Introduction: Mozambique has a generalized HIV epidemic, among pregnant women, HIV prevalence is estimated at 15.8% with a vertical transmission rate of 14%, more than double global targets. We evaluate electronic national health information system (SIS-MA) performance to verify if the data flow procedures met its objectives and evaluated the prevention of mother-to-child transmission (PMTCT) surveillance system to access its attributes and usefulness. Methods: we conducted a descriptive, cross-sectional evaluation of the PMTCT surveillance system in eight facilities in Gaza and Inhambane provinces using the centers for disease control and prevention guidelines (2001). For data quality, we cross-referenced patient registries from health facilities against the SIS-MA. We also interviewed 34 health technicians, using a Likert scale, to assess the following attributes of the PMTCT surveillance system: simplicity, stability, flexibility, acceptability, timeliness and data quality, usefulness of the system and knowledge of PMTCT. Results: regarding the simplicity measure, we verified that the registry books contain more than 30 variables. The system was 83% flexible in maintaining functionality with the introduction of new health facilities in the system. The completeness of the data was 50% and concordance of data from the register book and monthly reports was 89%. Conclusion: the PMTCT SIS-MA is useful in supporting the collection, analysis, interpretation and continuous and systematic dissemination of health data that are used to define and monitor public health policies in Mozambique. However, continued efforts are needed to improve data quality to ensure that the SIS-MA can adequately monitor the PMTCT program and contribute to reduced vertical transmission.
Study design: we carried out a descriptive, cross-sectional evaluation of the PMTCT surveillance system between April and November 2016. The evaluation of the system was conducted per centers for disease control and prevention (CDC) guidelines (2001), which outline key attributes of the system including: simplicity, flexibility, data quality, acceptability, sensitivity, predictive positive value, representativeness, timeliness and stability. For this evaluation, we focused on the attributes of simplicity, flexibility, data quality, acceptability, representativeness, timeliness and stability using observation, data review (PMTCT register books, monthly summary reports and the SIS-MA database) and interviews. It was not possible to evaluate the sensitivity and the predictive positive value due to the absence of variables to evaluate these attributes in the system. Study site: the study was conducted in Gaza and Inhambane provinces, located in the southern region of Mozambique (Figure 1). We selected these provinces because they are located in the southern region, which has the highest prevalence of HIV infection among pregnant women (23.6%) in Mozambique [8]. Study population, sample size consideration and sampling technique: using a semi-structured questionnaire, we interviewed 28 maternal and child health (MCH) nurses responsible for PMTCT in the eight health facilities and eight district-level health statisticians responsible for entering data in SIS-MA. The questionnaire assessed their level of knowledge about the system and evaluated the following attributes: simplicity, flexibility, acceptability, representativeness, timeliness, stability and usefulness. We selected at convenience sample of four health facilities (HF) from Gaza and Inhambane provinces. These facilities reported the highest numbers of MCH patients in 2016 and also offered antiretroviral treatment (ART) in the antenatal consultation (ANC) and HIV exposed infants/at-risk children services [7]. For the review of the option B+ registry books and monthly SIS-MA summary reports, one month was randomly selected for each health facility from April to December 2016. Data analysis: the interviews were first transcribed. The interviews were then coded and then we analyzed knowledge and system attributes, as shown in the panel below. We used the Likert scale, a psychometric response scale used in questionnaires and opinion surveys in which the interviewees specified their level of agreement with a statement, based on a scale [9]. The Likert scale is a verified tool used as a basis to measure parameters similar to ones in this evaluation [10-12]. Data was analyzed using SPSS 16.0, descriptive analysis presented in frequency tables and for the evaluation of attributes, parameters described in a table using Microsoft Excel were used. Simplicity: structure and ease of operation of the health information system [13]. The following elements were evaluated for simplicity: number of key variables collected from the register books; number of documents to be filled out; data flow (health facility>district>province>national); number of trained personnel. Score: 1 to 8 classification: simple=2 points; complex=1 point. Final score: simple: ≥6 points; complex <6 points [11]. Flexibility: the system´s ability to adapt to changes in information needs. We observed the ability of the system to remain functional during the introduction of new health facilities and the change of trained personnel. Final score: yes: flexible or no: not flexible. Acceptability: the willingness of individuals and institutions to use the surveillance system. We evaluated the interest of the professionals participating in the system (based on the data notification rate) and the opinion of the technicians about the registration and data entry in the system. Final score: low: <70%; moderate: 70 to 89%; high: ≥90% [10]. Representativeness: the ability of the system to accurately describe the occurrence of a health-related event over time and its distribution in the population by place and person. Person: age, sex, clinical status; time: date of 1st consultation; place: province, district, health facility, residence. Final score: high: ≥85%; low: 10min – not timely); the time interval between the aggregation of the data in the monthly report and the reporting to the district statistics team (≤1 week = timely; ≥1 week = not timely); the time necessary to enter the data in the system (until 24 hours = timely; >24 hours = not timely). Final score: timely: ≥50% of stages satisfactory; mot timely: ≤50% of stages. Stability: the ability of the system to be operational when needed. We evaluate the interruption of the system from the beginning of the functionality; availability of the data in the system after the introduction of the same in the database; capacity of the SIS-MA to be operational even if it is offline. Final score: high =yes; low =no. Usefulness: an ability of the system to contribute to the prevention and control of adverse health events, including a better understanding of the public health implications of such events: we evaluated the ability of SIS-MA to support the collection, analysis, interpretation and continuous and systematic dissemination of health data. Final score: yes=useful; no= not useful [11]. Data quality: the integrity and validity of the data recorded in the system. We evaluate the number of variables/fields filled out with in completing the data. Data concordance we compared the data from the book, monthly summary and system database. Score: low <70%; moderate – 70 to 89%; high ≥90%. Knowledge of the health professionals: the knowledge of procedures for various PMTCT services: counselling; analysis of CD4, hemogram and biochemistry; the type of ART provided, prophylaxis with cotrimoxazole and isoniazid; early infant diagnosis; exclusive breastfeeding; childbirth; invitation of the partner for testing. Final score: low 0 – 2, acceptable 3 – 4 and good 5 – 7. Ethical considerations: ethical approval was obtained from the institutional committee of bioethics in health of the Faculty of Medicine and Maputo Central Hospital (CIBS FM & HCM). We obtained informed written consent from the interviewees. Participation was voluntary and anonymity was guaranteed by not indicating the interviewees´ name on the questionnaire. We coded the identification (name) of the participants so that there is a link between the code and the information.
N/A