This study used a convergent parallel mixed-method that combined quantitative and qualitative approaches to evaluate system functionality, analytics capacity, and data use practices in the selected sub-national units in Ethiopia. Both statistical measurement and a deeper understanding of the barriers and facilitators of using digital health systems were employed. The assessment took place from August to September 2025.

The study was conducted in ten woredas or towns across six regions and administrative areas of Ethiopia. The study population included people and institutions involved in generating, managing, and using health information at various levels of the health system. A purposive sampling strategy ensured representation of urban, rural, and pastoralist settings.

All health facilities in the selected woredas were included for the quantitative part, along with one purposively selected woreda health office for each site, resulting in a total of 53 facilities and 9 health offices. One respondent per site, including data clerks, clinicians, and Health Management Information System (HMIS) officers, participated. For the qualitative part, one key informant from a health facility and one from a woreda health office were selected. Site selection considered factors like infrastructure readiness, data management capacity, geographic diversity, program relevance (HIV, TB, malaria), accessibility, and institutional readiness scores. ¹¹

Data collection involved tools such as a facility checklist, a structured survey, and KII guides. Desk reviews included national documents like the Digital Health Strategy and Health Information System (HIS) Roadmap. ¹¹ Standard tools, such as the WHO Digital Health Readiness Tool and national Performance of Routine Information System Management (PRISM) frameworks, were adjusted to evaluate infrastructure, system use, and workflows. ¹² The data were collected using Kobo Toolbox/ODK to gather information on digital tool use, data literacy, and reporting practices. The KIIs were used to examine governance, data culture, and interoperability challenges.

To ensure data quality, the team conducted training, supervision, digital data validation (using timestamps and GPS), and transcription verification. Quantitative data were analyzed using SPSS v27, while qualitative data were coded and thematically analyzed using ATLAS.ti v9. ¹³ The findings of both methods were integrated through triangulation to provide a comprehensive understanding of the results.

Smartphones (58.1%) and laptops (43.6%) were less common among the 62 health facilities assessed than desktop computers (91.9%). Only 42.1% of the 463 functional computers available across 56 facilities were utilized to manage HIV, TB, malaria, and Resilient and Sustainable Systems for Health (RSSH) data. Despite 74.2% of facilities reporting internet access, only 45.7% had consistently reliable internet access ( Table 1).

The qualitative findings showed few and non-functional computers, forcing users to use personal devices and paper-based systems, and increased workload and errors. Participants emphasized improving key priority areas such as Information and Communication Technology (ICT) infrastructure, reliable internet, and power supply, Electronic Medical Record (EMR) implementation, DHIS2 strengthening, system integration, dedicated HIS units, adequate budgets, and facility improvements to enhance data use.

Capacity Building and Trainings

The assessment indicated that of the 3,192 Primary HealthCare (PHC) employees, only 107(3.4%) received data use and analytics training. The majority of training took place off the job through mentoring and experience sharing. Although 71.0% reported partner support, overall training coverage and practical capacity remained low ( Table 2).

Strong demand for training in data quality, DHIS2, HMIS indicators, data analysis and visualization tools, and digital systems (Electronic Community Health Information System-eCHIS, EMR) was revealed by qualitative findings. Along with refresher training and capacity building on dashboards and new technologies for HIV, TB, malaria, and RSSH programs, participants also emphasized mentoring, data triangulation, and program-specific tools.

The assessment showed low digital analytics capacity across all facilities, with about 60.0% had no dashboards, while 66.0% of staff were not trained in digital systems. Limited facilities created analytical tools, and 33.9% were with limited data interoperability. There was little exposure to advanced tools like Apache Superset, indicating large gaps in digital capability ( Figure 1).

The Key Informant Interview (KII) participants also suggested additional digital systems and tools, including web-based platforms for HIV data, Excel and DHIS2 enhancements, Superset applications, SMART Care, Dagu system, Human Resource Information System (HRIS), Logistics Management Information System (LMIS), eCHIS, EMR, Tableau, and Power BI. They also recommended improved dashboards and new digital tools to support HIV, TB, malaria, and RSSH programs.

Institutional Capacity that supports Regular Data Analysis and Use

The assessment showed that 29.0% of facilities lacked focal persons, indicating uneven institutional capacity for data analysis and use. Best practices were not shared by the majority (87%.0). There are gaps in technical support and knowledge sharing for HIV, TB, malaria, and RSSH programs, as only roughly half received regular mentorship, mostly quarterly ( Table 3).

Qualitative findings showed that the Performance Monitoring Team’s (PMT’s) function is inconsistent, limiting effective data review and use. The members’ capacity gaps to interpret and apply data reduce its impact. To increase institutional capacity, participants stressed the need for improved infrastructure, digital tools, training, mentorship, and easier access to HMIS.

Availability of Data Systems (specifically for TB, HIV, Malaria, and RSSH)

The assessment showed that the DHIS2 was the widely available system, which 77.8% of facilities used online, but the rest continued to use paper records. Only 29.0% of facilities had access to the eCHIS system, but 55.6% of them used it offline, and only 21.0% of facilities had access to the LMIS (Dagu and related systems) system, but 53.8% of them used it online.

