key: cord-0291440-iyd4mo5h
authors: Godwin-Akpan, T. G.; Chowdhury, S.; Rogers, E. J.; Kollie, K. K.; Zaizay, F. Z.; Wickenden, A.; Zawolo, G. V. K.; Parker, C. B. M. C.; Dean, L.
title: Recommendations for an Optimal Model of integrated case detection, referral, and confirmation of Neglected Tropical Diseases: A case study in Bong County, Liberia
date: 2022-02-02
journal: nan
DOI: 10.1101/2022.02.01.22269709
sha: f158bc6614761d5495963df6ca24dd1bcdc09d03
doc_id: 291440
cord_uid: iyd4mo5h

Background: People affected by Neglected Tropical Diseases (NTDs), specifically leprosy, Buruli ulcer (BU), yaws, and lymphatic filariasis, experience significant delays in accessing health services, often leading to catastrophic physical, psychosocial, and economic consequences. Global health actors have recognized that Sustainable Development Goal 3:3 is only achievable through an integrated inter and intra-sectoral response. This study evaluated existing case detection and referral approaches in Liberia, utilizing the findings to develop and test an Optimal Model for integrated community-based case detection, referral, and confirmation. Finally, this study evaluates the efficacy of implementing the Optimal Model in improving the early diagnosis of NTDs. Methodology/Principal Findings: The study used mixed methods, including key informant interviews, focus group discussions, participant observation, quantitative analysis, and reflexive sessions to evaluate the implementation of an Optimal Model developed through this study. The quantitative results from the testing of the optimal model are of limited utility. The annual number of cases detected increased in the twelve months of implementation in 2020 compared to 2019 (pre-intervention) but will require observation over a more extended period to be of significance. Qualitative data revealed essential factors that impact the effectiveness of integrated case detection. Data emphasized the gendered dynamics in communities that shape the case identification process, such as men and women preferring to see health workers of the same gender. Furthermore, the qualitative data showed an increase in knowledge of the transmission, signs, symptoms, and management options amongst CHW, which enabled them to dispel misconceptions and stigma associated with NTDs. Conclusion/Significance: This study demonstrates the opportunity for greater integration in training, case detection, referral, and confirmations. However, the effectiveness of this approach depends on a high level of collaboration, joint planning, and implementation embedded within existing health systems infrastructure. Together, these approaches improve access to health services for NTDs.

Liberia was one of the first countries to adopt an integrated approach to the case management of 111 (IDM) NTDs to improve access and universal health coverage for people with NTDs. This 112 approach responded to lessons learned from the Ebola Virus Disease (EVD) outbreak and two 113 national policies: "Essential Package of Health Services" and "Investment Plan for building a 114 resilient health system," which were developed to build resilient health system (19, 20) . Maryland, and Nimba, alongside the integration of national-level activities including budgeting, 131 performance monitoring, data management, and drug supply chain. The NTD Program 132 implemented multiple and diverse community-based case detection models in the pilot counties 133 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. ; https://doi.org/10.1101/2022.02.01.22269709 doi: medRxiv preprint to increase early case detection, as shown in Fig 1. However, the effectiveness of these differing 134 models in increasing patient access to care and early diagnosis had not been documented. This study aimed to evaluate existing approaches to case detection and referral in Liberia using 140 an action research cycle, use the findings to develop an optimal model for integrated community-141 based case detection, referral, and confirmation. Finally, this study evaluates the efficacy of the 142 optimal model in improving early diagnosis NTDs. This paper presents findings from the 143 formative research phase, describes the optimal model produced, and the key learnings from its 144 implementation over twelve months. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. ; https://doi.org/10.1101/2022.02.01.22269709 doi: medRxiv preprint School of Tropical Medicine, University of Liberia Pacific Institute for Research and Evaluation) 151 through COR-NTD funding. There was a collaboration between the national program, lower 152 levels of the health system, and persons affected by NTDs during the entire development 153 process. The model was developed in two phases over nine months: phase one, the formative 154 phase, and phase two, the planning phase. During phase one, we conducted a participatory 155 qualitative evaluation of the program's three models of case detection to identify strengths and 156 weaknesses. We used the results to design an Optimal Model for case detection. This phase drew 157 on qualitative participatory methods, including key informant interviews and focus group 158 discussions with health workers and affected persons on understanding their experiences of 159 existing case detection approaches. During phase two, the planning phase, a two-day workshop 160 was held in Monrovia, Liberia, with stakeholders including persons affected by NTDs, 161 community health workers (CHWs), Bong County Health team, NTD program, and other health 162 stakeholders from the central, county, and district levels to collaboratively develop the optimal 163 model based on the findings from phase one. The findings included vital themes such as 164 inconsistencies and delays in remuneration, geographical barriers, opportunity costs throughout 165 pocket expenditure of CHWs on transport and calls, and insufficient training and supervision of 166 DMDI services. Challenges regarding gender, disability, and other axes of social disadvantage 167 patients face were also explored to understand how they affect case detection; poverty, social 168 isolation, stigma, and gendered access were highlighted as critical issues. The optimal model 169 included comprehensive training on integrated approaches to identify, refer, diagnose and 170 manage NTDs at the community level. The training included a referral process, supervision 171 structure, and incentive packages. The Optimal Model developed is summarized in Fig 2 and  172 detailed in S1 File. 173 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 2, 2022. Bong County was purposively selected as the study setting for this research based on the 189 following criteria: Bong County is one of the five counties of the integrated CM-NTDs pilot; co-190 endemic for all priority diseases; has mixed success in terms of case reporting across its nine 191 health districts (i.e., some have identified more cases than others); and all three models of case 192 detection (Fig 1) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 2, 2022.

