Volume 22, No. 3/4
2001

[Table of Contents]

Measuring Self-Reported Sunburn: Challenges and Recommendations

Jean A Shoveller and Chris Y Lovato

Abstract

Sunburn is a major preventable risk associated with the development of malignant melanoma and basal cell carcinoma. It is considered a key epidemiological concept to assess in prevention research and a core component of routine behavioural surveillance and program evaluation efforts. This review examined 38 English-language survey instruments and research reports published between 1990 and 1999 that used self-report data or parent-proxy reports of sunburn outcome. A qualitative review of the instruments and reports identified several methodological issues: the conceptual and operational definitions of sunburn, the recall period, and the use of self-reports and parent-proxy reports. As there was little consistency in definitional issues or recall periods across the studies, it is difficult to meaningfully compare their findings. We examine key issues that program evaluators and researchers should consider in determining the strengths and limitations of various definitions, measures and approaches and include recommendations for measurement of sunburn and for further research.

Key Words: prevention; skin cancer prevention; surveys

Introduction

Skin cancer has been described as an emerging public health problem in North America in terms of morbidity, mortality, health care costs, and personal disfigurement. In 1999, 740 Canadians died of malignant melanoma and approximately 66,000 new cases of skin cancer were diagnosed.¹ In 2000, 47,700 new cases of melanoma were diagnosed in the United States. It is estimated that this form of cancer will kill 7,700 Americans during the upcoming year.² Cumulative exposure to ultraviolet radiation from sunlight and other sources seems necessary for the development of squamous cell carcinoma, while solar exposure received as a result of sunburn, may be more important in the development of cutaneous malignant melanoma and basal cell carcinoma, especially amongst people who may have high melanocyte density or who may be genetically predisposed.^3-10

Given the emerging magnitude of this health problem in North America,^11-14 the number of practitioners and researchers working in skin cancer prevention has increased rapidly over the past decade. Since melanoma and non-melanoma skin cancers may have distinct patterns of occurrence and etiology, this review focuses on sunburn as an important risk factor, and therefore a key outcome, associated primarily with cutaneous malignant melanoma and basal cell carcinoma.^15,16 Other outcomes, such as cumulative lifetime sun exposure, may be more relevant to the etiology of squamous cell carcinoma and may demand different measurement and prevention strategies.

Hill et al.¹⁵ suggested that a sunburn is a useful outcome for researchers and prevention program evaluators to assess because it "can be taken as an objective indication that a biologically effective dose of ultraviolet radiation (UVR) has been received, regardless of the measured amount of environmental UVR." These authors asserted that sunburn is a good "after the fact" indicator of inadequate sun protection behaviour and concluded that sunburn represented a good measure of the "UVR dose received." They also pointed out that since sunburn is at least partly under the control of the individual, it represents a worthwhile focus for those interested in evaluating prevention programs.

While increasing numbers of program evaluations, behavioural surveillance surveys and case-control studies have measured sunburn using self-reported data, concerns have been raised about the validity and reliability of these data.^16-19 A lack of standardized approaches to measuring self-reported sunburn has also prevented advancement in this area.

It is important to recognize that no single method of measurement is capable of addressing the needs of all epidemiological researchers and program evaluators working in the area of skin cancer prevention. Age at exposure to severe sunburn may be critical to understanding the etiology of melanoma, but less relevant to a program evaluation designed to improve current sun protection behaviour. To improve program planning, evaluation, and research in these areas, it is important to be able to make meaningful comparisons across studies. We need to be able to compare how sunburn rates in one community might differ from those in other regions. We must also be able to meaningfully compare results to determine if one intervention approach is more effective than another in reducing sunburn.

There is little information to guide researchers and practitioners in assessing existing self-reported measures of sunburn and comparing sunburn results across studies. There is little discussion in the literature to help practitioners and researchers select self-reported measures and study approaches that best fit their particular aims. This paper reviews examples of various definitions, questions, and study approaches used to assess self-reported sunburn. We discuss the implications of using self-reported sunburn data on reliability and validity and address some practical issues concerning measurement.

