Community tuberculosis screening, testing and care, Uganda

Abstract Objective To assess the effectiveness of a community-based tuberculosis and leprosy intervention in which village health teams and health workers conduct door-to-door tuberculosis screening, targeted screenings and contact tracing. Methods We conducted a before-and-after implementation study in Uganda to assess the effectiveness of the community tuberculosis intervention by looking at reach, outputs, adoption and effectiveness of the intervention. Campaign 1 was conducted in March 2022 and campaign 2 in September 2022. We calculated percentages of targets achieved and compared case notification rates during the intervention with corresponding quarters in the previous year. We also assessed the leprosy screening. Findings Over 5 days, campaign 1 screened 1 289 213 people (2.9% of the general population), of whom 179 144 (13.9%) fulfilled the presumptive tuberculosis criteria, and 4043 (2.3%) were diagnosed with bacteriologically-confirmed tuberculosis; 3710 (91.8%) individuals were linked to care. In campaign 2, 5 134 056 people (11.6% of the general population) were screened, detecting 428 444 (8.3%) presumptive tuberculosis patients and 8121 (1.9%) bacteriologically-confirmed tuberculosis patients; 5942 individuals (87.1%) were linked to care. The case notification rate increased from 48.1 to 59.5 per 100 000 population in campaign 1, with a case notification rate ratio of 1.24 (95% confidence interval, CI: 1.22–1.26). In campaign 2, the case notification rate increased from 45.0 to 71.6 per 100 000 population, with a case notification rate ratio of 1.59 (95% CI: 1.56–1.62). Of the 176 patients identified with leprosy, 137 (77.8%) initiated treatment. Conclusion This community tuberculosis screening initiative is effective. However, continuous monitoring and adaptations are needed to overcome context-specific implementation challenges.


Introduction
Tuberculosis remains one of the leading causes of death worldwide. 15][6][7] For example, in Ethiopia a case finding intervention that consisted of advocacy, training, stakeholder engagement and active case finding by health extension workers led to an increase in both case notification and treatment success. 8][11] Uganda is a country with a high tuberculosis burden; its incidence has remained fairly constant over the years at around 198 cases per 100 000 population per year, with a high proportion of missed tuberculosis cases. 12The case notification rate had been declining until 2018, at which point the national tuberculosis programme implemented active case finding strategies.Case notifications subse-quently improved but, during the coronavirus disease 2019 (COVID-19) pandemic, notifications declined by 7.5% (dropping from 65 857 cases in 2019 to 60 887 in 2020). 13,14o address the issue of missed tuberculosis cases in Uganda, in 2022, its national tuberculosis programme embarked on a new country-wide, community-based, active case finding campaign dubbed CAST TB, with the focus on community awareness, screening, testing, prevention, and treatment to end tuberculosis (hereafter called the community tuberculosis campaign or intervention).The aim of this community tuberculosis campaign was to (i) empower communities to confront tuberculosis; (ii) expand service coverage; and (iii) accelerate progress towards the targets of the End TB Strategy. 15The objectives were to improve tuberculosis case identification, management and prevention.
The community tuberculosis campaign in this study also incorporated screening for leprosy.Subsequent rounds which offer tuberculosis, leprosy, HIV, malaria, malnutrition, antenatal care for pregnant mothers, immunization, and COVID-19 services at the community level have been dubbed CAST+.The analysis presented here is limited to the evaluation of campaign 1 and campaign 2 which had a similar scope of diseases (tuberculosis and leprosy).
In this article we present an effectiveness evaluation of two of such community tuberculosis campaigns held in 2022.

Study design
We conducted a quasi-experimental before-and-after study using the RE-AIM framework 16 to assess the reach, outputs, adoption and effectiveness of the community tuberculosis intervention.
We calculated each as the percentage of achievements compared to set targets.We used four outputs.These outputs were the number of people who were: (i) screened during the campaign; (ii) identified as presumptive tuberculosis patients; (iii) identified with bacteriologically-confirmed tuberculosis and had initiated treatment; and (iv) contacts assessed and initiated on tuberculosis preventive therapy.We measured adoption by the number of districts and implementing partners that embraced the community tuberculosis intervention, and their willingness to integrate it into routine case-finding practices across the country.
To determine effectiveness (impact), we conducted a cross-sectional analysis of bacteriologically-confirmed tuberculosis case notification before (control) and after the community tuberculosis intervention was implemented.For campaign 1 in March 2022, we compared tuberculosis case notifications in the first and second quarters of 2022 with those from the corresponding quarters in 2021 (control).For campaign 2 (September 2022), we compared tuberculosis case notifications in the third and fourth quarters of 2022 with those from the corresponding quarters in 2021 (control).This approach accounts for delayed notifications of tuberculosis cases identified during the intervention month at the end of each quarter, which may only be officially reported in the subsequent quarter.

