Strengthening Human Resources for Health to Achieve Epidemic Control: A Case Study of Democratic Republic of the Congo

Background The United States President’s Emergency Plan for AIDS Relief (PEPFAR) in Democratic Republic of the Congo (DRC) continues to fund a robust portfolio of programs aimed at achieving epidemic control in three provinces where 30 percent of people living with HIV/AIDS in the country reside. Challenges around human resources for health (HRH), including inadequate stang and limited capacity, impede the delivery of quality HIV/AIDS services in DRC. Methods In partnership with the United States Health Resources and Services Administration (HRSA), PEPFAR, and DRC Ministry of Health (MoH), ICAP at Columbia University worked with 16 PEPFAR-identied high-priority health facilities (HFs) in DRC and developed HRH-specic interventions to address challenges in achieving 95-95-95 targets. These potential interventions were then prioritized for implementation using a collaborative, criteria-driven approach considering factors such as feasibility, viability, and time-to-impact. Through interviews at all 16 HFs, the joint teams developed an intervention framework, determined short-term priorities, and prepared to implement short-term HRH improvements to reach 95-95-95 targets across all HFs. Interviews used an adapted version of the PEPFAR HRH Rapid Assessment tool to capture key HRH information including stang levels by type of clinical or administrative position, key barriers to achieving 95-95-95 targets, and perceptions of needed HRH-specic improvements. Results Site-level interviews occurred in April 2019, and the in-country team created a list of possible interventions across six domains: stang, training, workplace environment, medical supplies and equipment, and monitoring and evaluation. Thirty-ve interventions were hypothesized and prioritized into short, medium, and long-term priorities using a framework focused on desirability, feasibility, viability, and time-to-impact. Some interventions were applicable to all HFs while others applied only to selected HFs, the national MoH, or to implementation partners. Twelve interventions were selected as highest priority, and budget allocations and task planning were developed for each of the high-priority interventions. These high-priority interventions were then launched for implementation and evaluation within six months.


Democratic Republic of the Congo (DRC) Context and Challenges
Democratic Republic of the Congo (DRC) gained independence in 1997 and has received United States President's Emergency Plan for AIDS Relief (PEPFAR) funding since PEPFAR's inception in 2003 [1,2].
DRC is a country of over 2 million square kilometers that borders seven sub-Saharan African countries. Decades of violent con ict and instability have taken a devastating toll on the country's economy, human resources, and infrastructure, and, coupled with a largely rural population, present several challenges to providing healthcare [3].
DRC has a generalized HIV epidemic with an estimated prevalence of 1.2 percent among adults aged 15 to 49. HIV infection is more prevalent in urban settings (1.6 percent) than rural areas (0.9 percent). An estimated 404,894 people are living with HIV/AIDS, and around 10,535 people die from AIDS related conditions each year [4]. The HIV/AIDS epidemic has drastically impacted two of DRC's most populous provinces, with general population prevalence estimates of 1.6 percent in Kinshasa and 2.6 percent in Haut Katanga [5].

Human Resources for Health (HRH): Priorities and Challenges
Strengthening HRH-in terms of quantity, quality, and distribution-has been a priority focus of many global public health and epidemic control efforts, especially in African nations struggling to control outbreaks of malaria, TB, HIV/AIDS, Ebola, and other infectious diseases [6,7]. In DRC, numerous training and policy measures have aimed to address gaps in managing a disparate and diverse healthcare workforce, including nurses, community health workers (CHWs), and midwives, across a geographically and culturally diverse country [8,9]. Since healthcare (and especially HIV/AIDS treatment) is often centralized at major health facilities (HFs) in large cities, decentralized efforts to reach DRC's largely rural population (often with higher infection rates) are constrained by lack of trained health care workers able to provide the full breadth of HIV/AIDS services. [10,11]. Further, the abilities of these community staff and HFs to in uence local healthcare policies and procedures are constrained by distance, staff burnout and a sense that rural health services may not "matter" as much [12,13]. These challenges can lead to staff feeling underappreciated, burned out, or unclear of job responsibilities and expectations [14,15]. Thus, for ICAP and its partners and funders, implementing new HIV/AIDS-speci c interventions can sound terri c in principle but can be quite di cult in practice.
In response to the urgent need to scale-up access to antiretroviral therapy (ART), in close collaboration with the Ministry of Health (MoH) and Nursing Council and Associations, ICAP successfully implemented a PEPFAR-funded program administered by the United States Department of Health Resources and Services Administration (HRSA), the Global Nursing Capacity Building Project (GNCBP), across 10 countries including DRC between 2011 and 2018. Working closely together, HRSA and ICAP partnered with in-country nursing leaders and ministries of health and education (MoH/MoE) to strengthen the nursing and midwifery workforce. This project addressed critical gaps in nursing education and training focusing on seven building blocks of health workforce strengthening: infrastructure improvement, curricula revision, clinical skills development, production, in-service training and faculty development and building partnerships with MoH and MoE stakeholders for policy and regulation with the overarching goal of enhancing the quality and quantity of the nursing and midwifery workforce. Therefore, ICAP and HRSA have a demonstrated track record of collaborative success in HRH improvements in DRC working at both the pre-and in-service levels.

