Implementation of GeneXpert for TB Testing in Low- and Middle-Income Countries: A Systematic Review

Scott Brown; Justine E. Leavy; Jonine Jancey

doi:10.9745/GHSP-D-21-00121

Figures & Tables

Figures

Tables

Download figure
Open in new tab
Download powerpoint
FIGURE
PRISMA Flow Chart for Study Selection on Implementation of GeneXpert for TB Testing in Low- and Middle-Income Countries
Abbreviation: PRISMA, Preferred Reporting Items for Systematic reviews and Meta-Analyses.

Tables

Figures

View popup

TABLE 1.

Characteristics of Studies on Implementation of Xpert for TB Testing in Low- and Middle-Income Countries

Author (Year)	Country	Implementation Approach	Health Setting	Testing Algorithm	Public Health Impact
Cattamanchi et al.¹⁵ (2020)	Uganda	Hub-and-spoke model: Testing ‘hubs’ linked to 3–5 microscopy unit “spokes” Monitoring of results centralized through National TB Reference Laboratory.	Regional or district public hospital for testing, with community health center for sample collection	Any person with presumed TB.	Nearly 4-fold increase in confirmed MDR-TB from 2009–2017 Increase in TB CNR from approximately 41,000 cases pre-2010 to 57,756 cases in 2017
Cowan et al.²⁶ (2015)	Mozambique	Hub-and-spoke model:Xpert installed in 4 public hospitals in 4 districts. Transportation network established from select health centers to transfer samples for testing.	District and urban public hospital for testing. Urban and remote health centers for sample collection	Two-step algorithm for people suspected of having pulmonary TB. GeneXpert testing occurs after 2 separate smear-negative results using smear microscopy.	Increase in diagnosis of bacteriologically confirmed pulmonary TB by 69%
Sikhondze et al.²⁷ (2015)	Swaziland	Hub-and-spoke model:23 Xperts installed in 19 TB diagnostic laboratories. Community health sample transportation covers 78% of country. NGO covers the remaining regions.	TB diagnostic laboratories for testing	Not stated.	Not stated.
Nalugwa et al.²⁸ (2020)	Uganda	Hub-and-spoke model.249 Xpert machines in 227 of 1500 TB diagnostic units. Motorcycle riders employed by Central Public Health Laboratories transport samples from community health centers.	Regional or district public hospital testing hub, with community health center for sample collection	At time of study, Xpert testing available to PLHIV, health care workers, contacts of DR-TB, pregnant women or breastfeeding mothers, prisoners, patients from refugee camps, and diabetics.	Not stated.
Newtonraj et al.²⁹ (2019)	India	Hub-and-spoke model:Xpert installed at the Intermediate Reference Laboratory in a government hospital for chest diseases in Puducherry district. Samples are received from 27 designated microscopy centers in medical colleges or district-level hospitals.	Centralized testing in Intermediate Reference Laboratory within government hospital for chest diseases. Sample collection from district hospitals.	Initial diagnostic for extrapulmonary, pediatric, and HIV-associated TB. Xpert is also an add-on test for sputum microscopy negative patients.	CNR reduced from 118 to 97 per 100,000 population between 2010 and 2017
Rendell et al.³⁰ (2017)	Mongolia	Hub-and-spoke model:3 Xpert machines installed across the country. Samples collected at community TB clinics and results returned using several paper-based delivery options.	Testing available at the National TB Reference Laboratory, the Regional Diagnostic and Treatment Centre, and a northern province hospital. Samples collected at community/district level TB clinics.	All smear-negative pulmonary TB cases, patients with presumed TB diagnosed with HIV, patients with presumed MDR-TB, and all smear-negative new cases aged 15–35 years	Number of diagnosed cases increased from 2,783 in 2012 to 3,209 in 2015
Gidado et al.³¹ (2018)	Nigeria	Point-of-care model: 176 Xperts installed at clinics that meet necessary installation requirements. Test results monitored centrally, as well as the procurement of supplies.	Primary, secondary, and tertiary facilities	Not stated.	Not stated.
Hoang et al.³² (2015)	Vietnam	Point-of-care model: Xpert installed in TB units of district health center in 35/63 provinces. Provinces chosen based on known prevalence of MDR-TB and/or HIV.	TB units in district health centers	Presumptive MDR-TB cases, defined as belonging to a risk category including TB treatment non-converters; contact of a person with MDR-TB; person coinfected with TB/HIV; >1 month using TB drugs.	37.8% of estimated presumptive MDR-TB patients tested 75% of identified MDR-TB patients completed treatment and cured.
Joshi et al.³³ (2018)	Nepal	Point-of-care model: Xpert installed in 26 health facilities under TB Reach Project and operated by either government or NGO. In 2014, all machines donated to government. Samples collected from patients for smear microscopy, and an additional sample collected for Xpert testing, where available.	Government health facilities such as District Public Health Office laboratory, hospital and primary health centers located throughout the country	Targeted to specific populations as per WHO recommendations, including children aged younger than 15 years, PLHIV, severe forms of TB, and in presumptive MDR-TB.	Xpert diagnosed 28% of the total bacteriologically confirmed TB cases in 2015/2016.
Mustapha et al.³⁴ (2015)	Nigeria	Point-of-care model: Xpert implemented at 22 sites by NGO in partnership with government. Governance oversight by the National TB Control Program in the form of an advisory committee.	10 secondary health facilities, 10 tertiary hospitals, 2 private health facilities	Targeted to specific risk groups, including PLHIV with presumptive TB, those with poor response/relapse to TB treatment, contact of known MDR-TB case, TB cases at risk of resistance.	Not stated.
Awan et al.³⁵ (2018)	Pakistan	Two-pronged point-of-care approach: A “private-public mix model” with an Xpert installed at the TB lab of 6 public hospitals and 1 private site participating in Programmatic Management of Drug-Resistant TB. Active case finding occurred among outpatients and in wards of hospitals. The second “social business model” introduced Xpert at 3 TB centers for testing, with community screeners identifying symptomatic patients from nearby private-sector clinics and referring them to TB clinics.	Public and private hospitals and private community clinics	Initial diagnostic for people with presumptive TB.	43% increase in diagnosed DR-TB 83.2% of TB cases found in the public-private mix model.