Setting
Administratively, Uganda's 146 districts are divided into 15 health regions.There is regional variation in tuberculosis incidence and notification; the highest case notification rates are in the Karamoja region. 14ganda's decentralized health system is structured with national and regional referral hospitals at the top.These hospitals support general (district) hospitals and health centres (ranging from levels II to IV) and, at the community level, village health teams.Around 1700 of the 6937 public and private facilities are tuberculosis diagnostic and treatment units.

Intervention description
The community tuberculosis campaign is a biannual, national, communitybased, active tuberculosis case-finding initiative which is implemented in March and September of each year.Key components include community mobilization and sensitization; voluntary symptom screening for tuberculosis and testing of presumed individuals for tuberculosis using GeneXpert® (Cepheid, California, United States of America); and linkage to care for those with a confirmed tuberculosis diagnosis.Contacts of peopled diagnosed with tuberculosis are identified and assessed for preventive tuberculosis therapy.
National and district-level planning meetings were held before any activity on the intervention, while mass media was used to raise community awareness.
The tuberculosis programme held a one-day meeting before campaign implementation to orientate village health teams and primary health workers on the package.The programme then assigned each village with a village health team member to operate within their own communities; on average, each member was responsible for visiting 7-15 households per day.
At the start of implementation of the intervention, village health teams introduced themselves and the campaign to household members and requested verbal consent to conduct tuberculosis sensitization and screening.They were equipped with an educational flipchart and one-pager flyers containing basic tuberculosis and leprosy messages, translated into local languages.Every household visited received a flyer.
Village health teams used symptom screening to identify presumptive tuberculosis patients regardless of human immunodeficiency virus (HIV) status: persons with a cough for two weeks or more, or persistent fever, or excessive night sweats or weight loss were identified as presumptive tuberculosis patients.
Village health teams were associated with a tuberculosis diagnostic and treatment unit (hereafter also referred to as "facility"); health workers from these facilities performed contact investigations, targeted screenings in hotspots, tuberculosis diagnosis and treatment initiation.
For contact investigation, health workers first identified index tuberculosis patients whose contacts had not yet been evaluated.They then scheduled appointments with contacts and assessed them for tuberculosis and, where relevant, eligibility for tuberculosis preventive therapy.
Health workers used their facility data to assign hotspots (i.e.areas with the most people with tuberculosis) for targeted community screening.Health workers then coordinated with community leaders for convenient screening times and locations.
All sputum samples collected during screening were analysed using GeneXpert®.Information on confirmed tuberculosis patients was then sent to the facility by a phone call or short message service; the facility in turn relayed the information to the village health teams to link confirmed individuals to treatment at the nearest facility.
Village health team members were given a daily basic allowance of 10 000 Ugandan shillings (around 2.7 United States dollars, [US$]); each health worker received a daily allowance of 12 000 Ugandan shillings (around US$ 3.2) and a transport refund of 30 000 Ugandan shillings (around US$ 7.9).
The community campaign also incorporated screening for leprosy.The village health teams were sensitized on identifying leprosy lesions, and individuals presumed to have leprosy were referred to the health facilities for appropriate evaluation and treatment.

Implementation process
Campaign 1 and campaign 2 were implemented for 5 days in March and September 2022, respectively.Regions and districts were at liberty to choose a week during March and September 2022 to implement the campaign.
The community tuberculosis intervention was implemented in seven key steps (Fig. 1).Throughout the campaign, national, district and implementing partner staff provided technical assistance to ensure quality assurance for both facility-and community-le vel inter vent ions.Participants did not incur any costs.

Variables
The primary outcome of our analysis was the number of tuberculosis cases notified.This was defined as the number of bacteriologically-confirmed tuberculosis patients diagnosed and notified within the national surveillance system in a quarter.