Epidemic Control and PEPFAR
In support of the Joint United Nations Programme on HIV/AIDS (UNAIDS) 95-95-95, PEPFAR-supported programs place emphasis on those activities which are proximal to reaching these targets and rapid treatment initiation and clinic and community efforts which focus on retention in care and viral suppression. Therefore, this paper discusses efforts to strengthen HRH at 16 PEPFAR-identi ed "highpriority" HFs in the Haut Katanga region towards achievement of 95-95-95 goals. These HFs were prioritized based on the volume of patients on ART and regional HIV/AIDS prevalence and, as such, appears to be a good target collaboration between ICAP and HRSA/PEPFAR collaborations. The HIV/AIDS prevalence in Haut Katanga is 2.5 percent compared to 1.2 percent in the capital, Kinshasa and is challenged by a largely rural population and large land mass borders with both Zambia (HIV/AIDS prevalence of 11.4 percent among adults aged 15-49) and Malawi (10.0 percent among adults aged 15-49) [5,16,17].

Resilient and Responsive Health Systems (RRHS) Projects
With PEPFAR funding through HRSA, in 2017, ICAP was awarded the RRHS project to continue strengthening HRH in DRC. Using the World Health Organization (WHO) conceptual framework for HRH development, the rst two years of the RRHS project built on extensive HRH capacity building and infrastructure improvements for student nurses and midwives through HRSA funding the Nursing Education Partnership Initiative (NEPI) as a follow-on to the previously funded GNCBP [18]. Since nurses are the largest HRH component in DRC, this NEPI-focused work centered on nurses' readiness for clinical practice through curricula reform and development of innovative pedagogy including use of skills labs and simulation-based training. Over the past two years, HRSA and ICAP have leveraged these pre-service strengthening efforts to shift focus to in-service capacity building and addressing wider HRH limitations affecting epidemic control. Improving clinic personnel collaboration, implementing healthcare worker and management training in HIV/AIDS and other leadership skills, upgrading infrastructure for HRH effectiveness, and improving morale and job stability via rural rotations of nursing school graduates are all examples of RRHS activities that focus on HRH in DRC.
The value-added of such HRH interventions extends beyond C&T programs and thus may support faster achievement of PEPFAR 95-95-95 goals. These faster outcomes are well-known from prior research as are the pitfalls if such HRH needs are not addressed [19][20][21]. Thus, this paper outlines ICAP's method to develop targeted, prioritized, and sequenced interventions-at the HF, MoH, and ICAP levels-to improve HRH effectiveness. With improved HRH effectiveness, ICAP and HRSA hope to demonstrate faster and more e cient achievement of 95-95-95 goals than would have been possible with only C&T interventions, alone [22]. Further, such HRH improvements can inform country-and regional public health policy to more accurately and quickly identify health workforce needs moving forward [8].

