Dear Editor,
Since its endorsement by the WHO in 2010, Xpert® MTB/RIF (Cepheid Inc, Sunnyvale, CA, USA) has been utilised across the world for TB diagnosis.1 Many high-burden countries adopted a “hub-and-spoke” sample transport and centralized testing model for Xpert scale-up,2 but this may not effectively improve test turnaround times (TAT) and pre-treatment losses to follow-up. Implementing decentralized, onsite Xpert testing in peripheral settings may help address these issues, but the decision to invest in such a strategy will require careful evaluation of the trade-offs in costs, logistical and operational factors compared to the current practice (e.g., centralized testing model).3-5
As part of a pragmatic cluster-randomized implementation trial comparing the effectiveness of decentralized point-of-care vs. centralized hub-and-spoke Xpert testing in Uganda (XPEL TB, clinicaltrials.gov: NCT03044158), we conducted a pre-trial assessment of the per-patient operational cost of on-site smear microscopy followed by specimen transport and centralized Xpert testing at five participating peripheral health centers using a “bottom-up” micro-costing method from the health systems perspective.6 The health centers were purposively sampled based on the size of the corresponding sub-county, distance to the capital city, and volume of patients with TB. We categorized and assessed cost data into six categories: human resources, building, equipment, overheads, supplies and reagents.
Building space and human resource cost allocations were based on a time-and-motion (TAM) study in which we measured the amount and proportion of staff time dedicated to specific TB-related activities. TB health care workers were recruited at each site and, after providing informed consent, were asked to self-report activity-based time estimates to different activity codes similar to methods described here.7 We also collected facility-level cost data by interviewing key health facility staff and reviewing documents to supplement and validate interview responses. At each centralized testing facility, we collected daily logs from hub riders who transported laboratory specimens from peripheral health centers; these logs included the number and type of laboratory samples collected from, and results delivered to, each facility visited. Additionally, we surveyed three of four implementing partners who finance the sample transport networks, regarding resource consumption. Costs per specimen transported were calculated by dividing the average daily cost of the network by the number of samples collected each day. We estimated the cost of Xpert testing in central laboratories from a previously published estimate,8 adjusted for service volume and inflation. All costs were assessed as 2018 US dollars (USD). Cost and price data in local currency and/or reported from earlier years were adjusted using the World Bank estimates for the 2018 exchange rate and the GDP deflator for Uganda.9 We annuitized and depreciated costs for capital assets linearly based on their expected useful lifetime (10 years for furniture and equipment assets, 30 years for buildings) with a 3% annual discount rate.10 We report “volume-weighted” means of costs weighted by the annual volume of tests run for each facility in order to reduce upward bias11. This study was approved by institutional review boards at Makerere University (Kampala, Uganda) and the University of California, San Francisco (San Francisco, CA, USA).
Across all personnel categories, person-time dedicated for TB was less than 20% of the total reported work hours of staff member involved in TB services. On average, it took laboratory staff 15 min (IQR 10–38) to collect sputum from a patient and 20 min (IQR 15–25) per patient to perform sputum smear microscopy (SSM). The weighted average per-patient cost of SSM was US$2.38 (range 1.98–2.88). The largest cost drivers were supplies and human resources with a weighted average unit cost of US$0.86 (range 0.73–1.08, 36% of total cost) and US$1.19 (range 0.83–1.71, 50% of total cost), respectively. On a per activity basis, sputum collection was the most resource-intensive activity associated with SSM, costing on average US$1.24 per patient (range 0.98–1.43). Supplies (primarily masks and sputum cups) contributed 36% of this cost. On average, hub riders collected 26.7 (standard deviation [SD] 19.5) laboratory specimens per day; 3.1 (SD 5.8) samples (12%) were for TB testing. The estimated cost per specimen transported was US$1.34 (SD 2.28). If the sample transport network were to be used only for transporting TB samples, the estimated cost per sample transported rose to US$10.43 (SD 5.55).
