RESEARCH
Paraffin-related injury in low-income South African communities: knowledge, practice and perceived risk
Traumatismes liés à l'utilisation de paraffine parmi les communautés à faible revenu d'Afrique du Sud: connaissances, pratiques et perception du risque
Lesiones asociadas a queroseno en comunidades de ingresos bajos de Sudáfrica: conocimientos, prácticas y riesgo percibido
David C SchwebelI,*; Dehran SwartII; Siu-kuen Azor HuiI; Jennifer SimpsonI; Phumla HobeII
IDepartment of Psychology, University of Alabama at Birmingham, 1300 University Boulevard, CH415, Birmingham, AL 35294, United States of America
IIThe Paraffin Safety Association of Southern Africa, Cape Town, South Africa
ABSTRACT
OBJECTIVE: To explore what individuals at risk of injury from using paraffin (also known as kerosene) know about paraffin safety, what they do to protect themselves and their families from paraffin-related injury, and how they perceive their risk for such injury. Also, to explore interrelations between these factors and age, sex, education and income.
METHODS: A sample of 238 individuals was randomly recruited from low-income housing districts near Cape Town, South Africa in 2007. Trained research assistants interviewed participants to explore their knowledge about paraffin-related safety and their perceived risk of injury from using paraffin. Researchers inspected participants' homes to evaluate paraffin safety practices. Descriptive and correlational analyses were conducted.
FINDINGS: Participants had relatively low levels of knowledge about paraffin-related safety. They had high levels of unsafe practice and their perceived risk of injury was moderate. Knowledge of paraffin safety and safe practices were positively correlated with each other. Greater knowledge showed a negative correlation with the perception of being at risk for injury, but safe practices showed no correlation with perceived risk of injury. Formal education, the number of children in the home and frequency of paraffin use were positively correlated with knowledge but not with safe practices. The only significant correlate to safe practices was greater income, perhaps a reflection of the impact of financial resources on paraffin safety practices.
CONCLUSION: To develop successful paraffin safety interventions, it is necessary to understand baseline levels of knowledge, practice and perceived risk of injury among at-risk populations. Our findings could be of value for designing interventions that will increase knowledge, improve safe practices and lead to the accurate perception of the risk of injury from using paraffin.
RÉSUMÉ
OBJECTIF: Étudier ce que les individus exposés à un risque de traumatisme en raison de l'utilisation de paraffine (également appelée kérosène) savent sur la sécurité de ce produit, ce qu'ils font pour se protéger eux-mêmes et leur famille de ce type de traumatisme et comment ils perçoivent le risque associé. L'étude vise aussi à étudier les interactions entre ces facteurs et l'âge, le sexe, l'éducation et le revenu.
MÉTHODES: Un échantillon de 238 individus a été recruté en 2007 de manière aléatoire dans des quartiers pauvres proches de la ville du Cap en Afrique du Sud. Des assistants de recherche formés ont interrogé les sujets de l'étude pour déterminer ce qu'ils savaient sur la sécurité de la paraffine et comment ils percevaient le risque de traumatisme associé à l'utilisation de ce produit. Les chercheurs ont réalisé des inspections au domicile des sujets pour évaluer les précautions prises à l'égard de la paraffine. Des analyses descriptives et des analyses de corrélation ont été menées.
RÉSULTATS: Les sujets de l'étude avaient relativement peu de connaissances sur la sécurité de la paraffine. Ils appliquaient fréquemment des pratiques dangereuses et leur perception du risque de traumatisme était limitée. La connaissance des notions de sécurité concernant la paraffine et celle des pratiques sans risque d'utilisation de ce produit étaient corrélées positivement entre elles. Il existait une corrélation négative entre la possession de connaissances étendues et la perception d'être exposé à un risque de traumatisme, mais l'application de pratiques sans risque n'était pas corrélée à la perception d'un risque de traumatisme. On relevait une corrélation positive entre l'éducation formelle des sujets, le nombre d'enfants au domicile et la fréquence d'utilisation de la paraffine d'une part, et le niveau de connaissance d'autre part, mais non entre ces paramètres et la mise en œuvre de pratiques sans risque. Le seul paramètre corrélé significativement à l'application pratiques sans risque était le niveau de richesse, ce qui reflète peut être l'impact des moyens financiers sur les pratiques d'utilisation de la paraffine.
