Secondary Logo

Journal Logo

Epidemiology and Prevention

Implementation and Operational Research: Clinical Impact of the Xpert MTB/RIF Assay in Patients With Multidrug-Resistant Tuberculosis

Padayatchi, Nesri MBChB*; Naidu, Naressa MMedSc*; Yende-Zuma, Nonhlanhla MSc*; O'Donnell, Max Roe MD*,†,‡; Naidoo, Kogieleum MBChB*; Augustine, Stanton MBChB*; Zumla, Alimuddin FRCP§; Loveday, Marian PhD

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: September 1, 2016 - Volume 73 - Issue 1 - p e1-e7
doi: 10.1097/QAI.0000000000001110

Abstract

INTRODUCTION

Tuberculosis (TB) is a leading cause of morbidity and mortality worldwide, and the emergence and continuing spread of multidrug-resistant TB (MDR-TB) is of great public health concern. Only 6.8% of the estimated 440,000 MDR-TB incident cases, which emerged worldwide in 2008, were detected and reported.1 The protracted nature of conventional diagnostic tools, such as culture and drug susceptibility testing (DST), impedes case finding and delays appropriate treatment initiation, which in turn potentially fuels transmission, increases morbidity and mortality, and amplifies resistance.2,3

The advent of a polymerase chain reaction–based Xpert MTB/RIF assay (Cepheid, Sunnyvale, CA) in 2010 represented a new breakthrough in TB diagnostics, simultaneously detecting both Mycobacterium tuberculosis (MTB) complex DNA and rifampicin resistance in just 2 hours.4 Although rifampicin resistance is not a complete surrogate marker for MDR-TB, it is considered the most important indicator of MDR-TB as greater than 90% of rifampicin-resistant isolates are additionally resistant to isoniazid.5,6 In fact, Boehme et al7 found, in their multicenter study, that Xpert MTB/RIF correctly identified 97.5% of MDR-TB samples. In December 2010, the World Health Organization (WHO) endorsed the Xpert MTB/RIF assay for use in TB endemic settings, and by the end of June 2014, 3269 Xpert MTB/RIF instruments had been procured in the public sector across 108 countries.6

South Africa has the third highest burden of TB after India and China, and the fifth highest burden of MDR-TB globally. More than 60% of patients in South Africa are coinfected with HIV. South Africa rapidly rolled out and led the implementation of the Xpert MTB/RIF assay, accounting for 56% of all Xpert MTB/RIF assay cartridges procured worldwide.6 There have been numerous studies evaluating the accuracy and operational implementability of the Xpert MTB/RIF assay from various geographical regions in patients with all clinical types of TB, including MDR-TB.4,7–9 The Xpert MTB/RIF assay gives a result operationally within 24 hours of obtaining sputum. The diagnostic accuracy of the Xpert MTB/RIF assay has therefore been well validated, with an assay sensitivity of 89% and specificity of 99% when used as an initial test replacing smear microscopy.10 When used as an initial test replacing conventional DST, the Xpert MTB/RIF assay sensitivity demonstrated 95% sensitivity in the detection of rifampicin-resistant TB cases, with a specificity of 98%.10

Although the diagnostic advantages of the Xpert MTB/RIF assay over conventional microbiological methods are apparent, its clinical usefulness in improving treatment outcomes in patients with MDR-TB remains undefined. It is hypothesized that MDR-TB patients diagnosed by Xpert MTB/RIF will be more rapidly initiated on MDR-TB treatment, therefore present with less severe morbidity and that these benefits would translate into improved treatment outcomes compared with patients with MDR-TB diagnosed by conventional DST. The primary aim of this study was to evaluate the clinical impact of the Xpert MTB/RIF assay on treatment outcomes in patients with MDR-TB.

