Study design
The study was cross-sectional in design, using mainly primary quantitative data. Additionally, secondary data on reported AEFI were reviewed.
Review of secondary data on reported AEFIs
A quantitative descriptive analysis and inferential analysis were conducted on 4,389 spontaneous AEFI reports from COVID-19 vaccines deployed in Ghana from March 2021 to December 2022 to assess the timeliness of reporting and case detection rates defined as number of AEFIs identified per 100,000 surviving infants in a population, for the different tools for reporting. The data consisted of serious and non-serious AEFI reports from five COVID-19 vaccines deployed in Ghana, namely, AZ Covishield, Sputnik V, Moderna, Covid-19 Vaccine Janssen and Pfizer-BioNTech) received at Food and Drugs Authority, Ghana between March 2021 and December 2022. All AEFI reports received at the FDA through the passive reporting pathway for COVID-19 were included. AEFI reports received through active monitoring, i.e cohort event monitoring (CEM) were excluded.
Cross sectional study
Participants for the cross-sectional survey comprised public health nurses, general nurses, midwives and disease control officers selected from Tema Metropolitan and Ga North Municipal Health Directorates of the Ghana Health Services in the Greater Accra Region.
Sampling and sample size determination
A total of 147 healthcare professionals were interviewed during a survey to assess the awareness of the use of the Med Safety App for reporting AEFIs in Ghana. Healthcare professionals who were encountered during data collection at the health facilities visited within the district were interviewed. The sample size for the survey was determined using the Cochran sample size formula 8. Cochran’s formula was used because it is preferable in a scenario where a sample of any given size may be required to provide more information about a smaller population than a larger one.
Where;
e is the desired level of precision/ acceptable sampling error (i.e. the margin of error), e = 0.05
p is the (estimated) proportion of the population that has the attribute in question, p = 0.1
q is 1 – p, q = 0.9, confidence interval of 95%, Z = 1.96
$$\:n=\frac{1.962\:\left(0.1\right)0.9}{{0.005}^{2}}$$
n = 138
A 30-item novel self-administered structured questionnaire developed using Google Forms was administered to participants; this was adapted from a study by Mehmeti et al., 2017). The Participants’ consent was obtained prior to the filling of the questionnaire with objectives of the study clearly explained to the participant (Ewudzie-Sampson 2023, Unpublished manuscript).
The questionnaire consisted of four sections:
Demographic characteristics: age, sex and place of residency and educational level.
Professional experience: information on the type of healthcare professional, length of service as a healthcare worker and experience about vaccination.
General knowledge of AEFIs and AEFI reporting in Ghana: to assess level of awareness of healthcare professionals about AEFIs vaccines and vaccines related AEFIs.
Awareness of the use of the Med Safety App for reporting AEFI: focused on assessing the level of awareness for healthcare professionals on the use of the Med Safety App for reporting AEFIs and other reporting channels for reporting AEFIs.
The knowledge of the definition of an AEFI by the healthcare professionals was assessed by providing them two options as the definition of an AEFI. For option “A”, an AEFI was defined as “any untoward medical events that follow immunization, and that do not necessarily have a causal relationship with the immunization” and option “B, “any untoward medical events that follow immunization which is caused by a vaccine after vaccination”. Per WHO, option “A” is the correct definition of an AEFI.’
Inclusion criteria
The survey involved public health nurses, disease control officers and nurses/ midwives in the Tema Municipal and Ga North Municipal Health Directorates in Greater Accra Region.
Exclusion criteria
Public health nurses, disease control officers, nurses/ midwives who were unwilling to participate in the survey were excluded. Only participants in the selected area were enrolled due to cost and convenience. Only healthcare professionals who were present at work at the time of the data collection were included; those who were absent at the time were not included.
Data Analysis
The quantitative data obtained from the survey was downloaded as a Microsoft Excel 2016 spreadsheet and cleaned to ensure accurate analysis. Analysis of data was done using Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC; 2019.
For the secondary data, reporting rates of AEFI was calculated by age groups, sex, vaccine types, and reaction categories for the different reporting tool used 10. AEFI reporting rates was calculated using doses administered as denominators. Doses administered provided a more accurate comparison of AEFI reporting rates. The reporting rate was calculated to assess trends among the reporting channels relative to the Med Safety App.
Timeliness was estimated by finding the difference between the day of reporting of AEFI by the vaccinee/ healthcare professional to the NPV Center and the onset of the AEFI. 11
In order to establish case detection rate for the various channels as part of the sensitivity measure, the study adopted the recommended WHO case detection rate of 10 per 100,000 surviving infants since there was no peculiar established rate for only the COVID-19 vaccination globally 11. This was obtained by using the total dose administered as denominator.