The yield of tuberculosis contact investigation in low- and middle-income settings: a systematic review and meta-analysis

Abstract

Background: Contact investigation, the systematic evaluation of individuals in close contact with an infectious tuberculosis (TB) patient, is a key active case-finding strategy for global TB control. Better estimates of the yield of contact investigation can guide strategies to reduce the number of underreported and underdiagnosed TB cases, approximately three million cases per year globally. A systematic review (Prospero ID # CRD42019133380) and meta-analysis was conducted to update and enhance the estimates of the yield of TB contact investigation in low- and middle-income countries (LMIC). Pubmed, Web of Science, Embase and the WHO Global Index Medicus were searched for peer-reviewed studies (published between January 2006-April 2019); studies reporting the number of active TB or latent tuberculosis infection (LTBI) found through contact investigation were included. Pooled data were meta-analyzed using a random effects model and risk of bias was assessed.

Results: Of 1,644 unique citations obtained from database searches, 110 studies met eligibility criteria for descriptive data synthesis and 95 for meta-analysis. The pooled yields of contact investigation activities for different outcomes were: secondary cases of all active TB (defined as those bacteriologically confirmed or clinically diagnosed) 2.87% (2.61-3.14, I² 97.79%), bacteriologically confirmed active TB 2.04% (1.77-2.31, I² 98.06%), and LTBI 43.83% (38.11-49.55, I² 99.36%). Yields are interpreted as the percent of contacts screened who are diagnosed with active TB as a result of TB contact investigation activities. Pooled estimates were substantially heterogenous (I² ≥ 75%).

Conclusions: This study provides methodologically rigorous and up-to-date estimates for the yield of TB contact investigation activities in low- and middle-income countries (LMIC). While the data are heterogenous, these findings can inform strategic and programmatic planning for scale up of TB contact investigation activities.

Keywords: Contact investigation; Mycobacterium tuberculosis; Systematic Review; Tuberculosis.

Fig. 1

Flow diagram for study selection

Fig. 2

Distribution of studies per year based on the start date of data collection (n = 110). *: there was no information on the start date of data collection, but the publication of these studies was before September 2012

Fig. 3

Overall results of the risk of bias assessment (n = 110)

Fig. 4

Forest plot of the yield of contact investigations for all active tuberculosis (confirmed and clinically/ radiologically diagnosed)

Fig. 5

Forest plot of the yield of contact investigations for confirmed tuberculosis

Fig. 6

Forest plot of the yield of contact investigations for LTBI