The study uses a Community-Based Participatory Research (CBPR) approach comprising phases 201 depicted in Fig 4. The CBPR approach promotes the equitable participation of community 202 members, researchers, and all stakeholders throughout the research process (24) . During phases 203 1-2 (nine months period), the optimal model was developed (as described above) and 204 implemented during phase 3 (one-year period). The findings presented in this study were 205 identified in phase four (four months period), the observation and reflection phase, based on the 206 activities in phases 1,2, and 3. Phase four used a mixed-methods approach to evaluate the 207 implementation of the optimal model, which is both a process and outcome evaluation. 208 209 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 2, 2022. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. Four health facilities in Bong were purposively selected based on the geographical variation. A 227 CHSS trained to supervise CHW in the community in each facility was interviewed as a key 228 informant, using an interview guide. We were particularly interested in their experiences and 229 perspectives of training and supervision, specifically, how the optimal model has impacted case 230 identification, referral, and confirmation at different health system levels. 231 232 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. Reflexive sessions involving a participatory group discussion to reflect on the optimal model's 244 design, delivery, and evaluation were conducted with the core research team and the County 245 Health Team. The core research team remotely participated in a reflexive session over Skype. In 246 contrast, a reflexive session with the county health team was conducted at the health facility, 247 including the NTD focal person and the Community Health focal person in Bong County. The 248 reflexive session included key themes on collaboration, participation and power, the use of 249 evidence, and reflections on the challenges, strengths, and sustainability of the optimal model. 250

Data Analysis 251 All interviews were digitally recorded and transcribed in English. All data were anonymized and 252 stored securely in a password-protected database. Debriefing sessions with data collectors 253 focused on key findings, identifying saturation themes, and refining inquiry lines. Data were 254 analyzed thematically using a framework approach which involved the research team reading 255 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. ; https://doi.org/10.1101/2022.02.01.22269709 doi: medRxiv preprint and re-reading the transcribed data. A coding framework was developed, transcripts were coded 256 using NVivo software, with emerging themes discussed and coding refined. The coded 257 transcripts were charted and summarized in narratives for each theme. 258

Quantitative Methods 259 During the optimal model pilot test period (January-December 2020), pre-existing and routine 260

NTDs program data reporting tools were used to collect quantitative data on the utilization of 261 NTDs services. Existing health systems data was disaggregated by gender and other socio-262 demographic factors and utilized to collate information about the impact of the optimal model. 263

