The results of this study are presented as follow as characteristics of selected districts, performance of districts and quantitative findings.
Quantitative study findings
Characteristics of selected districts: quantitative data were collected from eight districts in four regional states of Ethiopia. On average, there are about 121,328 inhabitants in each district. Health services are offered to the community through three primary hospitals, 37 health centers, 164 health posts, and 175 primary schools (table 1).
Performance of district health systems
All twinning partnership targeted districts were rated on three occasions: baseline, midterm and end-line stages. Figure 3 below depicts the five and three districts that were categorized as low and medium performers at the baseline measurement stage, respectively. These scores were improved at midterm as one, four and three district categories as high, medium and low performing districts, respectively. The end-line results revealed that out of eight districts, four fulfilled the transformation criteria, a district was categorized as a medium performer and the remaining three districts - despite improving their scores, fell in the low performing district category.
Table 2 below shows that the correlation (r) results of district health system performances. There was a strong degree and statistically significant relationship between baseline and midterm (r>0.978**), baseline and end-line (r>0.936**), and midterm and end-line (r>0.987**) scores.
Difference in woreda management standard scores over time
At baseline, the mean score in woreda management standards (WMS) was 6.42, at midterm, it was 7.52, and at ‘end-line’, it was 8.15. There was a statistically significant difference in WMS scores at baseline, midterm and end-line measurements, χ2(2) = 14.250, p = 0.001. Post hoc analysis using the Wilcoxon signed-rank test was conducted with a Bonferroni correction applied, resulting in a significance level set at p< 0.017. Median (IQR) WMS at baseline, midterm and end-line were 6.45 (5.35 to 7.42), 7.65 (6.25 to 8.85), and 7.95 (6.95 to 9.30), respectively. There were statistically significant positive differences between the WMS2 – WMS 1 (Z = -2.524, p = 0.012), WMS3 – WMS2 (Z = -1.963, p = 0.050), and WMS3 – WMS1 (Z = -2.521, p = 0.012).
Difference in model village scores over time
At baseline, the mean score of model villages (MV) was 11.58, at midterm, it was 14.88, and at ‘end-line’, it was 20.85. There was a statistically significant difference in MV coverage at baseline, midterm and end-line measurements, χ2(2) = 16.000, p = 0.001. Post hoc analysis using the Wilcoxon signed-rank test was conducted with a Bonferroni correction applied, resulting in a significance level set at p-< 0.017. Median (IQR) MV coverage at baseline, midterm and end-line were 10.65 (7.12 to 18.65), 14.56 (9.52 to 21.37), and 22.05 (14.70 to 26.32), respectively. There were statistically significant positive differences between the MV2 – MV1 (Z = -2.533, p = 0.011), MV3 – MV2 (Z = -2.527, p = 0.012) and MV3 – MV1 (Z = -2.524, p = 0.012).
Difference in high performing PHCU scores over time
At baseline, the mean score of high performing PHCUs (HP PHCU) was 17.96, at midterm, it was 20.43, and at ‘end-line’, it was 23.32. There was a statistically significant difference in HP PHCU scores at baseline, midterm and end-line measurements, χ2(2) = 11.677, p = 0.003. Post hoc analysis using the Wilcoxon signed-rank test was conducted with a Bonferroni correction applied, resulting in a significance level set at p< 0.017. Median (IQR) HP PHCU at baseline, midterm and end-line were 17.10 (14.02 to 22.57), 20.55 (15.97 to 24.22), and 22.80 (19.35 to 28.12), respectively. There were statistically significant positive differences between the HP PHCU2 – HP PHCU1 (Z = -1.960, p = 0.050), HP PHCU3 – HP PHCU2 (Z = -2.536, p = 0.011), and HP PHCU3 – HP PHCU1 (Z = -2.366, p = 0.018).
Difference in community-based health insurance scores over time
At baseline, the mean score of community-based health Insurance (CBHI) was 15.0, at midterm it was 17.51, and at ‘end-line’, it was 19.74. There was a statistically significant difference in CBHI coverage at baseline, midterm and end-line measurements, χ2(2) = 13.556, p = 0.001. Post hoc analysis using the Wilcoxon signed-rank test was conducted with a Bonferroni correction applied, resulting in a significance level set at p< 0.017. Median (IQR) CBHI coverage at baseline, midterm and end-line were 14.40 (10.87 to 19.72), 18.75 (11.25 to 22.27), and 21.39 (11.85 to 25.65), respectively. There were statistically significant positive differences between the CBHI2 – CBHI1 (Z = -2.207, p = 0.027), CBHI3 – CBHI2 (Z = -2.371, p = 0.018), and CBHI3 – CBHI1 (Z = -2.371, p = 0.018).