CONCLUSION: Pour mettre au point une intervention réussie en faveur de l'utilisation sans risque de la paraffine, il faut connaître les niveaux de départ en termes de connaissances, de pratique et de perception du risque de traumatisme parmi les populations à risque. Nos résultats pourraient être utiles à la conception d'interventions visant à renforcer les connaissances, améliorer la sécurité des pratiques et affiner la perception du risque de traumatisme dans l'utilisation de paraffine.
RESUMEN
OBJETIVO: Determinar los conocimientos de las personas con riesgo de sufrir lesiones por uso de queroseno acerca de los problemas de seguridad asociados; las medidas que toman para protegerse a sí mismos y a su familia de las lesiones relacionadas con el queroseno, y su manera de percibir el riesgo que corren de sufrir tales lesiones. Se analizó además la relación entre esos factores y la edad, el sexo, el nivel de instrucción y los ingresos.
MÉTODOS: En 2007 se reclutó aleatoriamente a una muestra de 238 habitantes de distritos de viviendas sociales cercanas a Ciudad del Cabo, Sudáfrica. Ayudantes de investigación debidamente adiestrados entrevistaron a los participantes para determinar su nivel de conocimientos sobre las medidas de seguridad que requiere el queroseno y su percepción del riesgo de sufrir lesiones como consecuencia de su uso. Los investigadores inspeccionaron los hogares de los participantes para evaluar las prácticas de seguridad relacionadas con el queroseno. Se realizaron análisis descriptivos y de correlación.
RESULTADOS: Los participantes tenían unos conocimientos relativamente escasos sobre las medidas de seguridad que exige el queroseno. Presentaban una alta frecuencia de prácticas peligrosas y un nivel moderado de riesgo percibido de sufrir lesiones. El conocimiento de las normas de seguridad para el queroseno y la observancia de prácticas seguras estaban correlacionados positivamente. La posesión de unos mayores conocimientos sobre el tema estaba correlacionada negativamente con la percepción del riesgo de sufrir lesiones, pero no había ninguna correlación entre la percepción y las prácticas seguras. El nivel de escolaridad, el número de niños en el hogar y la frecuencia del uso de queroseno estaban relacionados positivamente con los conocimientos, pero no con las prácticas seguras. La única correlación significativa con esas prácticas fueron unos mayores ingresos, quizá como reflejo de la repercusión de los recursos financieros en las prácticas de seguridad relacionadas con el queroseno.
CONCLUSIÓN: Si se desea desarrollar intervenciones eficaces de fomento de la seguridad del manejo del queroseno, es necesario comprender los niveles basales de los conocimientos al respecto, las prácticas y el riesgo percibido de sufrir lesiones entre las poblaciones en riesgo. Nuestros resultados podrían ser de utilidad para diseñar intervenciones que amplíen los conocimientos y propicien tanto unas prácticas más seguras como una percepción precisa del riesgo de sufrir lesiones por uso de queroseno.
Introduction
Paraffin, also known as kerosene, is among the most frequently used fuels for cooking, lighting and heating among people living in low-income communities lacking electricity in low- and middle-income countries. Paraffin is relatively inexpensive to produce and consume and a wide range of appliances are equipped to burn it, but it can endanger human health from burns and poisoning.
Risks
Paraffin is highly flammable and can lead to fires originating from malfunctioning paraffin appliances; improper placement of appliances (e.g. near curtains or tablecloths); use of polluted paraffin (since paraffin is highly explosive when mixed with water or other fuels); use of paraffin by children or intoxicated individuals; homicide, assault, suicide or arson attempts; and many other causes. Fires can be particularly dangerous in low-income urban communities, where dwellings are densely located and constructed of flammable materials such as wood and cardboard.
Young children are at greatest risk of paraffin poisoning. Because paraffin has the consistency and appearance of water and in some places is stored in reused beverage containers without child-resistant caps, unsupervised children are at high risk of consuming it.