Abbreviations: CNR, case notification rate; DR-TB, drug-resistant TB; MDR-TB, multi-drug resistant TB; NGO, nongovernmental organization; PLHIV, people living with HIV.

View popup

TABLE 2.

Identified Barriers to Implementing Xpert TB Testing in Studies in Low- and Middle-Income Countries

Barriers	India	Mongolia	Mozambique	Nepal	Nigeria	Pakistan	Swaziland	Uganda	Vietnam
Patient-level factors
Distance to testing sites²⁹^,³³^,³⁵	X			X		X
Cost of testing in private health clinics³⁵						X
Human resources
Inadequate/inconsistent staff training on testing processes/guidelines, and/or limited awareness of availability²⁶^,²⁸^–³⁰^,³²^,³⁴^,³⁵	X	X	X		X	X		X	X
Low self-efficacy and confidence that Xpert improves outcomes¹⁵								X
Workload capacity in laboratories¹⁵^,³⁰		X						X
High staff turnover¹⁵^,³¹^,³²					X			X	X
Initial struggle with English software (since rectified)²⁶			X
Material resources
Inadequate power supply¹⁵^,²⁶^,²⁷^,³¹^,³³^,³⁵			X	X	X	X	X	X
Poorly equipped labs (e.g., limited space for patient assessment; no ventilation, workbench, air conditioning, and/or refrigerator)¹⁵^,²⁶^,³⁴			X		X			X
Inappropriate storage of cartridges³¹					X
Service implementation
Geographically dispersed TB laboratories³⁵						X
Transportation of sputum samples (e.g., inconsistent/delays in availability of deliveries; improper packaging/temperature control during transport)¹⁵^,³²^,³⁴^,³⁵					X	X		X	X
Inability to track/follow up with patients testing positive¹⁵								X
Determining an appropriate testing algorithm²⁶			X
Failure to identify eligible cases for screening³⁰^,³²		X							X
Limitations to accessing updated and clear standard operating procedures/internal audits²⁷^,³²							X		X
Poor quality samples collected²⁷^,²⁹^,³⁰^,³²	X	X					X		X
Service coordination
Supply chain for procurement of cartridges, reagents, and/or medicines resulting in lack of supplies¹⁵^,²⁶^,³⁰^,³²^–³⁵		X	X	X	X	X		X	X
Lack of referral pathways/communication between staff and health centers (e.g., referral pathways and transfer of results)²⁷^–³⁰^,³²^–³⁵	X	X		X	X	X	X	X	X
Insufficient oversight from national body/remote monitoring¹⁵^,²⁶			X						X
Delays in notification of results¹⁵^,²⁶^,²⁸^–³⁰^,³⁴	X	X	X		X			X
Limited ability to track positive cases and confirm treatment²⁶^,²⁸^,³⁵			X			X		X
Technical operations
Xpert maintenance (e.g., frequency of maintenance not always implemented as required; poor understanding of routine maintenance in dusty, non-temperature-controlled labs)²⁶^–²⁸^,³¹^,³³^,³⁵			X	X	X	X	X	X
Failure of calibration and required replacement²⁶			X
Lack of timely replacement of damaged modules³³				X
Module malfunction²⁸^,³¹					X			X
Limited internet connectivity²⁶^,³¹			X		X
Local repair options limited³⁰		X

View popup

TABLE 3.

Identified Enablers of Implementation of Xpert TB Testing Studies in Low- and Middle-Income Countries

Enablers	Country
Daily transport of samples¹⁵	Uganda
SMS communication of results to health centers¹⁵	Uganda
Collecting monthly performance feedback from staff for quality improvement purposes¹⁵	Uganda
Clear guidelines in local language³⁰	Mongolia
Purchase of uninterruptable power supply³⁰	Mongolia
Access to external experts³⁰	Mongolia
Peer learning for professional development³⁰	Mongolia
Consistent process of confirming of results between referring site and laboratory after sample sent³²	Vietnam
Laboratory personnel understood Xpert to be superior to smear microscopy³³	Nepal
Active case finding approach³⁵	Pakistan
Expanded diagnostic algorithm³⁵	Pakistan
Additional human resources³⁰^,³⁵	Pakistan, Mongolia
Close collaboration³⁵	Pakistan
Supervisory visits to improve maintenance and stock procurement³⁵	Pakistan