Descriptive comparisons are made between the optimal model intervention period (2020) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. The length of training was a key challenge. One day for training was too short, and many CHWs 307 requested longer training sessions or refresher training. Within this pilot, the duration of the 308 COR-NTD training deviated from the original optimal model, which was initially planned to last 309 for two days due to budget constraints. As a result, participants reported feeling rushed and 310 needed more time on the referral forms to practice filling them out correctly. 311 "The training is for…it's not supposed to be for twenty-four hours. It is 312 supposed to be for forty-eight hours or seventy-two hours because the notes' 313 training is so bulky. You can't use one day to put everything in your brain. You 314 have to take some time to study. Even though they had pretest and posttest. It 315 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. However, the training sensitized case detectors to consider gender preferences by involving 332 relatives or friends during case detection; many CHWs described that this resulted in affected 333 persons being more likely to cooperate. 334 'Though the person will find it very difficult to express their condition to you as 335 an opposite gender because traditionally they do not feel secure. 336 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. 

being getting in recent time, as soon I call the county focal person if that case 379 specimen supposed to be collected I get in touch. He will try to connect and try 380 to collect the specimen. As soon that case is confirmed, the drugs will 381 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. really become a problem in the facility. When the CHV report the cases, the focal person will 399 call you and say the case is confirmed; you will give your feedback to the CHV in the community. 400

They expect their five-dollar at that time but at time it will take long time before the five dollar. 401

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. increased compared to the previous year (Fig 5) . This was particularly noticeable for hydrocele 405 and suspected Buruli ulcer cases, with more than twice the number of cases reported in 2020 406 compared to 2019 (Fig 6) . The Optimal Model also shows an increase in the number of health facilities throughout the 423 county that reported cases (Fig 8) . However, many patients reported some level of movement 424 limitations at diagnosis due to ulceration. reported CM-NTDs cases for the first time in 2020 due to the Optimal Model (Fig 8) . 433

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. ; https://doi.org/10.1101/2022.02.01.22269709 doi: medRxiv preprint Impact of gender on equitable accessa higher proportion of male cases than females was 438 identified across diseases (Fig 9) . Five hundred ninety-five males and 109 females participated in 439 the case detection and referral training from the nine health districts in Bong County (Table 2) . 440

The CHW training included both males and females between the ages of 25-60 from catchment 441 communities of all the health facilities in Bong County. However, significantly more men than 442 women were in this training, as reflected in the cases detected (Fig 9) . A more significant 443 proportion of cases identified as males is shared across all disease conditions across 2019 and 444 2020, excluding leprosy, where case detection appears equal (Fig 9) . 445 446 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. ; https://doi.org/10.1101/2022.02.01.22269709 doi: medRxiv preprint CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. Gendered access to patients to identify and refer was a key challenge emphasized by 500 qualitative and quantitative data. The gender disparity in the number of cases diagnosed was 501 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. ; https://doi.org/10.1101/2022.02.01.22269709 doi: medRxiv preprint linked to the sex of those who received the training. CHWs reported seeking permission from 502 spouses or family members to screen suspected cases of the opposite sex due to cultural 503 norms and fear of stigmatization. This finding is consistent with findings from India that 504 women delayed hospital visits awaiting permission from their husbands or guardians (33). In 505 order to address gender inequities in CM-NTDs interventions, programs must invest in 506 frontline health workers to improve communication techniques to reduce shame and stigma as 507 barriers to healthcare seeking (34). Despite the commitment to gender equity within the 508 research, it was impossible to achieve gender parity in training as the community health 509 workforce in Liberia is predominantly male. National programs must proactively advocate for 510 gender balance in the recruitment of CHWs to ensure that those conducting activities truly 511 represent the gender makeup of such a community. Additionally, integrating gender in all stages 512 of planning, tool development, testing, implementation, and evaluation will improve the equity 513 and impact of case finding. 514