Epidemiologists have long documented the risk of paraffin-related burns and poisoning. The earliest reports came from the United States of America in the 1940s;1,2 however, as electricity became more widely available in that country and paraffin use dropped, poisonings and burn injuries decreased. In the 1980s, reports of paraffin-related injury were published based on data from Barbados3 and Saudi Arabia.4 Over the past 15 years, epidemiological data from Egypt,5,6 India,7,8 the Islamic Republic of Iran,9 Jordan,10 Kenya,11 Malawi,12 South Africa13 and Sri Lanka14 have shown that the risk may be greatest in Africa and southern Asia. In most cases, paraffin is cited as particularly dangerous to children; for example, it accounts for about 60% of paediatric poisonings in Kenya11 and South Africa.15
Beliefs and behaviours
Little is known about the beliefs and behaviour of individuals at risk for paraffin injury. To develop safety-related interventions, it is critical to understand what people know about safety in a particular domain, how serious they perceive the risks to be, and what they do to protect themselves and their families from those risks. According to theories of health-related behaviour change, individuals change their behaviour only when they understand the existing risk, perceive it to be alterable, perceive few and manageable barriers to changing the risk, and believe the benefits to reducing it are worthwhile.16 The most detailed information to date about these aspects of belief and cognition in populations at risk for paraffin-related injury was obtained through focus groups with parents and childcare professionals in low-income South African communities.17 Participants recognized that the risks for injury were multifaceted. However, parents tended to focus on environmental risks, whereas childcare professionals suggested that poor parental supervision was to blame for many children's injuries. Safety-related knowledge and practice were not carefully explored.
The present study
This study was designed to extend the findings from the South African study17 by posing four questions. First, what do at-risk individuals know about the risks of paraffin-related injuries? Do they know what to do to protect their families from paraffin-related burns and poisoning? Second, do at-risk individuals perceive a significant risk of injury, and how do they reduce that risk in their homes? Third, do individuals with greater knowledge and safer practices perceive a greater risk of injury? Finally, do basic demographic traits - age, sex, education and income - influence knowledge, practice and risk perception in people at risk for paraffin injury? Although a few successful paraffin safety interventions have been implemented,18,19 we believe that information about knowledge, safety and perceived risk among at-risk populations would be valuable in developing empirically supported intervention programmes.
Methods
Setting and participants
Participants were recruited from 2 July to 18 September 2007 from two low-income housing districts near Cape Town, South Africa - Phillipi (on the Cape Flats, a plateau area located east of Cape Town) and Du Noon (in the northern suburbs). These communities were chosen because they were receptive to the research and because they were at a distance from one another, so they represented two different subsamples near Cape Town. Each community had about 5000 inhabitants, and neither had any known history of organized paraffin safety interventions.
Data were collected verbally during visits to the participants' homes by teams of two trained researchers working together, at least one of whom spoke Xhosa, the primary language in the community. Homes were randomly selected within lower-income sectors of the communities, where paraffin usage was higher. Since women are more likely to be children's primary caregivers (and protectors from injury), the female head of the household was asked to participate. If she was not available or refused consent, the male head of the household was asked to participate, followed by other adults in the home. One adult in each household participated in the study; in the rare instance in which the head of the household was a minor (under 18 years of age by South African law), the household was deemed ineligible.
The consent and participation rate among the randomly selected homes was 95.2% (238/250 homes). As shown in Table 1, the sample was 81.5% female, had a mean age of 30.2 years (range: 18-65) and had a mean of 9.3 years of formal education (range: 0-12). Many homes (65.3%) included children, and the number of children in those homes ranged from 1 to 7 (mean: 2.0; standard deviation, SD: 1.2). All participants identified themselves as black apart from one, who identified himself as "coloured", a term that usually refers to mixed black and white ancestry in South Africa. (Note: The formal division of South African citizens into racial categories was repealed by the national Parliament in 1991, and the authors of this article strongly support equality between individuals of all racial and ethnic backgrounds.) Participants were given the option to complete the research in English, Afrikaans or Xhosa; all chose Xhosa except for two, who chose English. Materials were translated using standard translation and back-translation procedures.
After the study was completed, participants were given a small financial compensation (about which they were not informed until the end of the study for ethical reasons). All research was approved by appropriate institutional review boards, and all participants gave informed consent to participate.