METHODS

Study Design, Participants, and Procedures

We conducted a study of 921 patients with MDR-TB, who presented to a specialist drug-resistant TB facility in KwaZulu-Natal, South Africa, pre- and post-implementation of the Xpert MTB/RIF assay (Fig. 1). Clinical, laboratory, chest radiograph, and follow-up data from 108 patients with MDR-TB, post-introduction of the Xpert MTB/RIF assay (Xpert group) in November 2010, were analyzed and compared with data from 813 MDR-TB patients from the pre-MTB/RIF assay period (Conventional group), July 2008–2010. Data for both cohorts were obtained through patient clinical charts at a specialist drug-resistant TB hospital in the KwaZulu-Natal Province, South Africa, between 2008 and 2011. Eligible participants were adult patients (≥18 years of age) with pulmonary MDR-TB (as defined by the WHO6), confirmed by conventional DST, receiving standard of care at this facility in accordance with the South African Management of Drug-Resistant Tuberculosis: Policy Guidelines.11 The treatment regimens between Conventional and Xpert groups did not differ. Patients with resistance to a fluoroquinolone or second-line injectable agent were not eligible for study inclusion.

FIGURE 1.
FIGURE 1.:
Flow diagram of participant enrollment in the 2 study groups.

Before the Xpert MTB/RIF assay implementation in South Africa, resistance testing was only performed when indicated (poor clinical response to first-line treatment or those with a previous history of TB). After the introduction of the Xpert MTB/RIF assay, rifampicin-resistant cases were referred for MDR-TB treatment initiation while awaiting confirmatory culture and DST results.12 Patients diagnosed with MDR-TB or rifampicin resistance were referred to this facility if their place of residence lies within the catchment area. Patients were admitted as inpatients or outpatients, after being placed on an admission waiting list, and were treated with a standardized MDR-TB treatment regimen comprising an intensive phase, which includes an injectable agent that is administered for at least 6 months, followed by a continuation phase of oral drugs administered for approximately 18 months.12

Clinical Management Impact Outcome Measures

The Primary outcome measure was “treatment success” (WHO definition)13,14 at 24 months since initiation of treatment. Secondary outcomes were time to initiation of MDR-TB treatment and TB-related morbidity at treatment initiation (chest radiographs were used as a surrogate marker of disease severity). Factors predicting treatment success were also analyzed.

Data Collection

Data for both cohorts were extracted from clinical chart reviews. Information on demographic characteristics, HIV, treatment, laboratory results, and final treatment outcomes were collected prospectively. Baseline chest radiographs were used as a marker of disease severity and morbidity at the initiation of MDR-TB treatment and captured retrospectively.

Definitions

Treatment outcome was assessed at the end of therapy. Definitions are as described in the WHO Guidelines for the Programmatic Management of Drug-Resistant Tuberculosis.13,14

Cured

A patient with MDR-TB is considered cured if they have completed treatment according to the program's protocol and had at least 5 consecutive negative cultures from samples collected at least 30 days apart in the final 12 months of treatment. If only one positive culture is reported during this time, with no concomitant clinical evidence of deterioration, a patient may still be considered cured, provided that this positive culture is followed by a minimum of 3 consecutive negative cultures taken at least 30 days apart.

Treatment Completed

A patient with MDR-TB who has completed treatment according to the program's protocol but does not meet the definition for cure because fewer than 5 cultures were performed in the final 12 months of treatment.

Treatment success is considered to be the sum of “cure” and “treatment completed.”

Default

Treatment interruption, without medical approval, for at least 2 consecutive months.

Treatment Failure

Patients were considered to have failed treatment if 2 or more of the 5 cultures performed in the final 12 months of therapy were positive, if any of the final 3 cultures were positive, or a clinical decision was made to terminate treatment early because of poor clinical response or adverse events.

Death

All-cause mortality during MDR-TB treatment.

Transferred Out

Patients whose treatment outcome was unknown because of transfer to another recording and reporting facility.

Time to Treatment Initiation

The time from initial sputum collection to MDR-TB treatment initiation.

Chest Radiograph Analysis and Scoring

Chest radiographs were performed on all patients at this specialist facility on the day of treatment initiation. We used uniform distribution to select a random sample of 147 (78 in the Conventional group and 69 in the Xpert group) baseline chest radiographs, which were scored by 2 independent medical practitioners trained in the reading of chest radiographs. Medical practitioners were blinded to the diagnostic group and scoring of the second reader. Chest radiograph scoring was based on a combination of the “NICE”15 and “Brust”16 scoring systems. Each lung field was divided into 3 zones (upper, middle, and lower), and a score of 0–4 for each zone was recorded on a standardized case report form. Radiographic scores were assigned based on disease severity: 0 = no disease, 1 = 1%–24% of the lung zone diseased, 2 = 25%–49% of lung zone diseased, 3 = 50%–74% of lung zone diseased, and 4 = 75%–100% of lung zone diseased. Scoring was performed for each of the following disease components per zone: nodules, infiltrates, cavities, bronchiectasis, and fibrosis. Additional scores were allocated to chest radiographs, in which pleural effusion, hilar lymphadenopathy, and/or pneumothorax was present. Scores for all 6 zones across both lung fields were added with a total possible score of 125.

