Extra Active Tuberculosis Cases Detected and Costs Incurred by an Additional Approach for Household Contact Investigationin Myanmar

Abstract

Background: Household contact investigation has been recommended as an approach to enhance tuberculosis case detection in high-burden countries. In Myanmar, a conventional method of household contact tracing for active tuberculosis is currently implemented by basic health staff visiting the households of index tuberculosis patients and symptom screening among the household contacts. However, the diagnostic yield of new microbiologically proven cases is very low by using low-sensitivity symptom screening by basic health staff. There is therefore a need to improve the yield of household contact investigation. The objective of the study is to improve the yield by testing a hypothesis that the yield could be improved by extra efforts of the deployed nurses to persuade households to comply with screening procedures and to use chest X-ray and GeneXpert MTB/RIF assay instead of simply assessing signs and symptoms of tuberculosis. Methods: This was a special intervention study where household contact management was done firstly by the national conventional method and later followed by a supplementary intervention. The study was conducted in Pathein township, Ayeyarwady Region between November 2018 and January 2019. In the conventional method, the basic health staff visited households of index tuberculosis patients and assessed their household contacts for signs and symptoms of tuberculosis. Symptomatic contacts were then referred to the tuberculosis centre for sputum microscopy and chest X-ray. Only basic health staff were provided with a travel allowance for home visits. No transportation cost was provided to the contacts. Otherwise, all clinical investigations were provided free of charge. In the supplementary intervention, the index cases were telephoned by programme nurses and asked to bring any remaining household contacts to the centre to receive investigations free of charge with sputum microscopy and GeneXpert MTB/RIF assay performed for those with a positive chest X-ray. The study perspective and costing method for cost-effectiveness was based on the healthcare provider perspective and a micro-costing approach at the point of service delivery including only the direct health and non-health costs. The unit cost for investigation of both methods, based on the household contact, was obtained from the National Tuberculosis Programme. In addition to the travel expenses given to household contacts, the programme nurses, laboratory technicians, X-ray technicians and radiologist were reimbursed for their extra workload during the intervention period. The compliance of the contacts to have screening procedures performed, tuberculosis case detection and cost estimation for diagnostic procedures using both strategies were calculated. Results: A total of 376 household contacts from 128 consenting index cases were identified by the basic health staff. Approximately 55% of the home visits were verified by the index patients in a subsequent interview. Four contacts had symptoms of tuberculosis of which three presented and yielded one bacteriologically confirmed case. Subsequent telephone calls and arrangement by the nurses resulted in 111 index cases bringing 264 household contacts for investigation out of the remaining 373 contacts. Among the remaining contacts, a total of 264 (70.8%) contacts presented at the tuberculosis centre. Approximately 51.2% of the contacts presented at the tuberculosis centre with the invitation of programme nurses and another 19.6% with home visit by basic health staff. This intervention yielded four extra bacteriologically confirmed and additionally 13 clinically diagnosed cases. The yield for tuberculosis disease among the household contacts screened was 0.3% based on the conventional method and an additional 6.4% based on the supplementary intervention. The personnel costs were US$418.56 for the basic health and US$293.12 for the programme nurses. The total investigation cost from the supplementary intervention was US$1,197.69 on top of US$3.03 from the conventional method. The travel cost for household contacts was US$345.84 provided only with the intervention. The total cost using the supplementary intervention was US$2,255.21 on top of US$421.59 using the conventional method. The incremental cost-effectiveness from the intervention was (US$2255.21 - US$421.59 / (18 - 1) = US$107.86 for one extra case detected. Conclusion: This study revealed the very low yield of the conventional method conducted by the basic health staff using symptom screening alone. Additional follow-up telephone calls to households of index tuberculosis cases and subsequent screening with chest X-ray and GeneXpert MTB/Rif assay were cost- effective. This study should be repeated in other areas of Myanmar to determine whether the existing screening method should be changed to the method that we have developed.