Measures
Demographics
Participants responded to a brief demographic survey.
Knowledge assessment
We developed and used an 18-item questionnaire to assess knowledge about the dangers of paraffin, the ways to prevent paraffin-related injury, and the first aid to be applied in response to paraffin-related injury. Because some participants were not fully literate, the questionnaire was administered orally. The number of correct items was summed to obtain a score ranging from 0 to 18 (for some items, "partial" credit - e.g. 0.5 points - was awarded for answers that were correct but not optimal).
Home inspection
To assess paraffin safety practices, researchers thoroughly inspected participants' homes and collected information on 17 practices related to how paraffin, paraffin appliances and paraffin-related materials were stored and placed in the household. This inspection, to which all participants consented, was conducted at the same time as other data collection and by the same trained research team. The researchers used a check sheet on which they checked "yes" or "no" for each safety practice. Sample items include the type of vessel(s) used to store paraffin in the home (e.g. jugs with child-resistant caps versus reused beverage bottles); whether paraffin appliances were on stable surfaces out of reach from young children; and whether rooms where paraffin appliances operated were properly vented. When information could not be obtained through visual inspection (e.g. if paraffin was stored in closed cabinets), researchers queried participants to gather more information. All items were objective and straightforward to score. For every item, 1 point was given for the presence of a safe situation and 0 points for its absence; points were then summed to yield a total score ranging from 0 to 17.
During the inspection, researchers also collected basic information about the physical structure of the home (e.g. number of rooms and beds).
Perceived risk
Perceived vulnerability to paraffin-related injury was explored through three items. Perceived fire risk was queried through the item, "In your opinion, what is the likelihood that a fire might occur in your neighbourhood over the next 6 months?" Perceived poisoning risk was queried through the item, "What is the likelihood, in your opinion, that a child might discover and accidentally drink paraffin in your home?" Both those items were answered on a 7-point Likert scale that ranged from "impossible" to "definite". Finally, participants were asked, "What do you think is the leading cause of death for toddlers in your community?"
Statistical analyses
Statistical analyses proceeded in three steps. First, we examined the descriptive data surrounding paraffin-related safety knowledge and practice and the perceived risk of paraffin-related injury. Second, we considered bivariate relations between knowledge, practice and perceived risk using Pearson correlation. Third, we determined whether knowledge, practice and perceived risk varied across demographic subgroups using Pearson correlation.
Results
Table 1 presents descriptive data. As expected, participants were predominantly female, poorly educated and living on a low income (the mean income of 2.7 on the 8-point scale reflects an average of about 1000-1500 South African rands per month, equivalent to US$ 150-200 at the time of the study). An average of 2.3 adults and 1.3 children lived in the participants' homes.
The homes themselves were modest. Almost none had running water and most were constructed of some combination of tin and wood. They averaged 2.0 rooms and 1.4 beds. Although more than 80% of participants had electricity in their home - often through illegal tapping of community electricity sources - they also reported paraffin as their primary fuel for cooking (36.6%), lighting (27.4%) and heating (95.6%). Most (92.8%) participants reported daily paraffin use, and all but one reported at least occasional paraffin use.
On the knowledge questionnaire, participants answered a mean of 11.5 of the 18 items correctly, suggesting that they knew, on average, 64% of the basic facts about paraffin safety. Scores on the paraffin-related safety practice measure, as assessed by a home inspection, were less encouraging. On average, participants had in place only 6.3 (37%) of the 17 safety practices.
Participants perceived modest risk of a fire in their neighbourhood over the next 6 months (mean: 4.8, based on a 7-point scale in which a "4" refers to "possible" and a "5" to "likely"). They perceived some risk of a child being poisoned in their home (mean: 3.6, based on a 7-point scale in which a "3' refers to "unlikely" and a "4" to "possible").
When asked what the leading cause of toddler death in the community was, participants overwhelmingly (62%) cited the poverty and filth present in their environments, with paraffin the second most frequent response (17%) (Table 2).
Table 3 shows the results of Pearson correlations between paraffin-related safety knowledge, practice and perceived risk of injury. As expected, knowledge and safe practice were positively correlated with each other; more knowledgeable individuals engaged in safer practices. The perceived risks of fire and of poisoning were also positively correlated with each other. Knowledge was negatively correlated with the perceived risk of both fire and poisoning.