Methods
A needs-driven process was used to determine site-level HRH needs across the PEPFAR-identi ed 16 highpriority HRs and to develop e cient action plans to address these needs within 12-18 months. This protocol was submitted to the Columbia University Irving Medical Center Institutional Review Board (IRB) and received a "non-research" determination. A series of 13 steps were completed between October 2018 and June 2019 after which HRH interventions were launched in early July 2019: 1. The PEPFAR Rapid Site-Level Health Workforce Assessment (hereafter, "HRH assessment") was modi ed and translated into French [23,24]. The modi ed assessment collected no personal information-such as names of HF staff or management-and focused on challenges that HRH faced, not "time and motion" information regarding daily sta ng, vacations, etc. Exhibits 1A, 1B, and 1C provide excerpts of the original English HRH assessment tool, the French translation of the same, and a section of the (French-language) Microsoft Excel data collection tool, respectively 2. Ten high-priority sites were selected for survey piloting. Local ICAP staff were trained on the survey and then scheduled initial interviews with HF leaders and staff. All participants were advised that no personal information would be logged and that all responses would be anonymous. Any participant could opt-out of the interview-in whole or in part-at any time 3. Once the ten pilot site interviews were completed, the ICAP team reviewed initial ndings and suggestions for how to improve/streamline the survey questions and interviews 4. All 16 sites were then interviewed-including re-interviews of the ten pilot sites-with the modi ed HRH assessment and presentation of current Monitoring, Evaluation, and Reporting (MER) statistics for each site [25]. Exhibits 2A, 2B, and 2C show MER snapshot statistics for one of the high-priority HFs in DRC. (Each sub-exhibit is one of the three "95-95-95" MER sets.) Any target coded "green" was considered su ciently met. Yellow and red colors indicate improvements needed or lagging performance relative to PEPFAR expectations 5. The team reviewed all data and prepared summaries of key needs from HF interviews. This helped the team establish top HRH challenges and needs for each site 6. The team then assessed volume data-of patients, clinical staff, etc.-to determine what training gaps existed and where excessive patient loads were across the HFs. This allowed the team to determine if site-level HRH needs were related to patient volumes, to numbers of HIV/AIDS trained staff, or to other factors beyond the C&T domain 7. From this work several hypotheses for interventions that could address key HRH needs/challenges were developed. These were discussed with the local C&T teams to determine if any of ndings were misaligned with C&T goals and improvements already scheduled 8. Interventions were then prioritized using a prioritization framework with criteria such as desirability, feasibility, and time-to-impact (see detailed discussion below). Priorities were shared with key HF and HRSA leaders for review and comment. Prioritization of interventions was informed by frameworks used in similar contexts in other infectious disease treatment areas and African nations [26,27].
9. Interventions were then determined to be of short-term focus (within the rst six months, mediumterm (to be done in months 7-12), or long-term (to be completed in months 13-18, or by December 2020) 10. Once agreed, HRH interventions were slotted as being the responsibility of speci c sites, all sites, the MoH, or ICAP 11. Short-term HRH interventions-including establishing a July 2019 baseline of 95-95-95 MER statistics for future benchmarking-were then launched 12. Concomitant with short-term intervention launch, the team developed a tracking tool, comprising both C&T and HRH intervention efforts, that aligned efforts with all 95-95-95 targets 13. Since mid-2019, monthly reporting of MER data-and of key C&T and HRH intervention progresshas been provided to all in-country leadership and HRSA. Bi-weekly conference calls with in-country leaders, ICAP staff, and HRSA leadership are held to assess line-item progress and exception handling The collaborative process used focused on building trust with sites and empowering HF leaders to express needs for HRH improvement in ways the team could report back to funders (HRSA) and MoH. Building relationships with HF leaders was critical to success of any planned intervention especially to upscaling sta ng, upskilling health workers, and improving workplace conditions and culture. Trustbuilding is especially important when providing nancial incentives for performance is impossible and quickly adding quali ed staff is unlikely [28][29][30].