Statistical Analyses

Fisher exact test was used for the analysis of categorical data, and unpaired t tests or the Wilcoxon 2-sample test was used for the analysis of continuous data. To measure the interreader variability or test agreement, we used Krippendorff alpha (α) and Bland–Altman diagrams. We considered α ≥ 0.800 to be good agreement, whereas α ≥ 0.667 was an acceptable agreement.17 In the Bland–Altman plots, the limits of agreement were placed at a distance of 2 times the SD of the differences. Multivariate log binomial regression was used to identify predictors of TB treatment success. All statistical tests were 2 sided. Statistical analyses were performed using SAS, version 9.3 (SAS Institute, Cary, NC).

Ethics Approval

The study was approved by the South African Medical Research Council Ethics Committee (Ref EC028-10/2012).

RESULTS

Figure 1 depicts the flow of patient enrollment into the 2 study groups. Of a total of 187 patients who were eligible for inclusion at baseline in the Xpert group, 79 were excluded from analyses: 17 had no confirmatory DST result available, 23 were found to be negative for MDR-TB, 6 were found to have drug-sensitive TB, 17 were either isoniazid or rifampicin mono-resistant, and 16 exhibited other resistance profiles (poly-resistant, pre-extensively drug-resistant tuberculosis (pre-XDR-TB), or XDR-TB).

Baseline Demographic Characteristics

The demographic and clinical characteristics were similar between the 2 groups (Table 1); patients were mostly young adults with 50.8% versus 58.3% of the cohort represented by females in the Conventional and Xpert groups, respectively. In both groups, a large proportion of patients were HIV coinfected, 70.8% in the Conventional group and 78.7% in the Xpert group with 75.2% of patients initiated on antiretroviral therapy (ART) in the Conventional group and 77.6% in the Xpert group. Analysis of the demographic and clinical characteristics of patients in the Xpert group, excluded in the final analysis, showed that age, sex, and HIV status were not markedly different compared with Xpert MTB/RIF patients included in the final analysis (see Table 6, Supplemental Digital Content, https://links.lww.com/QAI/A853). The proportion of patients on ART was, however, 77.6% among those included in the final analysis versus 51.6% in patients excluded in the final analysis.

TABLE 1.
TABLE 1.:
Baseline Demographic Characteristics of Patients Included in the Final Analysis

Treatment Outcomes

Table 2 shows outcomes of treatment in both study groups. There were no significant differences in treatment success rates between the pre-Xpert MTB/RIF assay Conventional group and post-Xpert MTB/RIF assay group (54.0% versus 56.5%, P = 0.681). Cure rates were significantly higher in the Xpert group compared with the Conventional group (45.4% versus 34.4%, P = 0.032) while this trend was reversed for treatment completion (11.1% versus 19.6%, P = 0.035). Default rates were significantly higher in the Conventional group (23.8%) than in the Xpert group (18.5%) (P = 0.038). Mortality was higher in the Xpert group (19.4%) compared with the Conventional group (13.9%), although this was not found to be statistically significant (P = 0.415). In sensitivity analysis that included the 79 patients in the Xpert group who were initially excluded, the cure rate was reduced from 45.4% to 40.6% and there was no longer a statistically significant difference between the Conventional and Xpert groups (see Table 7, Supplemental Digital Content, https://links.lww.com/QAI/A853).

TABLE 2.
TABLE 2.:
Treatment Outcomes of Patients in the 2 Study Groups

Time to Treatment Initiation

The median time to treatment initiation was considerably reduced by 78% in the Xpert group compared with the Conventional group. Median time to treatment initiation was 20 days [interquartile range (IQR), 13–31] in the Xpert group versus 92 days (IQR, 69–120) in the Conventional group (P < 0.001).