In our final analysis we explored how the demographic characteristics of the sample were related to knowledge, practice and the perceived risk of paraffin-related injury (Table 4). Several statistically significant Pearson correlation results emerged. Knowledge about paraffin safety was positively correlated with years of formal education, number of children in the home and extent of use of paraffin. Surprisingly, paraffin-related safety practices did not correlate significantly with years of education, number of children in the home or paraffin usage. Instead, they were most closely related with participants' income.
The perceived risk of paraffin injury was related to only one demographic variable: the frequency of paraffin use. The perceived risk of fire and the perceived risk of poisoning were both negatively related to the frequency of paraffin use. This suggests that the more participants used paraffin, the lower their perceived risk of injury.
Discussion
Knowledge, practice and perceived risk
We chose to study knowledge, practice and perceived risk because understanding these factors is essential for developing interventions. The finding that people at risk of injury from paraffin know relatively little about paraffin safety is important because it suggests that interventions must not only change behaviour, but also educate individuals. Without basic knowledge about safe practices, people cannot possibly engage in them.
Paraffin safety practices were quite poor in the sample. This is not surprising because knowledge about safety was low, but safe practices lagged substantially behind knowledge. In other words, participants did not always practice safe behaviours even when they knew the safe thing to do. This result, similar to findings in other health-related research areas, suggests that people do not always practice healthy or safe behaviours, even when they know what they should do.20,21 Health-related behaviour is difficult to change, but paraffin safety interventions should comprise the various mechanisms that have proved effective in other domains of injury prevention and health behaviour change. For example, early paraffin safety interventions have focused on normalizing safe paraffin practice (e.g. storage of the fuel on upper shelves) and making people feel more vulnerable to the threat of injuries in their own home.18,19
Perception of risk must be a primary target for behaviour change because people are only likely to change when they feel vulnerable. We found that those at risk for paraffin-related injury perceive some degree of risk, which is promising, although perhaps still less than the actual threat, especially for fires.13 Health-related behaviour change to increase paraffin safety is feasible if the perception of risk can be built on by giving individuals the requisite knowledge and resources to make suggested changes and by convincing them that change will be beneficial.
Health risks to toddlers
When asked about the leading cause of toddler deaths in their communities, most participants mentioned poverty and filth. Data from Statistics South Africa support participants' impression. Intestinal infections are the leading cause of death among children aged 1-4 years in South Africa (23.6% of deaths in this age group), followed by influenza and pneumonia (12.8%) and then by non-natural causes, most of them unintentional injuries (9.5%).22
Interelationships
We found that individuals with a greater knowledge of paraffin safety had safer practices and perceived less risk, but that safer practices were not significantly related to changes in perceived risk. In one way, these findings are logical - one might imagine that greater knowledge leads to safer habits and to a concomitant perception of reduced risk, probably accurately so. On the other hand, the association between greater knowledge and a lower perception of risk is contrary to leading theories of health-related behaviour change, such as the health belief model.16 Such theories suggest that greater knowledge will lead to a greater perception of health-related risk. There are several possible explanations for this. First, the health belief model was developed and tested primarily in people from developed countries and may not apply as strongly to those from less developed ones, such as South Africans whose native language is Xhosa. For example, native South African populations are more likely than people from other cultures to have an external locus of control.23 If paraffin-related injuries are perceived, at least partially, to be due to external, uncontrollable causes, then safety-related knowledge may have a lesser effect on the perception of vulnerability to injury. A second possible explanation is that the health belief model was developed and tested primarily for explaining behaviour around disease prevention rather than injury prevention. It is unclear whether the model applies to injury prevention as strongly as it does to disease prevention.
Research is needed to explore more carefully the relations between knowledge, practice and perceived risk of paraffin-related injury. In particular, researchers should consider the extent to which increased knowledge leads to changes in perceived risk, and whether those changes are helpful in developing effective intervention strategies.
