Prevention of skin cancer in primary healthcare: An evaluation of three different prevention effort levels and the applicability of a phototest

Abstract

Background/objective: The high skin cancer incidence in western society, and its known association with sun exposure habits, makes the area an important target for prevention. We investigated, in a primary healthcare setting, differentiated levels of prevention efforts directed at the propensity of the patient to change his/her sun habits, sun protection behaviour, and attitudes, after information intervention. Additionally, the impact of the performance of a phototest to determine individual sun sensitivity was evaluated. Methods: 308 patients visiting a primary healthcare centre in southern Sweden completed a questionnaire concerning sun habits, sun protection behaviour, and attitudes, and were randomized into one of three groups, representing increasing levels of prevention effort in terms of resources. Feedback on their questionnaire and general preventive sun protection advice was given, in the first group by means of a letter, and in the second and third groups by a doctor's consultation. Group 3 also underwent a phototest, with a self-reading assessment and a written follow-up of the phototest result. Change of sun habits, behaviour, and attitudes, based on the Transtheoretical Model of Behaviour Change and on Likert scale scorings, was evaluated after 6 months, by a repeated questionnaire. Results: Prevention mediated by a doctor's consultation had a clearly better impact on the subjects. The addition of a phototest did not further reinforce this effect in the group as a whole, but it did for a subgroup of individuals with high ultraviolet (UV) sensitivity, as determined by the phototest itself, suggesting that this might actually be a tool to improve outcome in this high-risk group. Conclusion: A personal doctor's consultation is a valuable tool in the effective delivery of preventive information in the general practice setting. In individuals with high UV-sensitivity and thus high risk for skin cancer the performance of a photo-test reinforces a positive outcome in habits, behaviour and attitudes.

• Skin cancer prevention
• phototesting
• behavioural change
• self-assessment
• questionnaire

Introduction

Exposure to ultraviolet (UV) light from the sun is associated with a spectrum of biological effects of both healthy and harmful nature in the skin, including skin cancer. Over the last half century, UV light exposure in western populations has increased, and in Sweden the annual incidence of malignant melanoma has risen from 8.2 to 27.3 cases per 100 000 inhabitants between 1970 and 2006 [1]. Prevention of skin cancer has been pursued with varying emphasis around the world and in different forms, including extensive nationwide media campaigns, efforts focusing on specific populations or risk groups, or by prevention programmes within the frameworks of local healthcare services.

Reviews of campaign/intervention studies show varying results, but at least some studies have been able to demonstrate a marked effect on sun protection behaviour [2–4]. Recurring target groups for which significant change in sun protection behaviour has been demonstrated after focused intervention/information are “beachgoers” [5–7] and school/preschool children [8], [9]. Programmes to promote early detection of malignant skin tumours have also been an important focus of efforts, since early detection improves prognosis, and several “open house” campaigns for skin examinations have led to an increased diagnosis of skin cancer in the early stages [10–13]. Some authors have focused on investigating the influence on attitudes and awareness towards sun exposure, rather than actual behaviour [14], [15]. Although the correlation between awareness and behaviour has been questioned [16–18], both components appear to be of importance for the design of skin cancer prevention.

Weinstein et al. [15] noted, in a study on children and their parents, that the most commonly reported sources of information about sun exposure and sun protection behaviour were television and magazines, but that many expressed a desire to obtain more information from their dermatologist or general practitioner. In general practice, both primary and secondary prevention within a multiplicity of disease groups comprise a considerable part of the daily workload. The steady flow of patients in combination with close patient contact and follow-up possibilities make primary healthcare an almost ideal base for prevention. Examples of tools available for prevention are written information folders, follow-up questionnaires, or oral information mediated directly by healthcare personnel, such as nurses or doctors. Actual testing of light sensitivity for larger-scale prevention efforts has not been reported.

Phototesting (testing of individuals’ reactivity threshold to UV light) is routinely used in dermatology during the investigation of photodermatoses, or as a tool for monitoring UV dose in phototherapy. A considerably less investigated field of interest is whether the performance of a phototest might increase the impact of skin cancer prevention, by actually demonstrating individual light sensitivity. In a recently performed study, we investigated the reliability of patients in performing self-reading of a phototest, and found this to be surprisingly high [19]. Patient self-reading of a phototest and reporting of test result by postcard, e-mail, or fax etc. could enable phototesting on large groups of patients/individuals, obviating the need for a review visit. Test results thus obtained could then be a source of information for personalized feedback and prevention information.