Difference in overall district health system performance scores overtime
At ‘baseline’ the mean score of district health system performance (DHSP) was 50.97, at ‘midterm’ it was 60.3, and at ‘end-line’, it was 72.07. There was a statistically significant difference in district health system performances at baseline, midterm and end-line measurements, χ2(2) = 16.000, p = 0.001. Post hoc analysis using the Wilcoxon signed-rank test was conducted with a Bonferroni correction applied, resulting in a significance level set at p< 0.017. Median (IQR) district health system performance measurements at baseline, midterm and end-line were 47.55 (36.10 to 67.32), 60.50 (42.3 to 74.18), and 72.99 (52.28 to 89.25), respectively. There were statistically significant positive differences between the DHSP2 – DHSP1 (Z = -2.521, p = 0.012), DHSP3 – DHSP2 (Z = -2.521, p = 0.012), and DHSP3 – DHSP1 (Z = -2.524, p = 0.012). Therefore, it can be concluded that the performances of district health systems were significantly raised over time.
Qualitative study findings
Data collected from health workers with diverse professional backgrounds: health officers (14; 35.9%), nurses (9; 23.1%) and midwives (2; 5.1%) were used for the study. The majority 33 (84.6%), were male. The mean age with standard deviation was 28.5 ± 5.0 years. On average, 7.5 years of service were tenured by the KIIs (table 3).
Based on BMCF, three major categories were discussed under inputs. The categories discussed below are mission, partner resources and financial resources.
In this study, ‘mission’ means the main reason stated for the existence of the twinning partnership. A common understanding of the mission and the health system’s strategic priorities enhance access to quality primary healthcare services in an equitable manner. Such systematic interventions help the Ethiopian health sector create a resilient district health system that is responsive to the needs and demands of all individuals in all places. These collective and widespread comprehensions help members of the twinning facilities to develop a shared vision that inspires partner districts and their staff, share resources, develop a culture of serving communities outside their district boundaries, and grow and become stronger together while maintaining a sense of competitiveness among members. This indicates that the implementation of health sector reforms enhance the governance, capacity, quality, and equity of access to primary health care services. The following verbatims describe the opinions of health workers on their understanding of the mission and shared vision of the twinning partnership strategy.
“Although we, [members of twinning partnership] live in neighboring woredas and are familiar with each other as health workers, we had never talked with each other about enhancing our health services. If fact, we were reluctant to share information as we wanted to stand out and be better performers than other health centers located in the same [name] zone. The twinning partnership approach helped us to open our eyes and broadened our horizons. More specifically, we understood that through working together, we became stronger. Hence, the partnership helped us get closer to our ultimate goal of serving the community.” (004, Health Center Director, Health Officer, Oromia region)
Some of the twinning partnership districts developed a vision or mission statement that supports the achievement of the Ethiopian health system long term goal of achieving UHC by increasing the coverage of transformed districts: “…to create model primary health care units in both [name] and [name] districts.” (033, District Health Office Head, Master of Public Health, Amhara region). Another staff member commented that their vision was, “… to be a transformed district.” (001, Health Center Director, Health Officer, Oromia region)
A health center staff also expressed his observations on the benefits of empowering health workers through a shared vision: “Staff at our partner district were familiar with the health sector reforms. Therefore, we created a platform for our districts to organize a number of seminars for staff that assisted us in creating a shared vision.” (010, Vice head of Health Center, Health Extension worker, Oromia region)
Another health professional commented on the impact of having shared missions and visions saying; “Previously, other sector heads considered health as a well-financed sector through development partners. However, after visiting the health centers in person, they recognized that health is an expensive service for many beneficiaries who expressed several grievances. To fulfil the minimum standards, ownership and local financing can help the health system to achieve its major goal of preventing maternal and child deaths.” (012, Reform Core Process Owner, BA in Economics, Tigray region)
Narrowing the performance gaps of districts in a short time demands the utilization of internal and external resources. The twinning partnership strategy was facilitated by three to four committed staff members assigned from the district health offices and health centers. Zone Health Department (ZHD) assigned focal persons who liaise between districts. In addition, focal persons were assigned to each department such as the laboratory, pharmacy, health center - health post linkage, health information system, quality improvement and infection prevention in each partnering district. The majority of the respondents believed that having a dedicated partnering facility helped them adopt innovative tools and achieve better results. One of the respondents commented, “If it wasn’t for the technical support on facilitating the preliminary discussions, partnership development, facilitating basic twining partnership trainings, developing problem solving skills, partner districts would have continued doing things as usual.” (021, Health Center Director, Environmental Health Officer, SNNP region)