Results
All 16 high-priority HFs responded in detail to the HRH assessment and participated fully in on-site interviews. Results from the team's initial data collections indicated that, contrary to initial conjecture, gross sta ng level challenges were generally not an issue at the HFs. Exhibit 3 shows the sta ng across all 16 sites by type of staff (e.g., doctor, nurse, midwife) and those who were trained in HIV/AIDS care and treatment. Full-time vs. part-time sta ng is also noted. Total sta ng numbers include non-HIV/AIDStrained staff, but for the most part, HFs in aggregate were well-staffed for HIV/AIDS care. From these analyses, the team concluded that having enough staff was not a key HRH issue and that, instead, other needs likely deserved more attention (aside from the clear HRH resource needs in the ve sites mentioned in Exhibit 4). After compiling the qualitative results from the HF interviews, Exhibit 5 offers some interesting insights into HRH challenges in the HFs:

An evaluation of HIV/AIDS caseload was then performed to determine if appropriate levels of HIV/AIDS
While few barriers seem to be compensation-related, more HIV/AIDS clinical staff and training (clinical and non-clinical) are needed. Low morale seems tied to lack of job descriptions, evaluations, and career pathing Improvements in morale could lead to better recruiting and/or positive selection for employment at HFs, which could help "up-scale" these HFs [31,32] Community activities are not well-organized due to lack of peer educators and community health worker training and oversight. Inconsistent oversight of and non-standard documentation from these workers is not only in the HIV/AIDS space but also in TB and malaria domains Work environments require improvement-spaces to collaborate, rest, etc.-and workers need more protective clothing and communities of interest to share / learn best practices Improved ART distribution would help as would access to more advanced therapies. Budgeting support will also help some individual HFs [33] After a careful analysis of HF interview data, the team reviewed literature on HRH intervention research and determined that the barriers, above, could be addressed with interventions across the following six domains [34]: Determining responsible stakeholders for these interventions is also critical [36]. The team determined that four key stakeholders would be assigned responsibilities for any interventions across the six domains, above. Interventions would be implemented by all sites, selected sites, the MoH, and/or by ICAP.
Exhibits 6A and 6B show the "universe" of proposed interventions by key stakeholder/responsible party. For example, in the "Training" domain, "All Sites" will need to provide more HIV/AIDS-speci c training while only one "Selected Site" will need to hire dedicated HIV/AIDS staff members. The MoH would need to approve budget for such training and personnel, and ICAP would need to provide the proper job description(s) and training plan for new hires or training programs for existing HRH.
However, not all of these 35 interventions are of the same importance/urgency nor could all be launched simultaneously (assuming resources were even available to do so). Thus, the team constructed a set of prioritization criteria in order to stage and sequence these interventions in a more manageable way [37]: Desirability: Do HFs, the MoH, ICAP really want to pursue this intervention (based on past experience, etc.)?
Feasibility: If the intervention is desirable, do HFs, the MoH, ICAP, etc. have the (non-nancial) resources, skills, and tools ready to kick-off the intervention?
Viability: If stakeholders have the skills and time, do they have the funding, or can they quickly get the needed it to fund the intervention?
Sustainability: Once implemented, will the HRH interventions "stick;" i.e., will the improvements stand the test of time or be lost as just one-time changes?
Time to Impact: Finally, if stakeholders implement an intervention, have skills to do it, and can be funded, will they see a return on their efforts based on meaningful and timely impacts?
After working with HFs, HRSA leadership, ICAP, and MoH, the team prioritized the interventions, and Exhibit 7 shows the results of this collaboration. Efforts were prioritized in terms of what needed to be implemented in the short-term (within the rst six months), over the medium-term (within the following six months), and over the long-term (within the nal six months of the total intervention time frame of 18 months). Exhibit 8 shows only the short-term interventions noting that two of them had already been started by C&T teams.
Based on successes of similar HRH interventions in other sub-Saharan contexts, the team expects to see several improvements [35]. These are aligned with the major HRH domains mentioned previously: Sta ng: Increase numbers of patients tested/tested positive for HIV/AIDS and those treated if positive (including pregnant women) as more quali ed health workers, midwives, and peer educators are put into the HRH workforce Training: Install more staff able to help HIV/AIDS patients, improve satisfaction, and enhance clinic leadership. These should create more awareness for clinics (or "pull") for new and existing cases Workplace Environment: Improve workplace environments that lead to more team-based care, best practice implementations, and improved morale for staff The team hopes to use these data to perform program evaluations to determine intervention-and program-level effectiveness and to show how these efforts have led to better outcomes, as measured by MER reporting, than C&T interventions, alone, may have created.