Methods

A search was conducted in various electronic databases, Medline and CAB HEALTH CD-ROM, HealthStar, CancerLit, Social Sci Search and EMBASE, to identify and retrieve relevant published literature. To meet the preliminary study eligibility criteria, the literature must have been a primary research report that included or focused on sunburn as a behavioural outcome, and/or an intervention or descriptive study associated with melanoma and basal cell carcinoma prevention or epidemiology, written or published in English between 1990 and 1999. Additional reports were located by hand-searching selected cancer prevention journals and reference lists from retrieved articles. In all but a few cases, copies of survey instruments were obtained by contacting the authors. We chose to focus our review on the past decade because during this time skin cancer has begun to emerge as an important health issue in North America and has become a more widespread issue for program planners, practitioners and researchers.

This review focuses on studies that used self-report and/or parent-proxy reports of sunburn. It does not include studies that used the term "sunburn" for measuring a person's propensity to sunburn (i.e., as a measure of phenotype), nor does it include studies that focused exclusively on knowledge or attitudes concerning sunburn. We assessed the studies independently and resolved disagreements about eligibility through discussions until consensus was reached. Information on study approaches, definitions, questions or survey items, and an assessment of the quality of the instrument (including reliability and validity, where provided) was extracted and summarized. We then conducted a qualitative synthesis of the extracted information to identify key issues related to the reliability and validity of sunburn measures.

Results

Description of studies reviewed

This paper presents a critical review of 38 published reports and unpublished survey instruments that assessed self-reported sunburn as a behavioural outcome. All studies relied on self-reports of sunburn as a primary outcome. Of the reports reviewed, 13 studies^15,20-31 were conducted with adults from the general population, 10^32-41 with adolescents or youth, 11^42-52 with parents, and four^53-56 with adult dermatology patients or individuals identified through cancer registries. Nearly half of the studies (n = 16) were conducted in the US or Canada.^{21,22,26,27,30,32,38-40,43,47,48,50-52,55} Eleven (n = 11)^{15,23,25,28,31,33,34,36,37,44,49} were conducted in Australia and New Zealand, and eleven were conducted in Europe (including the UK).^{24,29,35,41,42,45,46,53,54,56} Most studies (n = 35) were descriptive in nature and relied on cross-sectional surveys or structured interviews, although three used a case-control study design.^53,55,56 Table 1, which can also be viewed at <http://www.healthcare.ubc.ca/shoveller/home.html> summarizes the studies and instruments included in this review.

On analyzing the information extracted from each article and instrument, we identified three issues that warrant further discussion:

• conceptual and operational definitions of sunburn;
• recall periods; and
• use of self-reported data and parent-proxy reports.

Conceptual and operational definitions

No standardized conceptual or operational definition of what constitutes the presence of sunburn was widely used across the studies. In some reports and instruments, researchers appeared to assume that respondents have an implicit understanding of the concept "sunburn". For example, one survey provided no operational definition for respondents to use as a reference point, asking only if individual sunbathers had "ever had an obvious case of sunburn."²⁴ In contrast, another study was very specific in asking whether sunburns were "so severe that they produced blisters or pain lasting two or more days."²²

Sunburn has also been defined in terms of the physical characteristics associated with it, such as redness, tenderness, blistering, and peeling of skin. Skiers, in one example, were asked to select from the following categories the term that best described their worst burn:

• Red without being tender;
• Red and tender, but not blistered; or
• Red, tender and blistered.²⁶

Some studies also asked respondents to self-report on the frequency with which they had been sunburned or had received a particular type of sunburn. For example, one study asked parents to describe their children's sunburn frequency using the following: "Do your children ever get sunburnt? Never; Sometimes (once per year); Often (several times per year); and Always."⁴² A different survey assessed the frequency of painful sunburn among students in the Australian state of Tasmania using the following categories: "Never, 1 time, 2-5 times, 6+ times."³³ In these studies, respondents provided data in the form of categorical variables. In other reports and instruments, open-ended questions were used to collect self-reported sunburn data in the form of continuous variables. For example, a case-control study of female dermatology patients used an open-ended response option to assess the number of sunburns received during the past year.⁵⁶

