Study setting
The study was conducted from February to March, 2017 in Ho central within the Municipality of the Volta Region of Ghana. The Ho central is one of the four sub-municipals in the Ho municipality comprising of twelve health facilities with the MHD responsible for management and providing support to the sub-municipals. Three of these facilities are hospitals; the regional, municipal and a private hospital and the rest clinics, Community-based health planning and services (CHPS) and family health units. These facilities are manned by trained and skilled health personnel who manage all kinds of health-related issues. Community health nurses at the various Reproductive and Child Health units perform vaccinations in the facilities and outreach services. Data are generated at the end of vaccination, recorded in assigned books and safely kept. The availability of skilled personnel therefore affects data quality. The study was specifically conducted in two clinics and a CHPS compound located within the central municipality.
Study population and design
According to the 2017 projected populated by the Ghana Statistical Service, the population of Ho municipality was 209,161 with under-five being 23,734 [21]. Immunization data on under-five children from sampled health facilities were studied. The study was descriptive cross-sectional. This design was chosen because it was appropriate in terms of time for the study. Data of immunized children under 18 months were studied. They were examined in tally books and reports and recorded into DQS tool [18] for analysis.
Sampling of vaccines and health facilities
Selection of vaccines and study facilities began at the MHD. The three vaccines - BCG, Penta 3 and Measles 2 were selected randomly. This was done using simple random sampling where names of the twelve vaccines were written on pieces of paper and picked without replacement. Health facilities in Ho central that performed immunization and had complete data on the selected vaccines for January to December, 2015 were included. All the hospitals were excluded because they did not vaccinate against Measles 2. Any facility that had zero records of BCG, Penta 3 and Measles 2 were also excluded. Three of the nine eligible facilities were selected using simple random sampling. Selection of three health facilities was based on previous similar study [14]. This was done by writing their names on pieces of paper and picking at random without replacement. Letters A, B and C were used to represent the facilities in the analysis for the purposes of anonymity.
Definition of terms
Data accuracy: the degree to which data has attributes that correctly represent the true value of the intended attribute of a concept or event in a specific context of use [22].
Data consistency: the degree to which data has attributes that are free from contradiction and are coherent with other data in a specific context of use [22].
Discrepancy: the data from two or more sources are not consistent.
Measurement of data accuracy and discrepancy
The DQS tool is a standard tool developed by the WHO and use to determine accuracy of immunization data. It is simple to use in comparing immunization data from a source such as tally sheets data to reports submitted to health directorate by same facility and period. This tool has been used in previous studies on data quality and findings establish inconsistency between source documents such as tally books and reports extracted from same data.
Accuracy ratio is obtained by dividing tallied figure by report figure and multiply by 100 percent. The discrepancy is obtained by subtracting the accuracy ratio from 100. An accuracy ratio less than 100 percent indicates overreporting whereas underreporting occurs when accuracy ratio is greater than 100 percent [23]. Overreporting gives an accuracy figure of less than 100 percent and positive discrepancy figure indicating more data being reported to the MHD than found in the tally registers. Underreporting gives an accuracy figure more than 100 percent and a negative discrepancy figure demonstrating less data being reported to the MHD than recorded in the tally registers.
Data collection procedure
Reported data from the sampled facilities on BCG, Penta 3 and Measles 2 were first obtained from the MHD. Data on reports were read, recorded and reread for each month per vaccine to avoid errors. Visits were then made to facilities selected and data on tally sheets which are the original source documents were recounted. Tallies against the vaccines involve were recounted thrice for each month per vaccine at all facilities. The WHO DQS tool [23] was adopted and modified in excel. The data were first recorded in a note book and then transferred into the tool in excel sheet for analysis. In all three facilities involved, visits were made twice because, some of the tally books were not available on the first visit.
Data analysis
Data collected from the different levels were entered into Microsoft excel for storage and analysis. Data upon entry into excel were cross-checked thrice to avoid transcription error. A template of the WHO DQS tool was created in excel. Data for each month per vaccine and for each facility was then entered into the tool to generate accuracy ratio and discrepancy level (100 – accuracy ratio). Descriptive statistics was done with excel and presented in simple graphs and tables.
Ethical issues
Ethical clearance for the study was obtained from Ghana Health Service Ethical Review Committee (GHS-ERC) with number GHS-ERC: 64/10/2016. Permission was sought from the MHD through an introductory letter. GHS data request form at the MHD was filled before accessing reports. Letters were rewritten by the Municipal Health Director and sent to the facility in-charges who gave approval before data were accessed. Letters A, B and C were used to represent the 3 facilities for the sake of anonymity.