The delivery of this study was limited by the impact of COVID -19 due to partial lockdowns and 516 social distancing restrictions. Liberia confirmed its first COVID-19 case on March 16, 2020. On 517 April 8, 2020, the President declared a 21-day renewable national state of emergency for 518 residents of all fifteen counties until further notice. This meant no inter-county travels between 519 the capital and other counties and a 3 p.m. to 6 a.m. curfew rather than a total lockdown. During 520 this time, residents were permitted to leave their homes only to procure food or health items, an 521 activity limited to one person per household for a maximum of one hour within his or her local 522 area. This state of emergency was extended two weeks until July 21, 2020. Therefore, case 523 detection activities were suspended, resulting in the low number of cases reported in the 524 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. implemented. Hand sanitizers and face masks were provided to CHWs to reduce their fear of 538 interacting with community members during case detection when possible. 539 Furthermore, many CHWs were recruited and trained to conduct COVID-19 surveillance, which 540 was done simultaneously with NTDs case detection and awareness activities. COVID-19 also 541 affected the work of peer advocates. While peer advocates' significant role in case finding was 542 recognized, the duration of their case-finding activities was far less than that of CHWs. As a 543 result, the impact of their work cannot be quantitatively measured for case-finding activities. 544

Recommendations for Health System Strengthening 545 Implementation of the optimal model did come with challenges, which are to be anticipated 546 when integrating interventions into emerging health systems. Despite the global impetus 547 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted February 2, 2022. ; https://doi.org/10.1101/2022.02.01.22269709 doi: medRxiv preprint towards integration, restrictions on donor funding limited the scope of some activities, e.g., 548

length of training and supervision., greater flexibility concerning the scope and integration 549 within the training of multi diseases would enable more effective case finding for NTDs. 550

Similarly, a significant limitation was the logistical support required to implement a robust 551 supervision model. The importance of supportive supervision was repeatedly emphasized; 552 however, the investment in this area was insufficient. Based on these challenges as documented 553 in the results, it is recommended that the time allocated for training should be increased. For 554 example, a three-day training was suggested to allow more time to practice the referral and 555 reporting forms. Supervisors need logistical support to provide on-spot mentoring for CHWs on 556 the supervision forms. Provision of adequate copies of the referral forms to health facilities is 557 suggested to improve documentation and reporting of the referral process. The remuneration 558 system of cash, certificates, and other rewards should be strengthened to avoid delays and 559 encourage CHWs in their work. 560

The challenges documented in the study also highlight the financial constraints and barriers 561 faced by NTDs programs in implementing such an integrated model that requires skilled human 562 resources and logistics for routine activities. Funding allocation will significantly impact the 563 sustainability and scaling up of the optimal model. Implementing the optimal model is largely 564 dependent on donor support; components of the optimal model, such as the incentive package, 565 will be harder to sustain without financial support. With the documented success of integrating 566 active case finding and collaboration at all health system levels through this study, national 567 budgets should prioritize funding for integrated approaches to the case management of NTDs. 568

Low reliance on donor funding will enable the NTDs program to make independent decisions 569 and lead in the prioritization of critical interventions. 570 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 2, 2022. ;

This study has shown how integrated case detection of NTDs is pivotal to achieving the global 572 target to leave no one behind(18). The second WHO NTDs roadmap 2021-2030 prioritizes 573 integrated approaches to achieving new global targets(36). The increase in the number of cases 574 detected and diagnosed at an early stage of the disease through the optimal model outlined in this 575 study enables evidence-based advocacy with policymakers. Optimal models for case detection, 576 referral, and confirmation of suspect NTD cases rely on integrated approaches to training, 577 supervision, referral, and remuneration that are embedded within existing health systems 578 infrastructure. Equity and inclusion are also critical, particularly the gender of health workers 579 and the inclusion of affected persons. The global NTD target SDG 3:3 can only be achieved 580 through optimal case detection, referral, and confirmation models. Together, these approaches 581 improve access to health services, thus reducing morbidity associated with NTDs. 582 Acknowledgment 583 We would like to extend our gratitude to the following people, without whose support this 584 project would not have been successful. Many thanks to the Ministry of Health Liberia, the data 585 collection team comprised of Mohammad Dunbar and Otis Kpadeh, and field drivers for their 586 commitment and endurance during the data collection process. Many thanks to the Ministry of 587