Demographic correlates
As expected, individuals with higher levels of formal education knew more about paraffin safety practices. In addition, individuals with greater income had safer paraffin-related habits, perhaps because financial resources are needed to implement many safety practices (e.g. using proper storage containers for paraffin; using well-functioning appliances; and having flat, elevated surfaces on which to place appliances and store paraffin). We also found positive correlations between knowledge about paraffin safety and both the use of paraffin and the number of children: that is, individuals who used paraffin more often and had more children had a better understanding of safety practices. This is encouraging because these individuals are also those at greatest risk. Unfortunately, those same variables did not correlate with safer practices, perhaps because families who use paraffin more often and who have more children are also more likely to lack financial resources and are thus unable to implement the safety-related changes they know they should adopt.
We found negative correlations between paraffin use and the perceived risk of paraffin-related injury. Individuals who used paraffin more often perceived less risk from its use. For interventions to succeed, ways of changing the perceived vulnerability of individuals to paraffin-related injury must be found. Such change might occur through the sharing of paraffin-related injury experiences by neighbours and friends, through strategies designed to increase at-risk individuals' sense of control over the perceived risks, or through changes in community norms.18
Implications for policy
Paraffin safety is a significant public health problem in South Africa and much of the world. Our results suggest that at-risk people know little about how to keep themselves safe and that they engage in many unsafe practices in connection with paraffin use. Policymakers should develop strategies to educate citizens through curricula and programmes at schools and adult education centres,24 through mass media outlets (e.g. radio advertisements, posters) and through investment in empirically supported, community-based educational campaigns.18,19
At the same time, we must recognize that it is very difficult to change the behaviour of all individuals. Therefore, there is a need for government-level interventions on the ways paraffin and paraffin-powered appliances are manufactured, processed, packaged and used. Standard SANS 1906:2006 for non-pressure paraffin appliances, which was made compulsory in South Africa in 2007, offers an example of legislation that can have a broad-ranging effect on paraffin safety.25
Conclusion
To our knowledge, this report represents the first exploration of the knowledge, practice and perceived risk of paraffin use among a sizable sample of low-income individuals who use paraffin. We discovered relatively low levels of safety-related knowledge and even lower levels of safety-related practice in our sample. These findings emphasize the urgent need for intervention development and implementation. 
Acknowledgements
We thank Phathiswa Bani, Nomhle Bentele, Tommy Coloba, Tiffany Davis, Sandile Hanisi, Busisiwe Jali, Elizabeth Kula, Makhabane, Noluthando Martins, Nomaphelo Moyeni, Tindi Mzi, Luvuyo Ngqabe, Noelene Nkondlwana, Ntombekhaya Qhalo and Nalini Sathiakumar for their assistance with the project.
Funding: This study was partially funded by a Faculty Global Health Grant from the Framework for Global Health Grant Program at the University of Alabama at Birmingham Sparkman Center for Global Health, and by the Paraffin Safety Association of Southern Africa.
Competing interests: None declared.
References
1. Deichmann WB, Kitzmiller KV, Witherup S, Johansmann R. Kerosene intoxication. Ann Intern Med 1944;21:803-23.
2. Lesser LI, Weens HS, McKey JD. Pulmonary manifestations following ingestion of kerosene. J Pediatr 1943;23:352-64. doi:10.1016/S0022- 3476(43)80016-0
3. St. John MA. Kerosene poisoning in children in Barbados. Ann Trop Paediatr 1982;2:37-40. PMID:6186189
4. Mahdi AH. Kerosene poisoning in children in Riyadh. J Trop Pediatr 1988; 34:316-8. PMID:3221417
5. El-Badawy A, Mabrouk AR. Epidemiology of childhood burns in the burn unit of Ain Shams University in Cairo, Egypt. Burns 1998;24:728-32. PMID:9915673 doi:10.1016/S0305-4179(98)00097-7
6. Mabrouk A, El Badawy A, Sherif M. Kerosene stove as a cause of burns admitted to the Ain Shams burn unit. Burns 2000;26:474-7. PMID:10812271 doi:10.1016/S0305-4179(99)00180-1
7. Ahuja RB, Bhattacharya S. An analysis of 11,196 burn admissions and evaluation of conservative management techniques. Burns 2002;28:555-61. PMID:12220913 doi:10.1016/S0305-4179(02)00069-4
8. Gupta M, Bansal M, Gupta A, Goil P. The kerosene tragedy of 1994, an unusual epidemic of burns: epidemiological aspects and management of patients. Burns 1996;22:3-9. PMID:8719308 doi:10.1016/0305-4179(95)00082-8
9. Soltani K, Zand R, Mirghasemi A. Epidemiology and mortality of burns in Tehran, Iran. Burns 1998;24:325-8. PMID:9688197 doi:10.1016/S0305- 4179(98)00017-5
10. Shotar AM. Kerosene poisoning in childhood: a 6-year prospective study at the Princess Rahmat Teaching Hospital. Neuroendocrinol Lett 2005;26:835-8. PMID:16380681
11. Lang T, Thuo N, Akech S. Accidental paraffin poisoning in Kenyan children. Trop Med Int Health 2008;13:845-7. PMID:18363584
12. Chibwana C, Mhango T, Molyneux EM. Childhood poisoning at the Queen Elizabeth Central Hospital, Blantyre, Malawi. East Afr Med J 2001;78:292-5. PMID:12002105