In a study of Australian students³³ that compared data from two cross-sectional surveys conducted in 1992 and 1993, the problems of inconsistent approaches to defining self-reported sunburn type and frequency are well illustrated. As the authors note, comparisons between results from the two surveys were difficult to make because sunburn was defined and measured differently in each survey. In the 1992 survey, sunburn was defined as "pain lasting 2+ days," while in 1993, "1+ days" was the reference point. In addition, the categorical response options available for reporting frequency of sunburn differed in each survey. Girls appeared to be more likely to report higher rates of sunburn in the 1992 survey than the 1993 survey. On the other hand, boys seemed to be more likely to report higher rates of sunburn in the 1993 survey. No conclusions could be drawn on these differences, however, since they could be due to differential definition and measurement of self-reported sunburn.

Recall period

There was considerable variation in recall periods across the studies. Depending on the purpose of the study, some researchers focused on lifetime sunburn history and others focused on a specific period. For example, 12 reports and instruments included a measure of sunburn history that asked respondents to report on sunburns received during any point in their lifetimes.^{24,28,30,31,33,38,42,47,50,52,55} Two studies asked respondents to recall sunburns received at specific ages, although the categories provided varied between these two studies.^54,56 The term "sunburn history" was used frequently in published reports and instruments to refer to recollections of experiences with sunburn over the course of a defined recall period (e.g., lifetime, childhood, past year). Many studies (n = 21) asked respondents to self-report on their sunburn history using a recall period of one year or less, including previous day or weekend,^{15,20,21,25,26} previous summer or past year,^{27,29,32,35,37,38,41,46,50,54} or some combination of past year, previous summer and past weekend.^{28,33,39,43,51,55} The remainder of the reports or instruments we reviewed did not define a recall period.

Use of self-reported data and parent-proxy reports

All of the studies included in this review relied on either self-reported data or parent-proxy reports. Most studies did not describe measures of reliability or validity associated with the self-reported data. Although self-reported sunburn was rarely validated or verified in the studies reviewed for this paper, one study used a sun-behaviour diary to verify self-reported sunburn.⁵⁴ There were no significant differences between diary and survey reports in this study, although there was a tendency among those with multiple burns to underestimate the absolute number of sunburns when completing the self-report survey. The correlation between the number of sunburns reported in the diary and those reported on the survey was r = 0.60, d.f. = 40, P < 0.001. In another study, Shoveller et al.²⁷ described inconsistencies between self-reports in a national survey on protection and self-reported sunburns. That is, the self-reported prevalence of protection was much higher than would be expected given the large proportion of sunburns reported.

Use of parent-proxy data describing children's sunburns also presents problems when comparing results across studies. Eleven of the 38 reports assess children's sunburns using parent-proxy reports.^42-52 None of these studies uses comparable conceptual definitions or recall periods, which makes it difficult to accurately compare results across them. For example, Kakourou et al.⁴⁶ asked parents to estimate the number of blistering sunburns they and their children had experienced during the previous three summers. In a study of children presenting at emergency rooms, Zinman et al.⁵² administered a survey to parents to assess if their children had "ever had a blistering sunburn".

While self-report and parent-proxy report data may be of questionable validity, all of the publications included in our review relied on these measurements. It is difficult to determine whether differences in results found when comparing across studies are real or due to differences in the way questions are asked, or due to the validity of self-report data. Few authors discussed the limitations of relying exclusively on self-reported data, although Autier et al. (1995)⁵³ indicated that the lack of effectiveness of sunscreens in protecting against melanoma may be influenced by measurement error, either because they did not measure an unknown confounder or because they inaccurately assessed a variable known to influence sunburn outcomes. Intervention effects may not be observed because they do not exist. Eiser and Arnold²⁴ also argued that without independent validation or means of determining the consistency of the criteria for identifying a case of sunburn, self-reported findings should be interpreted with caution. They posited that a previous history of sunburn may reflect not only skin type, but may also be an indicator of past protective behaviour. Overall, there has been very little research to independently establish the validity and reliability of sunburn recall.