Ethiopia has fully adopted DHIS2 as its primary platform for managing routine health data. There are other systems, such as eCHIS, SMART Care, and EMR, but their functionality is limited, and their implementation is uneven. Interoperability between systems is still a challenge, hindering integrated data use across programs.

Data Quality Status

The data quality status across facilities and health offices showed that content completeness was not achieved by 53.2% of facilities during the survey year and 54.8% in the previous year. Regarding representative completeness, 51.6% of facilities or woreda health offices did not conduct both assessments in the last year and during the survey year. Timeliness of reports was performed by only 64.5% of facilities for both the previous year and this year, while 48.4% of facilities conducted data accuracy checks for both years ( Figure 2).

Data quality is still a major problem because of inconsistencies, errors, and a lack of capacitated staff, which makes people less likely to trust and use the data, even though some improvements were noted. Timeliness is a big challenge because delays at all levels make the data less useful. Poor infrastructure, limited staffing, workload, and weak connectivity are some of the root causes that affect the overall system performance.

Data Quality Assurance Practices

Most facilities (82.3%) conducted data quality audits (Lot Quality Assurance Sampling-LQAS), and routine cross-checking was practiced by 59.7% of facilities, while 67.7% of them conducted external supervision and validation, showing inconsistent data quality assurance practices ( Figure 3).

The qualitative findings indicate that though data quality assurance processes such as LQAS and Routine Data Quality Assessment (RDQA) are believed to support decision-making, their implementation varies across the facilities. Data reliability is affected by standardization gaps, even though there are improvements due to training and supervision. Routine data quality checks and promoting data use were identified as key to sustaining improvements in data quality.

Challenges in Data Quality

Major challenges affected data quality, and the utilization of HIV, TB, Malaria, and RSSH programs for decision-making at health facility and woreda health office levels were identified through the assessment. These challenges include a shortage of skilled staff for data management, limited integration between data systems, frequent system downtime, limited internet access that affects the system, delays in data reporting, and incomplete data entry in DHIS2.

Digital Analytical Practices and Capacity

The findings revealed weak data use practices, with 54.8% of facilities not conducting gap analysis, while 66% did not perform data triangulation in the past 12 months using both routine and non-routine data sources. Only 51.6% facilities conducted disaggregated or monthly data analysis, and only 4.2% of the total workforce had digital data analysis skills and competency ( Table 4).

Data Triangulation

The baseline assessment showed that data triangulation practices varied across methods, with the most common 46(74.2%) facilities performing comparisons of DHIS2 data with paper-based reports, whereas 24 (38.7%) facilities conducted verification of TB, HIV, and malaria program coverage dashboard data with facility registers, followed by 37.1% facilities performing matching of LMIS stock records with vaccine usage report ( Figure 4).

Data use practice

The qualitative findings showed that some facilities effective data use though there are problems. Woredas identified TB cases, did a root cause analysis, and conducted screening. This targeted, data-driven intervention improved case detection and reduced transmission of the disease, indicating the importance of data for decision-making. Other woredas successfully used data in malaria control, where facilities mapped cases and targeted high-burden areas with community interventions, such as sanitation campaigns, bed net distribution, and health education. Regular data reviews showed a reduction in malaria incidence. These data-driven, geographically targeted strategies improved outcomes and resource efficiency.

Barriers to Program Data Use

The participants identified multiple barriers affecting effective data use for HIV, TB, Malaria, and RSSH programs at health facility and woreda levels. These challenges include human resource capacity, infrastructure, organizational systems, and financial constraints, alongside issues of data quality and reporting timeliness.

Human resource and technical capacity gaps were evident. Many facilities depend on a single HMIS/HIT staff member, which creates vulnerability during absences or turnovers. High staff turnover results in lost expertise and disrupts data quality. Frontline health workforces often lack skills in data analysis and interpretation, limiting their use of real-time data. Computer illiteracy and limited familiarity with systems like DHIS2 and DATIM further restrict performance, while a weak culture of data use continues.

Infrastructure challenges include weak internet connectivity, frequent power outages, and computer shortages, which force reliance on paper-based systems.

Organizational problems like weak integration between systems, delayed reporting, weak feedback mechanisms, and weak coordination decrease data use. Financial issues, including insufficient budgets, supply shortages, and low support for supervision and data quality assessments, further declined general system performance.

Program data use enablers

Several enabling factors that could help with better utilization of data were pointed out by the respondents. These included the commitment of leadership, staff capability, the technology available, and external support. Leadership plays an important role in ensuring data quality and supportive supervision, among other practices that rely on the availability and utilization of data. Some staff-related enabling factors are motivation, willingness to embrace technology, access to capacity building on data utilization, quality, and computer skills. Technologies that can enable data utilization include the use of systems like DHIS2, EMR, and SmartCare, together with computers, the internet, and a power supply. Support from woreda offices and partners on training provision, experience sharing, supervision, and finance is also an important factor.

This study has limitations. The use of a smaller sample size and a purposive sampling approach limits the ability to generalize the findings to other study settings. Self-reported data may also introduce social desirability bias, particularly in responses related to data analysis, quality, and use practices. Finally, the cross-sectional study design restricts the ability to draw cause-and-effect conclusions between interventions and outcomes.