This broad category addresses the effort of twinned districts in mobilizing monetary resources. Experience sharing between medium and low performing districts helps executive and decision-making bodies understand the concept of performance management by institutionalizing minimum standards. A district health office staff member explains, “The experience sharing and learning tour helped the executive team understand the idea of achieving universal health coverage through district transformation and enabled them to compare our district’s achievements to that of our neighboring twin woreda, [name] district. They, [executive body] committed themselves to replicate what they saw in [name] Health Center and allocated $66,000.00 (sixty-six thousand USD) for availing essential drugs and the renovation of health facilities.” (037, District Health Office Head, BSc in Nursing, Amhara region).
Another district health office staff had this to say about the financial support they received from a development partner: “Without the financial support we received from the project, we might not be familiar with the concept of the twinning partnership which led us to develop projects and engage in the implementation of activities.” (018, District Health Office Head, master’s in public health, SNNP region).
Therefore, scanning health sector priorities and the establishment of missions led the participants to develop a shared vision of serving communities that motivated them to mobilize the necessary human, financial and other resources. These factors were the key elements of the inputs in the implementation of the twinning partnership strategy in Ethiopia.
Five major categories were presented under the theme of collaboration. The emerged categories discussed below are input interaction, leadership, formal roles and procedures, communication, and maintenance task.
The results of scanning missions, developing shared visions, mobilizing financial, human and other resources were steps of the collaboration that established inputs for the interactions. The majority of the respondents believed that the shared missions, shared visions, interactions with partners and allocation of financial resources had a positive effect on accelerating the performances of primary health care units and the district health system.
One of the district health office heads had this to say: “…staff mobilization and deployment, allocation of budget for fuel, covering staff accommodation costs and sharing of resources in the twinning partnership assisted us to motivate staff and led to the achievement of our shared vision.” (011, District Health Office Head, BSc nurse, Tigray region).
Similarly, a head of a health center expressed, “Our health center ran out of lab supplies for syphilis screening tests, and the antenatal services were not complete for pregnant mothers. Similarly, our partner district reported an unusual increase in the number of malaria cases observed in health facilities, while they lack antimalaria drugs. Both of us benefited from the established relationship as we were able to share human resources and essential drugs and supplies.” (034, Health Center Director, BSc Nurse, Amhara region).
The capacity of the leadership to scan their environment, focus on impactful interventions, exhibit motivating and inspiring behavior, and mobilize and align resources were pointed out as having a positive influence on the achievement of twinning partnership projects. A health worker affirms, “Assessing the needs and demands of the district health system’s environment, mobilizing and aligning resources coupled with leaders’ recognition and acknowledgement of staff cemented commitment for the twinning partnership.” (017, Health Center Director, Health Officer, Tigray region).
The majority of the respondents believed that the commitment of the assigned focal persons in planning, organizing, facilitating, implementing and monitoring the activities had a positive effect on the established partnerships. Almost all staff who were engaged in the twinning partnership were acknowledged by the top managers as having facilitated vehicles, per-diem and other resources effectively which led to the successful implementation of the developed twinning projects.
Formal roles and procedures
Each level of the healthcare system maintains a set of roles and responsibilities endorsed by partners ensuring the implementation of the twinning strategy at all levels of the health system. The majority of the respondents highlighted that the established clear structures, roles and responsibilities of all stakeholders helped them to achieve better results. One of the district health office managers had this to say: “The launching workshop showed us the structures, roles and responsibilities of the twinning partnership strategy. We also defined and shared this at the district health office and health center level which was instrumental for effectively carrying out our twinning partnership planned activities.” (002, District Health Office Head, BSc Nurse, Oromia region). Committed and willing districts and zone health departments signed a memorandum of understanding to work together for one full year.