Discussion

Three major issues associated with the reliability and validity of sunburn measures were identified as a result of reviewing the 38 reports and instruments included in this study:

• conceptual and operational definitions of sunburn;
• recall periods; and
• use of self-reported data and parent-proxy reports.

There is little consistency across the studies included in this review in definitional issues or recall periods. In addition, all of the findings of the studies we reviewed rely on self-reported data or parent-proxy reports. Thus, it is difficult to meaningfully compare findings across these studies.

In studies where respondents were asked open-ended questions about the frequency with which they were sunburned and the type of sunburn they received, it is possible to derive an overall score that reflects the severity and frequency of sunburn. These data may be important in estimating overall risk since it is biologically plausible to assert that more frequent and severe sunburns could be associated with increased risk of malignant melanoma and basal cell carcinoma. This kind of information may also help program planners to tailor public health messages on sunburn prevention.

Although no single method of measurement may be able to meet the needs of both epidemiologists and program evaluators, self-reported data are central to both. Concerns about measurement of sunburn may differ significantly depending on the intent of the study. The three case-control studies^53,55,56 reviewed in this article provide useful examples of the need to develop measures tailored to the purpose of particular studies. For example, epidemiological researchers may be more likely to undertake case-control studies requiring approaches to measuring sunburn that account for the age at which sunburns occurred. Program evaluators, however, may be more likely to measure the number of sunburns received before or following an intervention to demonstrate changes in patterns of exposure or protective behaviours. As has been suggested by other researchers,^16-19 reliance on self-reported data is an important problem facing researchers and practitioners working in the area of skin cancer prevention. Since self-reported measures of sunburn are widely used to assess risk, the development of standardized approaches to measuring this outcome represents an important area for further investigation. While technology, such as the colorimeter, has been used to assess sun exposure,¹⁹ self-reported measures are likely to remain the most widely used approaches to measuring sunburn since they tend to be most feasible and cost-effective for surveillance and evaluation purposes.

Program evaluators and researchers interested in assessing sunburn outcomes face some unique methodological challenges compared with other health behaviours or outcomes. For example, data collected during the summer season on recent sunburns (e.g., during the previous weekend) may vary within the region where the data are collected due to a number of factors, including variations in the weather, ultraviolet (UV) radiation levels (monitored by Environment Canada), and altitude. Reports on recent sunburns may be more susceptible to this sort of variability than reports on sunburn outcomes received during a longer recall period, although some studies, such as program evaluations, may require more time-specific information.

The seasonal nature of sun-related behaviours in many parts of the world presents additional challenges to skin cancer prevention researchers and practitioners. Deciding the best time of year to collect behavioural outcome data becomes important, particularly in locations such as Canada, the UK and the northern US where sun exposure tends to be seasonal. In most of North America, behaviours that are typically associated with sunburn are highest during the peak UV season during the summer or early autumn. Some researchers have attempted to collect data closer to the summer months, when sunburn is most likely to occur; however, there is little consensus among researchers on the best time of year to do so. Since response rates to large, population-based surveys are typically lower during the summer months, researchers who are interested in estimating the prevalence of sunburn at the population level should collect data in early September, but not beyond late October.

Ideally, researchers and practitioners should be able to compare results across studies with some degree of confidence. We found that the wide variation of measurement approaches to assessing sunburn made such comparisons difficult. Of the 38 studies we reviewed, we were able to identify only three that were sufficiently similar in approach to permit this kind of comparison. In Table 2, we compare reported rates of sunburn across these three studies. Each study used telephone interviews with large probability samples of adults living in Canada, Australia or New Zealand and used comparable questions to assess the frequency and severity of sunburn. It is important to note the methodological differences between these studies, including the timing of the data collection and differences in recall periods. The Canadian survey was completed in four consecutive weeks during early autumn and relied on recall of sunburn experienced several months before the survey, while the New Zealand and Australian studies were completed on a weekly basis over the summer months and asked respondents to report on sunburns received during the previous weekend. Nevertheless, these three studies warrant discussion since they were all conducted with large probability samples and their results have influenced intervention approaches in their respective jurisdictions.

