Background and motivation
Informal food vending has many benefits that range from food security for the urban poor, to contribution to local economies. Yet, the street food sector is not recognized and is treated as a major public health risk. This is based partly on perceptions of local authorities that street food vendors are undesirable, or are a temporary phenomenon that will disappear in due course as a result of development. Moreover, a number of studies have documented contamination of street foods with pathogenic microorganisms like Staphylococci aureus and enteric organisms. The former can grow and express virulence in foods such as meat and meat products post cooking. To date no participatory risk analysis methods have been applied to informal markets in Tshwane, to study the food value chain of ready-to-eat (RTE) chicken and quantify the risk of exposure to selected foodborne diseases, or to determine predictors for contamination of RTE chicken sold by informal vendors.
A cross-sectional analytical study design was adopted to achieve the objectives of the present study. The study area was city of Tshwane, Gauteng Province, South Africa. The study population was informal vendors selling RTE chicken to commuters at taxi and bus stations and the sampling design involved cluster sampling. Markets were the sampling units and the informal vendors the units of concern. All clusters (n=13) that were identified were included in the study and all informal vendors located within the markets were invited to be part of the study. A total of 237 vendors agreed to be interviewed, have checklists filled in for their vending stalls, and provided a sample of RTE chicken (n= 237) for laboratory analysis. Participatory Risk Analysis, a novel approach to understanding food safety in data scarce environments, was used to collect data for analysis. Participatory research methods employed in this study included: check-lists and structured interviews with informal vendors, proportional piling and focus group discussions. Enumeration of bacterial colonies from RTE chicken samples was performed using 3MTM PetrifilmTM plates (3M, St. Paul. Mn, USA). Data for the demographic profile of vendors (n=237) was analysed using descriptive statistics to establish means and proportions. Thematic analysis combined with descriptive statistical analysis was used to establish and quantify the food value chain of informally traded RTE chicken. Stochastic risk analysis using @Risk 4.5 (Palisade Corp., Ithaca, N.Y.), was used to estimate the risk of staphylococcal food poisoning from consumption of street-vended RTE chicken. Univariable and multivariable regression models were developed using Stata 9.0 (StataCorp, College Station, Texas, USA) to determine predictors for contamination of RTE chicken with E. coli and coliform bacteria respectively.
The majority of vendors in the markets studied were females (92%), and were between 25-50 years of age. Very few young people (< 25 years) were involved. More than half of the vendors (69.47%) ran the informal food outlet as a personal business. The minority were long term employees (15.04%) or were hired to help out for a short period of time (0.88%). A high level of literacy was observed among informal vendors. Assessment of hygiene practices showed a low level of compliance with regard to possession of a certificate of acceptance, washing of hands and utensils in the same container and controlling flies at the vending site.
The environment in which informal vendors operated was not conducive to production of safe food. This means that Key 1 of the WHO five keys to safer food that requires that food production takes place in a clean environment was violated. Contamination of RTE chicken was associated with the use of water collected and supplied in containers of questionable microbiological status. This was because vendors also violated Key 5 of the WHO five keys to safer food, which requires that raw materials, including water used to prepare food should be of a high microbiological quality. However, high compliance was observed with regard to the following aspects: not wearing jewellery while preparing food, using cutlery to pick up RTE chicken, washable floors (concrete slabs and cemented floors), use of potable municipal water, preparing food in a closed structure (temporary or permanent) and locating vending stalls within 30m from the toilets.
Four food value chains for RTE chicken, sold by informal vendors, were identified showing extensive cross-over from the formal to informal sector. This extensive cross-over, was corroborated by the fact that over 79.3% of the RTE chicken sold on informal markets in Tshwane metropolis, was sourced from formal markets such as supermarkets or wholesalers. The food value chain for RTE chicken was short, meaning that it involved very few stages from farm to fork. Furthermore the tendency was for the RTE chicken to be purchased, cooked and consumed the same day.
Although the prevalence of contamination of RTE chicken with S. aureus was high (44%), the risk of staphylococcal food poisoning was estimated to be only 1.3% (90% CI: 0% - 2.7%) for each meal of RTE chicken consumed. The mean S. aureus counts in the RTE chicken was 103.6 (90% CI: 103.3 – 103.9). This level was lower than the 105 CFU/g needed to induce staphylococcal food poisoning. While contamination of RTE chicken with E. coli (6.32 %) was low, contamination with coliforms (23.21 %) was moderate. This suggests a moderate level of food hygiene. The fly population at the stalls where the RTE was on display, intermittent washing of hands, the location of the stall >30 m from the toilets and holding RTE chicken at <70 C were identified as important predictors of contamination with E. coli. Whereas, the use of potable toilets without hand washing facilities and poor hand washing practices were a positive predictor for contamination with coliforms.
The informal trade in food in Tshwane is dominated by women and for the majority of these vendors the informal trade in food, is the main source of employment. A risk communication strategy needs to take this into consideration. There is need for intervention to improve the hygiene in the informal markets and ensure that informal vendors observe the WHO five keys to safe food. The high level of literacy observed among informal vendors is good news, as it presents a situation of educated vendors; a fertile ground for initiatives aimed at improving their lives and the service they offer to their customers. There is neither a food policy aimed at ensuring food security for the urban poor nor evidence it is being implemented. This is supported by the high number of vendors operating in temporary structures. There is a need for the municipal authorities to intervene, to provide appropriate structures. Planning of new taxi ranks should include a proper food market, with facilities that meet the legal requirements for production of safe food. Informal vendors have demonstrated an awareness of the need to practice good food hygiene as proved by the compliance with certain hygienic food handling practices. The link between the formal and informal markets suggests that the informal market of RTE chicken is well-established with a reliable source of raw chicken. Furthermore, the strong inter-linkage between the formal and informal value chain for RTE chicken, further confirms the informal sector as a potential market for locally produced product. It also emphasizes its potential to impact on the economy of the city. In addition, this strong inter-linkage, suggests that the risks associated with the formal sector might be mirrored in the informal. Therefore, as the supermarket proliferation takes root in the rest of Africa, this food value chain and its associated risks might become the norm on the continent.
The present study shows that participatory risk analysis is a good way to obtain data on informal markets. Such data can then be subjected to quantitative microbial risk analysis using sophisticated biostatistical techniques, to determine and quantify the microbial risks. The low risk of staphylococcal food poisoning proved that despite the poor working conditions under which the informal food vendors prepare food, they are able to produce food that is unlikely to result in food poisoning. However, due to the high prevalence of contamination of RTE chicken with S. aureus, there is a need for the informal vendors in these markets to be trained to improve hygienic food handling practices. On the positive side, the present study demonstrated that high prevalence of contamination does not always translate into high risk. The relatively low to moderate level of contamination with E. coli and coliforms respectively, suggests a low risk of enteric foodborne diseases from ingestion of RTE chicken. Given that some and not all hygiene practices were significantly associated with contamination of RTE chicken, monitoring authorities should emphasise identified predicators to reduce the risk of contamination and the subsequent risk of contracting foodborne diseases. The difference in hygiene scoring highlights that the one-size-fits all approach to monitoring of informal markets may not be effective. These need to be tailored to suit the individual informal markets.