During the implementation of the twinning partnership, project members communicated with each other through various channels including telephone conversations, written communication, creating groups through mobile phone applications, and through face to face communication. A health worker had this to say about improving performance of partners using telephone communication: “We usually use telephone communication to arrange meetings, experience sharing events, and invite experts. This helped us to maintain our friendly relationship.” (012, Health Reform Core Process Owner, BA in Economics, Tigray region).
Another health worker described the importance of face to face communication supplemented by written letters on sharing of drugs and supplies: “If we were requesting drugs and supplies we used formal letters along with face to face communication” (005, Head of Health Center - Pharmacy Department, Pharmacy Technician, Oromia region). The health manager explains how a mobile phone application was used to create groups for information sharing on the performance of the district’s health system. “We have a Telegram [application] group where we update each other on our day to day performances.” (028, Maternal and Child Health expert, Health Officer, SNNP region).
Another health center staff had to say: “Before we engaged in the twinning partnership, our communication was limited within district health teams as we had no means of sharing experience and supporting each other. We used to mainly meet during review meetings where we share and learn about experiences and successes. We also did not have the means to organize learning tours to other primary health care facilities. Now, our health center [name] formally communicates with the twinned health center located in [name] district.” (001, Health Center Director, Health Officer, Oromia region)
The basic twinning partnership training focuses on four main chapters, namely: health sector priorities, strategic problem solving, performance management and communication. During the implementation of the twinning partnership, the importance of clear reporting requirements and of sharing information were well addressed. One of the district health office staff had this to say: “During the basic twinning training, participants identified the current situation, developed desired measurable results, identified obstacles and prioritized solutions. In addition, a detailed action plan on resource mobilization as well as monitoring and evaluation was prepared. These activities helped us to share basic information and performance status.” (032, District Health Office Vice Head, BSc Nurse, Amhara region).
Another staff member explained how the achievements were garnered through integration with routine health system activities: “We organized supportive supervision, learning tours, and facilitated a number of workshops on the Ethiopian primary health care alliance for quality.” (023, Health Center Staff, Midwife, SNNP region),
A health center staff also expressed his opinion on effective maintenance: “The staff facilitated onsite and off-site trainings, experience sharing events, common review meetings and expert exchange platforms.” (038, Health Center Quality Improvement Officer, master’s in public health, Amhara region)
Three major categories were discussed under outputs. The summary of additive results, synergy, and antagonistic results will be presented below.
Additive results implied the implementation of health sector priorities separately, without considering the effects of implementing the twinning partnership strategy. The majority of twinning partnership targeted districts reported their engagement through orientation of health sector reforms, facilitated self-assessments against standards and provided routine health services. A district health office head said, “…the health office organizes and facilitates orientation of health sector reforms.” (011, District Health Office Head, BSc nurse, Tigray region).
Another health worker had this to say about the routine activities in their office implemented regularly: “Every quarter, the performance management team assesses performance against the standards.” (001, Health Center Director, Health Officer, Oromia region)
A health worker described the additive results of the established twinning partnership saying: “…though we had relatively higher performances than our partner low performer district within the partnership, we learned about experiences of implementing challenging interventions from our twin woreda, [name]. We also adopted the best practices and collected lists of items which are essential for maternal waiting homes and audio-visual job aids.” (034, Health Center Director, BSc Nurse, Amhara region).
Despite the additive results and synergy observed among partnering districts, there were some observed antagonistic results. During the experience sharing events as well as while conducting integrated supportive supervisions, some staff did not know the reason for the established partnership and the investment appeared to them as a waste of resources.
“While we were motivated to share our knowledge and skills to our partnering district staff, they perceived us as solely having travelled over 90 kilometers to get financial rewards through our meal and accommodation expense payouts.” (032, District Health Office Vice Head, BSc Nurse, Amhara region)
Some of the health workers pointed out that lack of transparency in decision making and lack of inclusion of all departments in the established twinning partnership had a negative impact on achievements. Furthermore, they describe the demotivating effects of lack of good governance on collaborative efforts.
“I was one of the active participants in the development of the one-year twinning project. However, it was not clear to some of us, the process through which leaders and managers hand-picked staff for the experience sharing events….” (005, Health Center Pharmacy Head, Pharmacy Technician, Oromia region)