Two studies conducted in the southern hemisphere found similar rates of sunburn using similar items and recall periods (e.g., previous weekend) to assess sunburn. In comparison, the Canadian study used an item that focused on sunburns received during the previous year and found higher rates of sunburn than the studies conducted in Australia and New Zealand. It is plausible to assert that differences in reported sunburn prevalence may arise because Australia and New Zealand have in place long-standing and aggressive prevention programs, whereas Canada has only recently begun to address skin cancer prevention. However, differences in reported sunburn prevalence rates between the studies conducted in Australia/New Zealand and Canada also may be due to differences in recall period and/or the unreliability of self-reports.

Recommendations for measuring sunburn and for further research

What should program evaluators and researchers consider in determining the strengths and limitations of various definitions, measures and approaches to assessing sunburn or examining the results of studies assessing sunburn? The most critical issue to consider is how the data will be used (e.g., program evaluation, population-level behavioural risk factor surveillance, or case-control study). Program evaluators are likely to need data that focus on individual episodes of sunburn during an intervention period. In contrast, those conducting surveillance research require questions that yield data pertaining to population estimates of prevalence patterns that can be compared over time to assess shifts in behaviour across an entire population. Reports on sunburn outcomes over a more extended recall period, such as the previous summer, may be sufficient for the purposes of program planning. Alternatively, researchers conducting case-control studies require measures that generate information on the frequency and severity of previous critical incidents of sunburn during specific age periods (e.g., less than 18 years of age).

We recommend that program evaluators and researchers tailor their approaches to sunburn measurement to the context within which their research is conducted and to how they plan to use the data. More standardized approaches to measurement, however, would help practitioners and researchers address some of the validity and reliability issues identified above. In Canada, we have attempted to develop a consensus on approaches to measuring sunburn to improve the comparability of results across studies. During the 1998 Canadian National Workshop on Measurement of Sun-Related Behaviours a group of practitioners and researchers working in this area developed several recommendations on measuring self-reported sunburn.¹⁶

In summary, the workshop participants made three recommendations on the assessment of sunburn for inclusion in omnibus style behaviour surveillance surveys and program evaluations. Sunburn is:

• an indirect measure of sun exposure and protection;
• important in the etiology of melanoma and basal cell carcinoma; and
• a relatively memorable and distinct event.

Sunburn was identified as the most important outcome to assess in omnibus style or program evaluation surveys, where space is often limited. The recommendations are perhaps less well suited for use in case-control studies. They are suitable for assessing sunburn outcomes using personal interviews, telephone surveys, or self-administered survey formats (see Table 3). They use a recall period of one year, since sunburns are not typically routine or frequent events, and were designed to capture data on both the frequency and severity of sunburns sustained during the previous year. In Canada, it is recommended that surveys using these recommendations are most appropriately conducted during the late summer or early autumn.

Further research is required to establish the reliability and validity of the recommendations presented in Table 3. Although few studies have attempted to validate self-reported sunburns, it is encouraging to note that one study, which used a sun-behaviour diary to verify self-reported sunburn, ascertained good correlation between self-reported items and diary entries. Future research may benefit from using this combination to measure the frequency and severity of sunburn. Additionally, research should be undertaken to develop and test self-report items that could be used in case-control studies.

Conclusion

Because sunburn is one of the most important indicators of risk for melanoma and basal cell carcinoma, it is important to improve the way this outcome is measured. Currently, a lack of standardized measurements inhibits comparison of results across studies and presents a serious barrier to progress in this area of research. Improvements in the measurement of self-reported sunburn can serve to enhance the overall quality of data collected during routine behavioural surveillance and program evaluation efforts. By collecting better quality data, researchers, planners and evaluators can work together more effectively on program and policy strategies to prevent skin cancer.