Health NTD program monitoring and evaluation team and the UL-PIRE for the approval to 588 conduct the research, and finally, our most tremendous thanks to the national stakeholders, 589 clinical staff, CHAs, CHVs, CDDs, patients, and communities for their time to participate in this 590 study. Finally, we appreciate Paul Saunderson and Stefanie Weiland for their review of earlier 591 drafts of the manuscript, which contributed immensely to the final draft. 592 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted February 2, 2022. ; https://doi.org/10.1101/2022.02.01.22269709 doi: medRxiv preprint

Neglected tropical diseases in the

Perceptions, health-seeking behavior and access to diagnosis and treatment initiation 597 among previously undetected leprosy cases in rural Maharashtra

Buruli ulcer and 600 leprosy, they are intertwined": Patient experiences of integrated case management of skin 601 neglected tropical diseases in Liberia

Social stigma towards neglected tropical diseases: a 604 systematic review. International Health

Factors 606 contributing to delay in diagnosis and start of treatment of leprosy: analysis of help-607 seeking narratives in northern Bangladesh and in West Bengal

Stigma and the social burden of neglected tropical diseases. PLoS Neglected 610 Tropical Diseases

):328. 614 8. World Health Organization. ACCELERATING WORK TO OVERCOME THE 615 GLOBAL IMPACT OF NEGLECTED TROPICAL DISEASES: A ROADMAP FOR 616 IMPLEMENTATION

Neglected Tropical Diseases, Cross-Cutting Issues Workshop

Netherlands: meeting report. International Health

Integrated Control and 622 Management of Neglected Tropical Skin Diseases. PLOS Neglected Tropical Diseases

Neglected tropical diseases: exploring long term practical approaches 625 to achieve sustainable disease elimination and beyond. Infectious Diseases of Poverty

632 14. World Health Organization. Department of Control of Neglected Tropical Diseases. 633 Integrating neglected tropical diseases into global health and development : fourth WHO 634 report on neglected tropical diseases

636 sanitation and hygiene for accelerating and sustaining progress on neglected tropical 637 diseases: a new Global Strategy 2015-20

Integrated Management of Skin NTDs-Lessons Learned from Existing Practice 639 and Field Research. Tropical Medicine and Infectious Disease

Routine Surveillance Data as a Resource for Planning Integration of NTD Case 642 Management

Leaving no one behind: a neglected tropical disease indicator and 644 tracers for the Sustainable Development Goals: Box 1. International Health

Republic of Liberia Investment Plan for Building a Resilient Health System in Liberia

Neglected tropical disease 653 as a 'biographical disruption': Listening to the narratives of affected persons to develop 654 integrated people-centered care in Liberia

Republic of Liberia 2008 Population and Housing Census

The LEAP Advisory Board, Straits KJE

Community-based participatory research (CBPR): Towards 659 equitable involvement of community in psychology research

Implementing 662 and integrating NTD programmes in Liberia: Reflections from key stakeholders

Impact of General Health Volunteers on 665 Leprosy Control in Nimba County, Liberia: an experience from the Ganta Leprosy 666 Rehabilitation Centre over 2013-2015

Active case detection methods for crusted 668 scabies and leprosy: A systematic review

Achieving the 671 endgame: Integrated NTD case searches

674 Contribution of the Community Health Volunteers in the Control of Buruli Ulcer in Bénin

The impact of community health workers (CHWs)

on Buruli ulcer in sub-Saharan Africa: a systematic review

The 680 contribution of community health workers to the control of Buruli ulcer in the Ngoantet 681 area

Investigating 683 barriers and challenges to the integrated management of neglected tropical skin diseases in 684 an endemic setting in Nigeria

Assessment of needs and quality care issues 686 of women with leprosy

A 688 call to action for universal health coverage: Why we need to address gender inequities in 689 the neglected tropical diseases community

New global targets for NTDs in the WHO roadmap 2021-2030. Plos NTDs

Detailed Optimal Model for case detection, referral, and confirmation