The policy evaluation study was carried out in August 2013, using in-depth assessment of the policy process and outcomes in Ghana, including agenda-setting, policy development, implementation, context, content and actors (local and international) for the policy as well as the policy outcomes [20]. Twenty-seven (27) semi-structured interviews were conducted with policy makers in Trade, Health and Agriculture; Implementers in Ghana Standards Authority, Ghana Food and Drugs Authority, Animal Extension and Customs Excise and Preventive Service (CEPS); Traders and importers; and Agricultural producers. Also, broad contextual questions regarding the historical trade policy context, changes in food prices and agricultural production, and the perceived effect of food imports on agricultural production were delved into [20].
Study design and population
Data for this study were collected between February to April 2015 using a cross sectional qualitative design among adults, aged 18 years and above living in suburbs of Kumasi Metropolis, the second largest city of Ghana. This study sought to assess participants’ awareness of the policy, what they thought the general awareness of the Ghanaian public was with regards to the policy, whether they supported the policy, and their thoughts about how well the policy has been enforced and its long-term impact. Participants’ attitudes towards fatty meat in general, as well as preference for and consumption of different meat types were also assessed.
Sample size
The sample size equation [Sample size = Z1-α/22p(1-p)/d2] [21] for cross sectional study (qualitative variable) was used to determine the sample size. Where, Z1-α/2 is standard normal variable at 5% type 1 error (p < 0.05), which is 1.96, p is the expected proportion of participants aware of the restriction set at half of the population, 50% or 0.5 (as this was unknown) and d is the precision/absolute error of 5% and at type 1 error of 5%. Using the above formula, a sample size of 384 was calculated for the study, although data was eventually collected from 377 participants.
Participants recruitment
Three hundred and seventy-seven (377) participants, aged 18 years and above living in 9 suburbs (Atonsu Agogo, Anloga, Atonsu, Ayigya, Ayeduase, Ayeduase Newsite, Kotei, Kwadaso, and Oforikrom) of Kumasi were recruited by final year undergraduate students from the Kwame Nkrumah University of Science and Technology (KNUST) trained for this study. The suburbs, although purposively selected, were all big towns in Kumasi, and were chosen to give a good spread of participants from the Metropolis. In each suburb, study enumerators went from house-to-house to recruit participants. Participants were recruited on first-come-first-service basis from house-to-house visits until at least 30 participants were recruited. The selection of houses was not systematic nor random since enumerators could enter any house upon entering the suburb and move to the next house afterwards. If a house had no adult available at the time of visit, the next house was entered. The direction of the next house from the preceding one was completely dependent on the enumerators. For each house entered, adults living in the house got the aims, objectives and other information about the survey explained to them, after which they were asked to volunteer. Only one adult was chosen from each house and it was usually the first person who volunteered to take part. Apart from the 30 participants selected from each suburb of the 9 suburbs by house-to-house visits, some participants were met on the road side in the suburbs and asked to participate. Some churches and workplaces were visited and people asked to volunteer until the required sample size was attained.
Data collection
A validated structured questionnaire was adapted for this study, which had been developed and used for assessing consumer responses to the turkey tail ban in Samoa [18]. The questionnaire was interviewer administered and apart from personal information, had close-ended questions regarding awareness of the policy, support for the policy, attitude and preference for meat types, comparing imported versus local meat types. Open-ended questions were used to assess participants’ perception of the outcome and impact of the restriction. The questionnaire also included a Food Frequency Questionnaire (FFQ). The FFQ used was based on a standard layout for FFQs but was adapted to suit the local food (meat and meat products) environment of Kumasi. The FFQ used had 7 meat types (local versus imported beef, chicken, turkey tail, pork, mutton, goat meat and processed meat) and 5 options for frequency of consumption: Weekly, Fortnightly, Monthly, Occasionally and Never. The entire questionnaire and study instruments were pretested by interviewers among adults living near the University campus as part of training for data collection. The study instruments were amended following the pretest.
Assessment of body composition and blood pressure were done using the OMRON body composition analyser and a digital sphygmomanometer respectively. For the body composition assessment, participants’ height in centimeters were first measured using a SECA stadiometer and then the heights, age and gender were entered into the analyser for the assessment of body composition. The purpose of this was to know the status of these parameters among participants.
Data analysis
The data were analysed using SPSS software, version 17.0.1 (SPSS Inc., Chicago IL, USA). Statistical significance was set at p < 0.05 for all analyses. Data from the food frequency questionnaire were used to describe and compare consumption pattern of imported versus local meat types and products among the participants. All other analyses were done using simple frequencies and percentages and presented as tables.