Acknowledgements

This research was conducted in part for the 1998 Canadian National Workshop on Measurement of Sun-Related Behaviours, which was supported by funding from Health Canada and the Terry Fox Workshop Program (administered by the National Cancer Institute of Canada). We wish to acknowledge the participants in the 1998 National Workshop on Measurement of Sun-Related Behaviours as well as the collegiality of those researchers who shared their unpublished reports and instruments. We also want to acknowledge the assistance of Ms. Ann-Louise Elwood, Ms. Wendy Klein and Ms. Laura Villeneuve in their gathering of published articles and original survey instruments.

References

1. National Cancer Institute of Canada: Canadian Cancer Statistics 1999. Toronto, Canada, 1999.

2. American Cancer Society. Cancer Statistics. www.cancer.org

3. Gallagher R, MacLean D, Yang P, Coldman A, Silver H, Spinelli J, Beagrie M. Suntan, sunburn, and pigmentation factors and the frequency of acquired melanocytic nevi in children. Similarities to melanoma: The Vancouver mole study. Arch Dermatology 1990;126:770-6.

4. Rosso S, Zanetti R, Martinez C, Tormo MJ, Schraub S, Sancho-Garnier H, Franceschi S, Gafa L, Perea E, Navarro C, Laurent R, Schrameck C, Talamini R, Tumino R, Wechsler J. The multicentre south European study "Helio": Different sun exposure patterns in the etiology of basal cell and squamous cell carcinoma of the skin. Br J Cancer 1996;73:1447-54.

5. Whiteman DC, Parsons PG, Green AC. Determinants of Skin. Arch Demaltol Res 1999;291(9):511-6.

6. Whiteman DC, Parsons PG, Green AC. P53 Expression and Risk Factors for Cutaneous Melanoma: a Case-Control Study. Int J Cancer 9-11-1998;77(6):843-8.

7. Consensus Development Panel: National Institutes of Health summary of the consensus development conference on sunlight, ultraviolet radiation, and the skin. J Amer Acad Dermatol 1991;24:608-12.

8. Gibbons L, Anderson L. (Eds.) Proceedings of the Symposium on Ultraviolet Radiation-related Diseases. Chronic Dis Can 1992;13 (Suppl 5):S7.

9. Health Canada. Report from Second Symposium on Ultraviolet Radiation-related Diseases. Vancouver, May 1996.

10. Elmets C, Mukhtar H. Ultraviolet radiation and skin cancer: Progress in pathophysiologic mechanisms. Prog Dermatology 1996;30:1-6.

11. Muir C, Waterhouse J, Mack T. Cancer incidence in five continents. Lyons, France. International Agency for Research on Cancer, 1987.

12. Gallagher RP, Hill GB, Bajdik CD, et al. Sunlight exposure, pigmentary factors, and risk of nonmelanocytic skin cancer. I. Basal Cell Carcinoma. Arch Dermatol 1995;131:157-63.

13. Gallagher RP, Hill GB, Bajdik CD, et al. Sunlight exposure, pigmentary factors, and risk of nonmelanocytic skin cancer. II. Squamous Cell Carcinoma. Arch Dermatol 1995;131:164-9.

14. Rivers JK. Melanoma. Lancet 1996;347:803-7.

15. Hill D, White V, Marks R, Theobald T, Borland R, Roy C. Melanoma prevention: behavioral and non-behavioral factors in sunburn among an Australian urban population. Prev Med 1992;21:654-9.

16. Lovato C, Shoveller J, Mills C, and Expert Panel. Report on the 1998 Canadian National Workshop on Measurement of Sun-Related Behaviours. Chronic Dis Can 1999;20:96-100.