Discussion
The tools developed-and preliminary results achieved in HRH effectiveness-would not have been possible without direct, site-level data collection and interviews from HF leaders (both clinical and nonclinical). By assuring respondents con dentiality of collected inputs, the team was able to obtain honest feedback about key HRH challenges and needs. The PEPFAR HRH rapid assessment tool proved useful in collecting these data quickly. Piloting the process rst, and then revising the data collection tool also proved valuable since the team could incorporate learnings from early site interviews to make the entire data collection more effective.
Sustained contact with stakeholders throughout the processes of setting MER baselines and HRH sta ng, hypothesizing possible interventions, prioritizing interventions, and launching and tracking implementations allowed the team to keep stakeholders aligned and focused on important endpoints throughout this study. Stakeholders agreed to the major domains of HRH needs, prioritization criteria, and the staged implementation plans. Since launch, stakeholders have appreciated the value of monthly MER reporting aligned with updates that show successes and challenges with corrective actions planned to mitigate poor performance (by month by HF).
Keeping focus on various stakeholder needs, too, has proven to be valuable as MoH priorities may not always align with HF needs (or those of funders, necessarily). Thus, with open communication and through monthly reviews and planning, these potential "de-railers" to HRH success can be made evident, planned for, and accommodated by multiple parties working collaboratively. This, in turn, has built trust across several constituents, all of whom have a vested interest in HIV/AIDS epidemic control through improved HRH sta ng, training, and collaboration. Prioritization and sequencing of HRH interventions was key to successful implementation.
But without tracking-of both intervention progress and the direct results of interventions through MER results-any program will likely produce suboptimal or unmeasurable results. Thus, the close coordination of C&T and HRH teams was a critical element of success, and this coordination has resulted in greater trust and improved knowledge-sharing across these teams.
The team looks forward to reporting more results from its evaluations of these HRH interventions-and their correlations to changes in MER improvements-in the coming months. Also as the team expands this study to non-PEPFAR funded areas/countries, it hopes to see similar results in other contexts (and, perhaps, beyond just HIV/AIDS-dominated areas).

Limitations
Some limitations to this approach are noteworthy. First, only PEPFAR high-priority HFs were interviewed for these targeted HRH interventions. Although the team believes these interventions are relevant to other HFs, the HRH assessment did not compare priority to non-priority sites and has not identi ed possible divergent HRH needs.
Second, given different country contexts in which these interventions may apply, results may vary based on MoH adoption, in-country and external funding, and in-country HRH sta ng and skills. The team hopes to pilot this study elsewhere to see if other HFs identify similar HRH needs and can realize similar gains from such interventions. National and local government priorities, which can change based on varying needs and "shocks," will always be exogenous factors that could affect any intervention e cacy.

Conclusions
While the correlation between number of health workers and care outcomes is well documented [22,38], less evidence exists regarding HRH interventions that go beyond the sta ng numbers. Here, the team determined that some HRH needs and challenges, for 16 high-priority PEPFAR HFs in DRC, were actually not related to sta ng levels but to other factors including training, workplace environment, morale, and a need to feel/be connected to other HFs and to best practices.

Funding
This research has been supported by PEPFAR through HRSA. Funding for this effort was provided to ICAP through HRSA grant UH6HA30739. John Ngulefac, one of the study's co-authors, is a Public Health Analyst, HRSA, and has supervised this funding.

Availability of Data and Materials
The amended rapid assessment survey used and site-level data collected for this study are available from the corresponding author upon reasonable request.

Ethics Approval to Consent and to Participate
The Columbia University Irving Medical Center's IRB reviewed this proposed research study and gave it a "non-research" determination (IRB-AAAS2598) on February 22, 2019. This determination was shared with DRC MoH, and the MoH also deemed this study to be of a "non-research" nature. Only site-level information was retained from interviews, and no speci c individuals' names or identifying data were captured as part of any interview or discussion. Although none did, any potential interviewee could optout of the site-level interview and/or refrain from completing the amended PEPFAR assessment tool in part or in full.

Consent for Publication
The views expressed in this article are those of the authors and should not be construed to represent the positions of HRSA or of the organizations with which the authors are a liated.