Chest Radiography as a Marker of Morbidity at Enrollment

Among the 78 patients with chest radiograph scoring in the Conventional group, 60 (76.9%) were HIV positive with 54 (90.0%) on ART while the HIV status of 2 (2.6%) patients was unknown. In the Xpert group, among the 69 chest radiographs that were scored, 54 (78.3%) were HIV positive with 50 (92.6%) on ART while the HIV status of 1 (1.5%) patient was unknown.

Baseline chest radiograph score for reader 1 and Krippendorff (α) estimates are shown in Table 3. The median chest radiograph score was 10 (IQR, 6–13) and 13 (IQR, 10–15) in the Conventional and Xpert groups, respectively (P < 0.001). Good interreader agreement was observed in both groups [α = 0.96, 95% confidence interval (CI): 0.94 to 0.98 in the Conventional group, and α = 0.81, 95% CI: 0.70 to 0.89 in the Xpert group]. This was also confirmed by the Bland–Altman plots (see Figure 2, Supplemental Digital Content, https://links.lww.com/QAI/A853). Because baseline morbidity was only evaluated for a subset of randomly selected patients, to exclude selection bias, subanalyses confirmed that treatment outcomes between the Conventional and Xpert groups were similar (Table 4). Interestingly, although treatment completed and cure are both treatment success, patients considered “cured” had greater baseline morbidity than patients with the outcome treatment completed [10 (7–13) cured versus 7 (6–13) treatment completed in the Conventional group and 13 (10–16) cured versus 12 (9–17) treatment completed in the Xpert group].

TABLE 3.
TABLE 3.:
Total Baseline* Chest Radiograph Score and Krippendorff (α) Estimates
TABLE 4.
TABLE 4.:
Median (IQR) Chest Radiograph Scores Stratified by Treatment Outcome

Predictors of Treatment Success

Table 5 shows multivariate log binomial regression analysis of factors associated with MDR-TB treatment success. HIV status, ART status, and age were the only predictors of treatment success. HIV-positive patients initiated on ART were more likely to achieve treatment success [risk ratio (RR): 1.32, 95% CI: 1.08 to 1.62, P = 0.007] than HIV-positive patients not on ART. Furthermore, HIV-negative patients had a significantly higher probability of treatment success (RR: 1.32, 95% CI: 1.06 to 1.64, P = 0.013) when compared with HIV-positive patients not on ART. Older patients had 4% higher chance of treatment success than younger patients (RR: 1.04, 95% CI: 1.01 to 1.06, P = 0.003). Baseline chest radiograph scores were not found to be predictive of treatment success.

TABLE 5.
TABLE 5.:
Multivariate Log Binomial Regression Analysis of Factors Associated With MDR-TB Treatment Success

DISCUSSION

Despite the Xpert MTB/RIF assay considerably reducing the time to treatment initiation, our data show several interesting findings; most importantly, there were no significant differences in treatment success rates (per WHO definition: sum of cure and treatment completed) between MDR-TB patients in the pre-Xpert MTB/RIF period who were diagnosed using conventional microbiological methods and those who were diagnosed in the period where the Xpert MTB/RIF assay was used. The rapid rollout and implementation of the Xpert MTB/RIF assay in South Africa provided a new rapid diagnostic test and increased referrals of patients suspected of having MDR-TB. Although Xpert MTB/RIF has recently shown no effect on morbidity or mortality in drug-sensitive TB,18,19 given the protracted nature of both MDR-TB treatment regimens and conventional DST, a greater opportunity for impact of the Xpert MTB/RIF assay was anticipated. To our knowledge, this study is the first to evaluate the clinical impact of the use of the Xpert MTB/RIF assay on the treatment outcomes of South African patients with MDR-TB.

We had postulated that early and rapid diagnosis would result in reduced baseline morbidity, positively affecting treatment outcomes. There are several plausible reasons for the unexpected findings of our study. First, morbidity was found to be greater in the Xpert group; the increased disease at baseline negates any potential cumulative effect of reduced time to treatment initiation. Second, severely ill patients may have died awaiting conventional DST results and MDR-TB treatment initiation. These patients therefore did not enter the Conventional group of our study, possibly introducing a selection bias and accounting for the lower morbidity and mortality reflected in this group. This is supported by programmatic data showing that approximately 50% of patients diagnosed with MDR-TB, in South Africa, in 2009 were not initiated on appropriate treatment.11 Gandhi et al20 has additionally shown that mortality from MDR-TB and XDR-TB was high and often occurred within 30 days of sputum collection in a high HIV prevalence setting in KwaZulu-Natal. Third, although the Xpert MTB/RIF assay is a revolutionary diagnostic tool, it does neither alter the behaviors of patients who often delay seeking treatment nor address poor medication adherence to protracted and toxic MDR-TB treatment regimens.21