The aim of the present study was to investigate in a primary healthcare setting differentiated levels of prevention initiatives directed at skin cancer, and how the propensity of the patient to change his/her sun habits/sun protection behaviour and attitudes towards sun bathing was affected. Additionally, the impact of the performance of a phototest with self-reading by patients as a complementary tool in skin cancer prevention was investigated.

Material and methods

The study was approved by the Ethics Committee of Linkoping University, no. M187-04.

Study population

The study was performed at Kärna Primary Healthcare Centre, situated in a suburb of Linköping, southern Sweden, a population comprising suburban and rural/outer-metropolitan inhabitants, totalling about 13 500 individuals. All patients >18 years of age visiting the healthcare centre during a period of 3 weeks in February 2005 (winter in Sweden) were offered the opportunity to complete a questionnaire concerning sun habits and sun protection behaviour. The questionnaire was given to the patient at registration in the reception of the healthcare centre regardless of the cause and purpose of the visit (visits to doctors, nurses, physiotherapists, occupational therapists, and others, or for laboratory investigation). Written information about the study was enclosed with the questionnaire, and the participants gave their written consent for voluntary participation and to be contacted for follow-up of the questionnaire.

Questionnaire

The initial questionnaire consisted of four parts: demographic questions, questions on sun habits/sun protection behaviour, mapping of readiness to change behaviour based on the Transtheoretical Model of Behaviour Change, and questions concerning attitudes towards sunbathing. It constituted a set-up of questions based on experience from previously performed studies [5], [7], [18], [20–22].

The demographic part (questions 1–8) consisted of questions on age, sex, educational level, skin type according to Fitzpatrick's classification of sensitivity to sun exposure (skin type I: always burns, never tans; skin type II: always burns, sometimes tans; skin type III: sometimes burns, always tans; skin type IV: rarely burns, always tans; skin type V: ethnic groups with moderately pigmented brown skin; skin type VI: ethnic groups with markedly pigmented dark or black skin) [23], history of sun-related skin diseases (photodermatoses or skin cancer), and the presence of skin cancer in the family.

The second part (questions 9–16) contained questions on sun habits/sun protection behaviour, with answer alternatives presented on a five-point Likert scale. The questions and scores are displayed in Figure 1.

Figure 1.  Questions on sun habits/sun protection behaviour.

The third part of the questionnaire consisted of four questions (question 17a–d) addressing readiness to change behaviour, based on the Transtheoretical Model of Behaviour Change. This model was presented in 1992 by Prochaska et al. [24], [25], and has since gained wide acceptance and extensive use, both in studies on sun exposure [5], [7], [20] and in other risk behaviour situations, not least tobacco smoking [26], [27]. The theory behind the model claims that the individual is in one of five schematic stages of behaviour change, each stage of which is represented by a statement. Individuals in the first stage (pre-contemplation) have no intention to change their behaviour. In the second stage (contemplation) the individual is seriously considering changing his/her behaviour, and in the third stage (preparation) the individual has decided to change his/her behaviour. In stage four (action) the individual has taken action to change his/her behaviour, and finally in the fifth stage (maintenance) the individual works to prevent relapse and consolidate the gains attained during the action stage. The four behavioural items investigated were a) giving up sunbathing, b) using clothes for sun protection, c) using sunscreens, and d) staying in the shade during the hours of strongest sunlight. For each item, the subjects were asked to mark the statement alternative best corresponding to their own attitude, each statement representing the different stages of change. Thus, for the use of sunscreens, the choice of the statement “I have never thought of using sunscreens” represented the pre-contemplation stage; “I would consider using sunscreens” represented the contemplation stage; “I intend to start using sunscreens” represented the preparation stage; “I have started to use sunscreens” represented the action stage; and “I have used sunscreens for a long time” represented the maintenance stage. Similar statements were used for the three remaining behavioural items. At analysis, the five stages of change were scored as 1–5, from maintenance to pre-contemplation stage, respectively.

The fourth and final part (questions 18–22) concerned attitudes towards sunbathing. These questions, again with responses presented on a five-point Likert scale, are displayed in Figure 2.

Figure 2.  Questions on attitudes towards sunbathing.

Group intervention/randomization

Participants fulfilling the inclusion criteria were randomized in numerical order to one of three groups (groups 1, 2, and 3), each group intended for a specific level/form of feedback. In all groups, information and feedback were related to the individual risk profile, based on the information gained in the questionnaire, regarding skin type, sun habits/sun protection behaviour, and heredity or history of skin cancer. Sun habits/sun protection behaviour was presented as the mean value of the 5-point Likert scale scoring, for which a higher mean value represented a higher risk behaviour.