13. Experts forum report, June 2004. Cape Town: Paraffin Safety Association; 2004.
14. Laloë V. Epidemiology and mortality of burns in a general hospital of Eastern Sri Lanka. Burns 2002;28:778-81. PMID:12464477 doi:10.1016/S0305- 4179(02)00202-4
15. de Wet B, du Toit N, van Schalkwyk D. Paraffin (kerosene) poisoning in South Africa - some errors in a poster designed for black viewers. S Afr Med J 1994;84:722-3. PMID:7495004
16. Rosenstock IM. The health belief model and preventive health behavior. Health Educ Monogr 1974;2:354-86.
17. Munro S-A, Van Niekerk A, Seedat M. Childhood unintentional injuries: the perceived impact of the environment, lack of supervision and child characteristics. Child Care Health Dev 2006;32:269-79. PMID:16634972 doi:10.1111/j.1365-2214.2006.00593.x
18. Schwebel DC, Swart D, Simpson J, Hui S-KA, Hobe P. An intervention to reduce kerosene-related injury in low-income South African communities. Health Psychol. 2009;28:493-500. PMID:19594274 doi:10.1037/a0014531
19. Swart L, van Niekerk A, Seedat M, Jordaan E. Paraprofessional home visitation program to prevent childhood unintentional injuries in low-income communities: a cluster randomized controlled trial. Inj Prev 2008;14:164-9. PMID:18523108 doi:10.1136/ip.2007.016832
20. Morrongiello BA, Dayler L. A community-based study of parents' knowledge, attitudes and beliefs related to childhood injuries. Can J Public Health 1996; 87:383-8. PMID:9009394
21. Prochaska JO, DiClemente CC. Stages and processes of self-change of smoking: toward an integrative model of change. J Consult Clin Psychol 1983;51:390-5. PMID:6863699 doi:10.1037/0022-006X.51.3.390
22. Mortality and causes of death in South Africa, 2005: findings from death notification. Pretoria: Statistics South Africa; 2007. Available from: http://www.statssa.gov.za/publications/P03093/P030932005.pdf [accessed on 20 May 2009] .
23. Riordan ZV. Locus of control in South Africa. J Soc Psychol 1981;115:159-68.
24. [Anonymous]. Final declaration. In: Paraffin users household energy summit, Kopanong Conference Centre, Johannesburg, 28 June 2007. Cape Town: Paraffin Safety Association; 2007. Available from: http://www.paraffinsafety.org/files/2007-09-25-paraffin-users-household-energy-summit-declaration-final-version.pdf [accessed on 20 May 2009] .
25. Lloyd PJ, Truran G. Safe paraffin appliances and their contribution to demand side management. In: proceedings of the sixteenth conference on the Domestic use of energy, Cape Peninsula University of Technology, Cape Town, 18-19 March 2008. Available from: http://www.erc.uct.ac.za/Research/publications/08Lloyd-Truran_Safe_paraffin.pdf [accessed on 20 May 2009] .
(Submitted: 4 August 2008 - Revised version received: 27 January 2009 - Accepted: 1 February 2009)
* Correspondence to David C Schwebel (e-mail: schwebel@uab.edu).