Interestingly, the proportion of patients who defaulted in the Xpert MTB/RIF assay group was significantly lower than those in the Conventional group. Although Theron et al19 found that the Xpert MTB/RIF assay reduced dropout (culture-positive patients not started on TB treatment) in drug-sensitive TB, it is unclear why patients were better retained in the Xpert group of our study. The fragile public health system has, however, been shown to be an impediment to those who try to seek medical care.22 It is possible that because patients in the Conventional group had to engage with the health care system for lengthier periods of time, they may have become increasingly disillusioned with it, further exacerbating the challenge of treatment interruption. Successful MDR-TB treatment outcomes have previously been found to correlate with health system performance.23

Although successful treatment outcomes did not differ between the Conventional and Xpert MTB/RIF assay groups, treatment completion was significantly higher in the Conventional group compared with the Xpert group with a trend reversal for patients considered cured. Because the difference between the definitions for treatment completed and cured is a lack of bacteriological evidence in the former,13,14 it may be that patients in the Xpert group were more likely to produce sputum, because of increased baseline morbidity. In fact, stratification of the median morbidity at baseline, by treatment outcome, supports this argument and shows that in patients successfully treated those in the Xpert group exhibited significantly higher baseline morbidity than patients in the Conventional group (Table 4). Patients were also 32% more likely to achieve MDR-TB treatment success if they were on ART, or HIV negative, compared with HIV-positive patients not on ART. This is consistent with the findings of other studies documenting the role of HIV status and ART initiation in MDR-TB patients.20,24 Increasing age was found to be a predictor of treatment success. This finding likely reflects the more adherent nature of older patients. Baseline chest radiographic scores were not found to be predictive of treatment success, possibly because of the small subset analyzed. Only a small subset of chest radiographs were scored so as to make the study more generalizable because chest radiographs are not readily available in most resource-limited settings.

WHO recommends the Xpert MTB/RIF assay for use as the initial test for individuals at risk of MDR-TB, on the basis of data that show it has similar accuracy as that of conventional culture and DST for rifampicin resistance.25 The advantage of Xpert MTB/RIF is that the assay provides results operationally within 24 hours rather than weeks for assays that use liquid media and months that solid media take to provide a result. Large-scale demonstration studies under research conditions have shown that Xpert MTB/RIF assay introduction is feasible in high TB-burden countries.4,19 The significant reduction in time to treatment initiation in our study is consistent with several others, within the context of both drug-resistant TB25–27 and drug-sensitive TB,4,19,28–30 and has important public health implications as it significantly reduces disease transmission. Given that the rapid diagnostic nature of Xpert MTB/RIF assay is 2 hours, the delay, about 3 weeks in our study, is still considerable. This alludes to the myriad of underlying programmatic challenges in the implementation of the Xpert MTB/RIF assay. We have previously noted the importance of scientific advances and health system strengthening as complementary processes.31 Indeed, a recent retrospective study in the Western Cape Province of South Africa showed that a combination of decentralized MDR-TB management and Xpert MTB/RIF assay implementation reduced treatment delay, in patients with rifampicin-resistant TB, to a median of just 8 days. One of the potential factors contributing to the delay in treatment initiation in our study was the placement of the Xpert MTB/RIF assay at a centralized laboratory rather than at the point of care. Numerous studies have demonstrated that point-of-care Xpert MTB/RIF assay testing significantly reduces the time to treatment initiation and often results in same-day treatment initiation.19,29,32 Although these studies were conducted within the realm of drug-sensitive TB, further decentralization and use of the Xpert MTB/RIF assay near the point of care could potentially improve existing system delays for the treatment of MDR-TB.27

Our study has several limitations. Although the data were collected prospectively for both cohorts of patients, pre- and post-implementation of the Xpert MTB/RIF assay, the study relied on routine programmatic data collected by health workers and data recording, which may have had errors; however, the errors would have been across both periods of study. The possible selection bias resulting in exclusion of terminally ill patients could have substantially reduced the impact of the Xpert MTB/RIF assay. Our study was performed shortly after the implementation of Xpert MTB/RIF, in a specialized drug-resistant TB facility, during a period where there was limited decentralization and access to drug-resistant TB treatment and substantial reliability on centralized care. There may have been subsequent improvements because of the expansion of decentralization and health care workers becoming more familiar with the implementation algorithms of Xpert MTB/RIF. We also note the disproportionate sample size between the 2 groups; however, this had no statistical impact. Finally, our study was conducted in a region with a high HIV burden and may therefore not be generalizable to other settings.