For patients who received Xpert testing, the estimated total per-person cost of the diagnostic process was US$25.04 (SD 11.83), which included US$3.72 (SD 2.31) for local SSM testing and transport of specimens and US$21.31 (SD 11.61) for centralized Xpert testing (Figure). Equipment (primarily the Xpert module) comprised 36% of the total cost, and reagents (primarily the Xpert cartridge) accounted for 44%. Specimen transport costs represented 5% of the total cost of TB diagnosis; however, if using a network delivering exclusively TB samples, specimen transport would represent 31% of the total cost of TB diagnosis and increase the total cost by 36%.
Figure.
Per-patient cost of each component of centralized Xpert testing for TB. All cost components sum on a per-patient basis to represent the total cost of diagnosis for the complete pathway. The largest contributor of overall costs is the GeneXpert equipment and cartridge represented by the “equipment” and “reagents” components of the central laboratory procedures, respectively. The total cost of the diagnostic process was US$25.04.
This costing analysis suggests that the cost of centralized hub-and-spoke Xpert testing in Uganda does not differ substantially from published estimates of the cost of point-of-care testing.12,13 Costs of Xpert testing at a centralized laboratory (primarily Xpert modules and cartridges) accounted for >80% of the total cost of TB diagnosis in this study. This analysis further demonstrates that transport of TB-related specimens can be effectively scaled up at a low per-patient cost if a transport network already exists for other clinical specimens.5 However, if a new network has be implemented specifically for TB diagnosis, this would be at additional cost (for such activities as training, coordination and purchasing vehicles), all of which would result in significant increase in cost of TB diagnosis. Such implementation costs may be underestimated in our analysis, where existing infrastructure was already in place. Our cost estimates also suggest that increasing the volume of testing may help to lower the costs of TB diagnosis by reducing the per-specimen costs of Xpert equipment. These dynamics would also affect peripheral testing.5
There are important limitations to this analysis. First, TAM activities were self-reported, which may result in recall and response bias. Second, with a limited sample size, it is difficult to assess how different operational factors and clinical characteristics influence service delivery costs and thus generalize to other settings.14,15 Finally, while we estimated the operational costs for sample transport and SSM, we did not capture the full capital implementation costs of these systems. Nevertheless, these estimates can be useful for policy and decision-makers in high-burden settings seeking to implement the most cost-effective and affordable strategies for TB diagnosis in the local context.
Table.
Mean per-patient cost of each component of centralized Xpert testing for TB. All cost components sum on a per-patient basis to represent the total cost of diagnosis for the complete pathway. The largest contributor of overall costs is the GeneXpert equipment and cartridge represented by the “equipment” and “reagents” components of the central laboratory procedures, respectively. The total cost of the diagnostic process was US$25.04.
| Cost Component |
Health Education (%) |
Sputum Collection (%) |
TB Microscopy (%) |
TB Results (%) |
Treatment Initiation (%) |
Specimen Transport (%) |
Centralized Laboratory Activities (%) |
Total (%) |
|---|---|---|---|---|---|---|---|---|
| Building | $0.03 (16%) | $0.05 (4%) | $0.04 (8%) | $0.02 (10%) | $0.03 (11%) | $ — (0%) | $0.70 (3%) | $0.87 (3%) |
| Equipment | $ — (0%) | $ — (0%) | $0.05 (10%) | $ — (0%) | $ — (0%) | $0.26 (20%) | $8.59 (40%) | $8.90 (36%) |
| Overhead | $0.003 (2%) | $0.01 (<1%) | $0.01 (1%) | $0.003 (1%) | $0.003 (1%) | $0.53 (40%) | $0.03 (<1%) | $0.58 (2%) |
| Staff | $0.15 (82%) | $0.35 (28%) | $0.27 (56%) | $0.18 (89%) | $0.25 (88%) | $0.17 (13%) | $0.34 (2%) | $1.71 (7%) |
| Supplies | $ — (0%) | $0.83 (67%) | $0.06 (12%) | $ — (0%) | $ — (0%) | $0.37 (28%) | $0.67 (3%) | $1.93 (8%) |
| Reagents | $ — (0%) | $ — (0%) | $0.06 (12%) | $ — (0%) | $ — (0%) | $ — (0%) | $10.99 (52%) | $11.05 (44%) |
| Total | $0.18 | $1.24 | $0.49 | $0.20 | $0.28 | $1.33 | $21.32 | $25.04 |
Acknowledgements
The authors would like to thank the health workers involved in TB care who participated in both the overarching XPEL study as well as the time-and-motion activities and patients undergoing TB evaluation at trial sites. The trial is funded by the National Heart, Lung, and Blood Institute, Bethesda, MD, USA, through SIMPLE TB R01 (project number: 5R01HL130192-05).