Data sources
During the campaign, village health teams used printed forms of the Community TB Sensitization and Screening Register to capture basic information; data were then manually transferred to facility registers.Health workers tallied the specified indicators used to compile facility-level summary reports.Reports were then submitted to the district tuberculosis and leprosy supervisor, verified, and entered into an online system from DEFEAT TB (Kampala, Uganda), by health information assistants and the district biostatistician.The platform has been set up for the tuberculosis programme and partners to monitor and coordinate efforts directed towards defeating tuberculosis in Uganda.The quarterly numbers of tuberculosis patients notified were obtained using District Health Information Software 2 (University of Oslo, Oslo, Norway).We obtained outputs from campaign records such as attendance lists.

Statistical analysis
We summarized the four outputs of each campaign both nationally and by Uganda's 15 health regions.
We compared the total number of bacteriologically-confirmed tuberculosis patients notified and the case notification rate during the eligible community tuberculosis campaign quarters of the previous year (before the campaign) with that of the corresponding eligible quarters during the campaign.Tallies of specified tuberculosis and leprosy indicators from the online system are presented in text, tables and figures.
We used the Stata csi command (StataCorp, College Station, USA) to estimate the case notification rate and case notification rate ratio from the number of tuberculosis cases notified in each period based on projected populations.
We also compared tuberculosis notifications during the quarters that the intervention took place in participating districts with notifications for corre-sponding quarters in non-participating (control) districts (further information is available in the authors' online repository). 17Again, we used the csi command to estimate the case notification rates and rate ratios.We present our results according to the reporting guidelines for implementation and operational research. 18

Ethical considerations
Ethical approval for the study was obtained from Mengo Hospital Research Ethics Committee (MH-2023-30).Screening, testing and treatment for tuberculosis and the provision of tuberculosis preventive therapy to community members was voluntary.

Results
We present the targets and achievements of the two campaign cycles in Table 1.In campaign 1, 111 out of 146 districts (76%) conducted district planning and orientation meetings.A total of 643 of all 1700 diagnostic and treatment units in the country (37.8%) trained and orientated village health teams.Campaign 1 screened a total of 1 289 213/44 212 800 people (2.9% of the Ugandan population), while campaign 2 screened 5 134 056 people (11.6% of the Ugandan population).
Of the 1 289 213 screened in campaign 1, a total of 179 144 (13.9%) were identified as presumptive tuberculosis patients.The proportion of those screened that were identified as presumptive tuberculosis patients differed by region; this proportion was highest in Bukedi (2703 out of 6955; 38.9%) and lowest in the West Nile region (7161 out of 106 216; 6.7%).Of all the presumptive tuberculosis cases in campaign 1, 4043 people (2.3%) were diagnosed with bacteriologically-confirmed tuberculosis.By region, Bunyoro had the highest proportion (451 out of 4821; 9.4%) of presumptive tuberculosis patients diagnosed with tuberculosis.Of the total of 4043 people whose diagnosis of tuberculosis was confirmed, 3710 (91.8%) initiated treatment.Five regions had treatment initiation rates below 90% (Ankole, Karamoja, Kigezi, Teso and West Nile) with losses-to-followup before treatment that ranged from 17.4% to 38.0% (100% minus percentage initiating treatment), while three regions (Bunyoro, North Central and Tooro) reported higher proportions of individuals   2).In campaign 2, all 146 districts (100%) and 1608 of 1700 (94.6%) diagnostic and treatment units conducted relevant orientation meetings.A total of 71 564 village health teams participated in campaign 2, screening an average of 30 households per village health team (Table 1).
Out of the 5 134 056 people screened in campaign 2, 428 444 (8.3%) were identified as presumptive tuberculosis patients and of these, 7144 (1.7%) were confirmed with tuberculosis (Table 2).By region, the proportion of presumptive tuberculosis patients among the screened population ranged from 3.3% to 18.7%.Of those diagnosed with tuberculosis, 5942 individuals (83.2%) were started on treatment.The treatment initiation rates in different regions ranged from 57.3% to 99.5%.During campaign 2, 23 547 people received tuberculosis preventive therapy.
Compared to hotspot screening and contact tracing, more cases were identified from the door-to-door screening (3410 for campaign 1 and 5969 cases for campaign 2), with village health teams visiting 615 438 households in campaign 1, and 2 174 595 households in campaign 2.
Campaign 1 identified 39 patients with leprosy of whom 22 (56.4%)initiated treatment; while campaign 2 identified 137 patients of whom 115 (83.9%) initiated treatment (further information is available in the authors' online repository). 17