CONCLUSIONS

Although the Xpert MTB/RIF assay has revolutionized rapid TB diagnostics, our study failed to show that Xpert MTB/RIF positively affects clinical treatment outcomes. Recent studies evaluating the Xpert MTB/RIF assay under programmatic conditions have also shown no effect on its use on morbidity or mortality in patients with drug-sensitive TB. Because rapid diagnosis and treatment initiation are critical to reducing the period of infectiousness and for interrupting transmission, the Xpert MTB/RIF assay may also have a role in reducing the burden of disease, and the public health impact of the Xpert MTB/RIF assay requires evaluation. The full potential of the Xpert MTB/RIF assay and its role in underresourced fragile public health systems requires further definition if we are to maximize their impact.

ACKNOWLEDGMENTS

The authors would like to acknowledge Garth Osburn, Prem Moodley, and Kumeren Govender for their contributions.

REFERENCES

1. World Health Organization. Multidrug and Extensively Drug-resistant TB (M/XDR-TB): Global Report on Surveillance and Response. WHO, Geneva, Switzerland: 2010.
2. Kipiani M, Mirtskhulava V, Tukvadze N, et al. Significant clinical impact of a rapid molecular diagnostic test (Genotype MTBDRplus assay) to detect multidrug-resistant tuberculosis. Clin Infect Dis. 2014;59:1559–1566.
3. Piatek AS, Van Cleeff M, Alexander H, et al. GeneXpert for TB diagnosis: planned and purposeful implementation. Glob Health Sci Pract. 2013;1:18–23.
4. Boehme CC, Nicol MP, Nabeta P, et al. Feasibility, diagnostic accuracy, and effectiveness of decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug resistance: a multicentre implementation study. Lancet. 2011;377:1495–1505.
5. Rahman A, Sahrin M, Afrin S, et al. Comparison of Xpert MTB/RIF assay and GenoType MTBDRplus DNA probes for detection of mutations associated with rifampicin resistance in Mycobacterium tuberculosis. PloS One. 2016;11:e0152694.
6. World Health Organization. Global Tuberculosis Report. Geneva, Switzerland: World Health Organization; 2014.
7. Boehme CC, Nabeta P, Hillemann D, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med. 2010;363:1005–1015.
8. Hillemann D, Rusch-Gerdes S, Boehme C, et al. Rapid molecular detection of extrapulmonary tuberculosis by the automated GeneXpert MTB/RIF system. J Clin Microbiol. 2011;49:1202–1205.
9. Nicol MP, Workman L, Isaacs W, et al. Accuracy of the Xpert MTB/RIF test for the diagnosis of pulmonary tuberculosis in children admitted to hospital in Cape Town, South Africa: a descriptive study. Lancet Infect Dis. 2011;11:819–824.
10. Steingart KR, Schiller I, Horne DJ, et al. Xpert MTB/RIF assay for pulmonary tuberculosis and rifampicin resistance in adults. Cochrane Database Syst Rev. 2014;1:CD009593.
11. National Department of Health. Management of drug-resistant tuberculosis: Policy Guidelines. Republic of South Africa, Pretoria, South Africa. 2013.
12. National Department of Health. Management of drug-resistant tuberculosis: Policy Guidelines. Republic of South Africa. 2011.
13. World Health Organisation. Guidelines for the Programmatic Management of Drug-resistant Tuberculosis. Emergency Update. Geneva, Swizerland: World Health Organization; 2008.
14. Laserson KF, Thorpe LE, Leimane V, et al. Speaking the same language: treatment outcome definitions for multidrug-resistant tuberculosis. Int J Tuber Lung Dis. 2005;9:640–645.
15. Kurashima A, Morimoto K, Horibe M, et al. A method for visual scoring of pulmonary Mycobacterium Avium complex disease: “NICE scoring system”. J Mycobac Dis. 2013;3.
