Inventory study on completeness of tuberculosis case notifications in Poland in 2018

Abstract

BackgroundEvaluating tuberculosis (TB) notification completeness is important for monitoring TB surveillance systems, while estimating the TB disease burden is crucial for control strategies.AimWe conducted an inventory study to assess TB reporting completeness in Poland in 2018.MethodsUsing a double-pronged inventory approach, we compared notifications of culture-positive TB cases in the National TB Register to records of diagnostic laboratories. We calculated under-reporting both with observed and capture-recapture (CRC)-estimated case numbers. We further compared the notifications by region (i.e. voivodship), sex, and age to aggregated data from hospitalised TB patients, which provided an independent estimate of reporting completeness.ResultsIn 2018, 4,075 culture-positive TB cases were notified in Poland, with 3,789 linked to laboratory records. Laboratories reported further 534 TB patients, of whom 456 were linked to notifications from 2017 or 2019. Thus, 78 (534 - 456) cases were missing in the National TB Register, yielding an observed TB under-reporting of 1.9% (78/(4,075 + 78) × 100). CRC-modelled total number of cases in 2018 was 4,176, corresponding to 2.4% ((4,176 - 4,075)/4,176 × 100) under-reporting. Based on aggregated hospitalisation data from 13 of 16 total voivodeships, under-reporting was 5.1% (3,482/(3,670 - 3,482) × 100), similar in both sexes but varying between voivodeships and age groups.ConclusionsOur results suggest that the surveillance system captures ≥ 90% of estimated TB cases in Poland; thus, the notification rate is a good proxy for the diagnosed TB incidence in Poland. Reporting delays causing discrepancies between data sources could be improved by the planned change from a paper-based to a digital reporting system.

Keywords: Poland; Tuberculosis; double-pronged; inventory study; underreporting.

Figure 1

Flow chart of the two pathways within the double-pronged inventory approach, which is used to estimate under-reporting in the National Tuberculosis Register, Poland, 2018

Figure 2

Results of deterministic record linkage between the laboratory and National TB Register datasets, Poland, 2018 (n = 4,075 TB cases in National TB Register)^a

Figure 3

Comparison of TB case numbers in the hospitals’ inventory and national TB register, stratified by (A) sex, (B) voivodeship and (C) age group, as well as proportions of TB patients in the hospitals’ inventory vs the National TB register (D) by voivodeship, Poland, 2018 (n = 2 datasets)