Intervention impact
In Fig. 2, we show maps of case notification rates by region before and during community tuberculosis campaign 1 and campaign 2.
In our before-and-after analysis, improvements in case notification rates with campaign implementation were observed for both campaign cycles 1 and 2. For campaign 1, the case notification rate increased from 48.1 per 100 000 population before the campaign to 59.5 per 100 000 during the campaign (Table 3); the case notification rate ratio was 1.24 (95% CI: 1.22-1.26).For campaign 2, the case notification rate increased from 45.0 per 100 000 population to 71.6 per 100 000; the case notification rate ratio was 1.59 (95% CI: 1.56-1.62).
In an inter-district comparison for campaign 1, districts that implemented the intervention had a higher case notification rate than districts that did not implement a campaign, at 51.6 and 49.0 per 100 000 population, respectively. 17he case notification rate ratio was 1.05 (95% CI: 1.01-1.10).

Discussion
We set out to evaluate the impact of expansive community tuberculosis screening in two community tuberculosis campaigns.Both campaigns successfully reached a substantial percentage of the targeted populations.Overall, both campaigns identified a notable proportion of individuals with presumed tuberculosis (13.9% and 8.3%, respectively), although there were some regional variations; in the absence of the intervention, such community members might not have sought care for their symptoms.Campaign 1 and 2 also resulted in increased case notification rates for bacteriologically-confirmed tuberculosis.
According to current World Health Organization guidelines, systematic screening for tuberculosis at the population level is recommended in areas where the tuberculosis prevalence is at least 5000 cases per 100 000 population. 19As it allows for intermittent community screening, the community tuberculosis campaign used in this study offers an alternative approach for high-burden countries that do not meet the threshold for systematic screening.
The community tuberculosis campaign described here was largely  20 The community campaign in this study used symptom screening during door-to-door visits, hotspot screening and contact tracing; the diagnostic yields observed underscore the large pool of undiagnosed tuberculosis within the communities. 21The slightly lower yield observed during campaign 2 could partly be attributed to campaign 1 having treated some cases of tuberculosis in the community. 20A previous systematic review of diagnostic yields from active case finding strategies showed variability in yields that were dependent on the population and tuberculosis screening and diagnostic approach. 22Although the presumptive and diagnostic yields were generally considerable, the substantial variation across the regions highlights the need for programme improvement by implementing tailored interventions, including refining the screening methods in regions with low presumptive and diagnostic yields.
The campaigns very likely contributed to the notable increase in the case notification rate for bacteriologicallyconfirmed tuberculosis because of the intervention.This observed increase in case notification rate contrasts with previous evidence suggesting that community-based tuberculosis screening did not consistently increase the case notification rate. 11That finding, however, could have been a consequence of using smear microscopy and chest radiographs for tuberculosis diagnosis.In another campaign, with no observed impact on the case notification rate, researchers attributed findings to rapid decline in tuberculosis burden. 23 major challenge that we experienced during implementation of the community tuberculosis intervention was delay in the initiation of tuberculosis treatment; this delay resulted in high levels of pre-treatment loss-to-followup, particularly during campaign 2. These pre-treatment losses are a concern in all community tuberculosis case finding interventions as highlighted by previous studies. 24,25Integrating patientcentred care approaches into community tuberculosis case finding interventions is vital to mitigate various provider and patient factors that influence losses. 26ur study has some limitations.First, we did not assess the treatment outcomes of the tuberculosis patients diagnosed during the intervention.Second, we also did not assess the economic and social benefits of the community tuberculosis intervention to families and households, and did not assess the impact of the intervention on tuberculosis epidemiology in communities.Third, analysis of data by age group and gender was not performed as data were not collected in a case-based format; this limited extended disaggregated analysis.We recommend that future studies prioritize these aspects for investigation.Further studies will evaluate the feasibility and cost-effectiveness of the intervention.
In conclusion, our study highlights the potential of community-based tuberculosis case finding campaigns in accelerating progress towards tar-  This is an open access article distributed under the terms of the Creative Commons Attribution IGO License (http://creativecommons.org/licenses/by/3.0/igo/legalcode),which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.In any reproduction of this article there should not be any suggestion that WHO or this article endorse any specific organization or products.The use of the WHO logo is not permitted.This notice should be preserved along with the article' s original URL.