16. Brust JC, Berman AR, Zalta B, et al. Chest radiograph findings and time to culture conversion in patients with multidrug-resistant tuberculosis and HIV in Tugela Ferry, South Africa. PloS One. 2013;8:e73975.
17. Krippendorff K. Content Analysis: An Introduction to its Methodology. 2nd ed. Thousand Oaks, CA: Sage; 2004.
18. Churchyard GJ, Stevens WS, Mametja LD, et al. Xpert MTB/RIF versus sputum microscopy as the initial diagnostic test for tuberculosis: a cluster-randomised trial embedded in South African roll-out of Xpert MTB/RIF. The Lancet Glob Health. 2015;3:e450–457.
19. Theron G, Zijenah L, Chanda D, et al. Feasibility, accuracy, and clinical effect of point-of-care Xpert MTB/RIF testing for tuberculosis in primary-care settings in Africa: a multicentre, randomised, controlled trial. Lancet. 2014;383:424–435.
20. Gandhi NR, Shah NS, Andrews JR, et al. HIV coinfection in multidrug- and extensively drug-resistant tuberculosis results in high early mortality. Am J Respir Crit Care Med. 2010;181:80–86.
21. O'Donnell MR, Wolf A, Werner L, et al. Adherence in the treatment of patients with extensively drug-resistant tuberculosis and HIV in South Africa: a prospective cohort study. JAIDS. 2014;67:22–29.
22. Loveday M, Padayatchi N, Voce A, et al. The treatment journey of a patient with multidrug-resistant tuberculosis in South Africa: is it patient-centred? Int J Tuberc Lung Dis. 2013;17(10 suppl 1):56–59.
23. Loveday M, Padayatchi N, Wallengren K, et al. Association between health systems performance and treatment outcomes in patients co-infected with MDR-TB and HIV in KwaZulu-Natal, South Africa: implications for TB programmes. PloS One. 2014;9:e94016.
24. Isaakidis P, Cox HS, Varghese B, et al. Ambulatory multi-drug resistant tuberculosis treatment outcomes in a cohort of HIV-infected patients in a slum setting in Mumbai, India. PloS One. 2011;6:e28066.
25. van Kampen SC, Susanto NH, Simon S, et al. Effects of introducing xpert MTB/RIF on diagnosis and treatment of drug-resistant tuberculosis patients in Indonesia: a pre-post Intervention study. PloS One. 2015;10:e0123536.
26. Cox HS, Daniels JF, Muller O, et al. Impact of decentralized care and the xpert MTB/RIF test on rifampicin-resistant tuberculosis treatment initiation in Khayelitsha, South Africa. Open Forum Infect Dis. 2015;2:ofv014.
27. Dlamini-Mvelase NR, Werner L, Phili R, et al. Effects of introducing Xpert MTB/RIF test on multi-drug resistant tuberculosis diagnosis in KwaZulu-Natal South Africa. BMC Infect Dis. 2014;14:442.
28. Cox HS, Mbhele S, Mohess N, et al. Impact of Xpert MTB/RIF for TB diagnosis in a primary care clinic with high TB and HIV prevalence in South Africa: a pragmatic randomised trial. PLoS Med. 2014;11:e1001760.
29. Hanrahan CF, Selibas K, Deery CB, et al. Time to treatment and patient outcomes among TB suspects screened by a single point-of-care xpert MTB/RIF at a primary care clinic in Johannesburg, South Africa. PloS One. 2013;8:e65421.
30. Yoon C, Cattamanchi A, Davis JL, et al. Impact of Xpert MTB/RIF testing on tuberculosis management and outcomes in hospitalized patients in Uganda. PloS One. 2012;7:e48599.
31. Padayatchi N, Loveday M, Naidu N. Drug-resistant tuberculosis control in South Africa: scientific advances and health system strengthening are complementary. Expert Opin Pharmacother. 2014;15:2113–2116.
32. Hanrahan CF, Clouse K, Bassett J, et al. The patient impact of point-of-care vs. laboratory placement of Xpert MTB/RIF. Int J Tuberc Lung Dis. 2015;19:811–816.
Keywords:

Xpert MTB/RIF; diagnostics; clinical impact; treatment outcomes; morbidity

Supplemental Digital Content

Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.