Background: Primary health care is recognized as a main driver of equitable health service delivery. For it to
function optimally, routine health information systems (HIS) are necessary to ensure adequate provision of health
care and the development of appropriate health policies. Concerns about the quality of routine administrative data
have undermined their use in resource-limited settings. This evaluation was designed to describe the availability,
reliability, and validity of a sample of primary health care HIS data from nine health facilities across three districts in
Sofala Province, Mozambique. HIS data were also compared with results from large community-based surveys.
Methodology: We used a methodology similar to the Global Fund to Fight AIDS, Tuberculosis and Malaria data
verification bottom-up audit to assess primary health care HIS data availability and reliability. The quality of HIS data
was validated by comparing three key indicators (antenatal care, institutional birth, and third diptheria, pertussis,
and tetanus [DPT] immunization) with population-level surveys over time.
Results and discussion: The data concordance from facility clinical registries to monthly facility reports on five key
indicators–the number of first antenatal care visits, institutional births, third DPT immunization, HIV testing, and
outpatient consults–was good (80%). When two sites were excluded from the analysis, the concordance was
markedly better (92%). Of monthly facility reports for immunization and maternity services, 98% were available in
paper form at district health departments and 98% of immunization and maternity services monthly facility reports
matched the Ministry of Health electronic database. Population-level health survey and HIS data were strongly
correlated (R = 0.73), for institutional birth, first antenatal care visit, and third DPT immunization.
Conclusions: Our results suggest that in this setting, HIS data are both reliable and consistent, supporting their use
in primary health care program monitoring and evaluation. Simple, rapid tools can be used to evaluate routine
data and facilitate the rapid identification of problem areas