambulatory system of care is used for several years, showed that there is a high risk of disease recorded in household contacts of patients with MDR- and XDR-TB (12). The fact that only 40.2% of discharged patients in our study provided a follow-up sputum sample further illustrates the challenges encountered in an ambulatory care model.