17. Buller DB, Borland R. Skin cancer prevention for children: a critical review. Health Educ Behav 1999;26:317-43.

18. Dwyer T, Blizzard L, Gies PH, Ashbolt R, Roy C. Assessment of habitual sun exposure in adolescents via questionnaire - a comparison with objective measurement using polysulphone badges. Melanoma Res 1996;6:231-9.

19. Eckhardt L, Mayer JA, Creech L, Johnston MR, Lue KJ, Sallis JF, Elder JP. Assessing children's ultraviolet radiation exposure: The potential usefulness of a colorimeter. Am J Public Health 1996;86:1802-4.

20. Baade P, Balanda K, Lowe J. Changes in skin protection behaviors, attitudes and sunburn in a population with the highest incidence of skin cancer in the world. Cancer Detect Prev 1996;20:566-75.

21. Campbell H, McGregor S, Medlicott T, Goos T. Report on pilot test of sun behaviour telephone survey. Calgary, Canada. Alberta Cancer Board, 1992.

22. Dennis L, White E, Lee J, Kirstal A, McKnight B, Odland P. Constitutional factors and sun exposure in relation to nevi: a population-based cross-sectional study. Am J Epidemiol 1996;143:248-56.

23. Douglass H, McGee R, Williams S. Sun behavior and perceptions of risk for melanoma among 21-year-old New Zealanders. Aus N Z J Public Health 1997;21:329-34.

24. Eiser J, Arnold B. Out in the midday sun: risk behaviour and optimistic beliefs among residents and visitors on Tenerife. Psychol Health 1999;14:529-44.

25. Hill D, White V, Marks R, Borland R. Changes in sun-related attitudes and behaviors and reduced sunburn prevalence in a population at high risk of melanoma. Eur J Cancer Prev 1993;2:447-56.

26. Leinweber C. Report on spring skiing survey with telephone follow-up. Calgary, Canada. Alberta Cancer Board, 1994.

27. Shoveller JA, Lovato CY, Peters L, Rivers JK. Canadian National Survey on Sun Exposure & Protective Behaviours: adults at leisure. Cancer Prevention & Control 1998;2:111-6.

28. McGee R, William S, Cox B, Elwood M, Bulliard J. A community survey of sun exposure, sunburn and sun protection. NZ Med J 1995;108:508-10.

29. Melia J, Bulman A. Sunburn and tanning in a British population. J Public Health Med 1995;17:223-9.

30. Newman W, Woodruff S, Agros A, Mayer J. A survey of recreational sun exposure of residents of San Diego, California. Am J Prev Med 1996;12:186-94.

31. Stender I, Lock-Andersen J, Wulf H. Sun-protection behavior and self-assessed burning tendency among sunbathers. Photodermatol Photoimmunol Photomed 1996;12:162-5.

32. Banks B, Silverman R, Schwartz R, Tunnessen W. Attitudes of teenagers toward sun exposure and sunscreen use. Pediatrics 1992;89:40-2.

33. Blizzard L, Dwyer T, Ashbolt R. Changes in self-reported skin type associated with experience of sunburning in 14-15 year old children of northern European descent. Melanoma Res 1997;7:339-46.

34. Broadstock M, Borland R, Hill D. Knowledge, attitudes and reported behaviours relevant to sun protection and suntanning in adolescents. Psychology and Health 1996; 11:527-39.

35. Hughes B, Altman D, Newton J. Melanoma and skin cancer: evaluation of a health education programme for secondary schools. Br J Dermatol 1993;128:412-7.

36. Lowe J, Balanda K, Gillespie, A, Del Mar C, Gentle A. Sun-related attitudes and beliefs among Queensland school children: the role of gender and age. Aus J Public Health 1993;17:202-8.

37. McGee R, Williams S. Adolescence and sun protection. N Z Med J 1992;105:410-3.

38. Oliphant J, Forster J, McBride C. The use of commercial tanning facilities by suburban Minnesota adolescents. Am J Public Health 1994;84:476-8.