Footnotes
Conflicts of interest: none declared.
References
- 1.Albert H, et al. Development, roll-out and impact of Xpert MTB/RIF for tuberculosis: what lessons have we learnt and how can we do better? Eur Respir J 2016. July 13; 10.1183/13993003.00543-2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kiyaga C, et al. Uganda’s new national laboratory sample transport system: a successful model for improving access to diagnostic services for early infant HIV diagnosis and other programs. PLoS One 2013; 8(11): 1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Raizada N, et al. Feasibility of decentralised deployment of Xpert MTB/RIF test at lower level of health system in India. PLoS One 2014; 9(2): e89301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Creswell J, et al. Results from early programmatic implementation of Xpert MTB/RIF testing in nine countries. BMC Infect Dis 2014; 14: 2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sohn H, et al. Informing decision-making for universal access to quality tuberculosis diagnosis in India: an economic-epidemiological model. BMC Med 2019; 17(1): 155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Reza TF, et al. Study protocol: a cluster randomized trial to evaluate the effectiveness and implementation of onsite GeneXpert testing at community health centers in Uganda (XPEL-TB). Implement Sci 2020; 15: 1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Tampi RP, et al. Operational characteristics of antiretroviral therapy clinics in Zambia: a time and motion analysis. BMC Health Serv Res 2019; 19(1): 244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Hsiang E, et al. Higher cost of implementing Xpert MTB/RIF in Ugandan peripheral settings: Implications for cost-effectiveness. Int J Tuberc Lung Dis 2016; 20(9): 1212–1218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.The World Bank. World Development Indicators. Washington DC, USA: World Bank, 2019. http://wdi.worldbank.org/table/4.16 [Google Scholar]
- 10.Walker D, Kumaranayake L. Allowing for differential timing in cost analyses: discounting and annualization. Health Policy Plan 2002; 17(1): 112–118. [DOI] [PubMed] [Google Scholar]
- 11.Clarke-Deelder E, Vassall A, Menzies NA. Estimators used in multisite healthcare costing studies in low- and middle-income countries: a systematic review and simulation study. Value Health 2019; 22(10): 1146–1153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Pallas SW, et al. Cost analysis of tuberculosis diagnosis in Cambodia with and without Xpert® MTB/RIF for people living with HIV/AIDS and people with presumptive multidrug-resistant tuberculosis. Appl Health Econ Health Policy 2018; 16(4): 537–548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Pooran A, et al. Point of care Xpert MTB/RIF versus smear microscopy for tuberculosis diagnosis in southern African primary care clinics: a multicentre economic evaluation. Lancet Global Health 2019; 7(6): e798–807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Engel N, et al. Addressing the challenges of diagnostics demand and supply: insights from an online global health discussion platform. BMJ Global Health 2016; 1(4): e000132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mugambi ML, et al. How to implement new diagnostic products in low-resource settings: an end-to-end framework. BMJ Glob Health 2018; 3(6): 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]