39. Reynolds K, Blaum J, Jester P, Weiss H, Soong S-J, DiClimente R. Predictors of sun exposure in adolescents in a southeastern US population. J Adolesc Health 1996;19:409-15.

40. Robinson J, Rigel D, Amonette R. Trends in sun exposure, knowledge, attitudes, and behaviors: 1986-1996. J Am Acad Dermatology 1997;37:179-86.

41. Wichstrom L. Predictors of Norwegian adolescents' sunbathing and use of sunscreen. Health Psychol 1994;13:412-20.

42. Bourke J, Graham-Brown R. Protection of children against sunburn: a survey of parental practice in Leicester. Br J Dermatol 1995;133:264-6.

43. Hall H. Sun protection for children - parent questionnaire (unpublished survey instrument). Atlanta, US. US Centers for Disease Control, 1998.

44. Harrison S. Sun exposure and melanocytic naevi in young Australian children. Lancet 1994;344:1529-32.

45. Jarrett P, Sharp C, McLelland J. Protection of children by their mothers against sunburn. BMJ 1993;306:1448.

46. Kakourou T, Bakoula C, Kavadias G, Gatos A, Bilalis L, Krikos X, Matsaniotis N. Mothers' knowledge and practices related to sun protection in Greece. Pediatr Dermatol 1995;12:207-10.

47. Lescano C, Rodrigue J. Skin cancer prevention behaviors among parents of young children. Child Health Care 1997;26:107-14.

48. Maducdoc L, Wagner R, Wagner K. Parents' use of sunscreen on beach-going children: the burnt child dreads the fire. Arch Dermatol 1992;128:628-9.

49. McGee R, Williams S, Glasgow H. Sunburn and sun-protection among young children. J Pediatr Child Health 1997;33:234-7.

50. Miller Dr, Geller AC, Lew RA, Koh HK. The Falmouth Safe Skin Project: Evaluation of a community program to promote sun protection in youth. Health Educ & Behav 1999;26:369-84.

51. Rodrigue J. Promoting healthier behaviors, attitudes, and beliefs toward sun exposure in parents of young children. J Consul Clin Psychol 1996;64:1431-6.

52. Zinman R, Schwartx S, Gordon K, Fitzpatrick E, Camfield C. Predictors of sunscreen use in childhood. Arch Pediatr Adolesc Med 1995;149:804-7.

53. Autier P, DoreJ-F, Schifflers E, Cearini J-P, Bollerts A, Koelmel K, Gefeller O, Liabeuf A, Lejeune F, Lienard D, Joarlette M, Chemaly P, Kleeberg U. Melanoma and use of sunscreen: an EORTS case-control study in Germany, Belgium and France. Int J Cancer 1995;61:749-55.

54. Brandberg Y, Sjoden P, Rosdahl I. Assessment of sun-related behaviour in individuals with dysplastic naevus syndrome: a comparison between diary recordings and questionnaire responses. Melanoma Res 1997;7:347-51.

55. Holly E, Aston D, Ahn D, Kristiansen J. Cutaneous melana in women: II. Phenotypic characteristics and other host-related factors. Am J Epdemiol 1995;141:934-42.

56. Westerdahl J, Olsson H, Masback A, Ingvar C, Jonsson N, Brandt L, Jonsson P, Moller T. Use of sunbeds or sunlamps and malignant melanoma in southern Sweden. Am J Epidemiol 1994;13:412-20. 

Author References

Jean A Shoveller, Centre for Community Health and Health Evaluation Research, BC Research Institute for Children's and Women's Health, and the Department of Health Care and Epidemiology, University of British Columbia

Chris Y Lovato, National Cancer Institute of Canada and the Department of Health Care and Epidemiology, University of British Columbia

Correspondence: Dr. Jean Shoveller, Centre for Community Health and Health Evaluation Research and the Department of Health Care and Epidemiology, University of British Columbia, 5804 Fairview Avenue, Vancouver, BC V6T 1Z3; Fax: (604) 822-4994; E-mail: jshovell@interchange.ubc.ca

[Previous] [Table of Contents] [Next]