Group 1 received feedback by means of a letter, consisting of standardized comments on skin type, sun habits/sun protection behaviour mean value, and skin cancer anamnesis. The letter concluded with a summarized risk assessment with adjusted sun protection advice, including information about the importance of early detection of skin cancer. Furthermore, an additional information folder from Apoteket (the Swedish public pharmacy) was included, containing general information on sun exposure risks and practical advice concerning sun protection.

Group 2 received feedback by means of a personal doctor's consultation at Kärna Primary Healthcare Centre. Each consultation took approximately 20 minutes, and consisted of the same (in this case oral) feedback on the questionnaire as well as information and advice concerning sun habits/sun protection behaviour. The consultation also included an inspection of naevi, and all subjects received the same written information from Apoteket as those in group 1.

Group 3 was handled following the same routine as for group 2, i.e., feedback in terms of a personal doctor's consultation, but with the addition of a phototest. The phototest (Skin-tester Kit, Cosmedico Medizintechnik GmbH, Scwennigen, Germany) was applied on the palmar side of the forearm, and consisted of six fields illuminated with varying increasing UV doses. The subjects themselves then performed the test reading, after 24 hours, by counting the number of erythematous reactions. The result was written down on a protocol they had been given, which the subjects then returned by mail to the primary healthcare centre. The reading instruction stated that each visible redness should be classified as a reaction. Feedback on the phototest result and comments on the risk aspect with regard to sun exposure were finally mailed back to the subjects.

Follow-up questionnaire

After 6 months, i.e., after the following summer season, all subjects in all groups received a follow-up questionnaire. This contained the same questions as the initial questionnaire, except for the demographic part, which was excluded. In order to investigate whether the time of year for completing the questionnaire (spring or autumn) could affect the outcome, a fourth group, recruited following the same procedure as for the first three groups, received their questionnaire to complete (including demographic data) in the autumn.

Statistical analysis

Mean values based on the five-point Likert scale, and for the five scored stages of change according to the Transtheoretical Model of Behaviour Change, were calculated for each separate question, and were then compared between the initial and the follow-up questionnaires, as a measure of change of behaviour and attitudes. A paired t test for significance was performed, and adjusted for age and sex. Analysis of variance (ANOVA) was performed for comparison between the outcome of the initial questionnaire in the three test groups (groups 1–3) and in the control group (group 4), and also to elucidate possible age-, sex-, educational level-, or skin type-dependent effects, between the initial and the follow-up questionnaires in groups 1–3. Calculation of power for ANOVA (based on an estimated sample size of 95 in each group and a difference between group mean values =0.3 as being clinically significant, with a standard deviation of 1.0) gave a power of 0.82. All calculations were made in MINITAB (Minitab Inc., State College, Penn., USA).

Results

Of 330 subjects who completed the initial questionnaire, 14 were excluded because of insufficient contact information or errors in inclusion. The remaining 316 subjects were randomized to one of the three intervention groups. The capacity for the number of possible doctors’ consultations required for groups 2 and 3 was limited to a total of 200 (100 in each group), leaving 116 subjects in group 1. A small number of subjects failed to attend the doctor's consultation, and were therefore excluded, leaving 97 in group 2 and 95 in group 3. Of subjects eligible for assessment (n = 308), 189 (62%) were female and 119 (38%) were male. Five per cent were aged between 18 and 25 years, 24% between 26 and 40 years, 47% between 41 and 65 years, and 24% >65 years. The response frequency for the follow-up questionnaire was 80% in total (72% in group 1, 97% in group 2, and 83% in group 3). Group 4, which only completed a questionnaire in the autumn, consisted of 107 subjects, with a similar age and sex distribution to the other groups.

ANOVA could not, for any of the questions, prove any significant difference between the outcome in group 4 (the autumn control group) and the outcome of the initial questionnaire in the three test groups, which indicates that the time of year did not affect the responses. Between the initial and follow-up questionnaires in the three test groups, ANOVA showed only sporadic effects of demographic factors: depending on age for questions 10 and 21, on sex for question 13b, on skin type for questions 14 and 21, and on educational level for question 11 (p < 0.05).

Sun habits/sun protection behaviour

Table I shows the change in mean values of the 5-point Likert scale scorings, between the initial and the follow-up questionnaire, for questions on sun habits/sun protection behaviour (questions 9–16). No statistically significant worsening of risk behaviour was seen for any question. Lowered risk behaviour was seen more often in groups 2 and 3.

Table I.  Change in mean values of the five-point Likert scale scorings, between the initial and the follow-up questionnaires in each group, for the questions concerning sun habits/sun protection behaviour (questions 9–16 in the questionnaire).

		Change of mean value
Question no.	Questions on sun habits/sun protection behaviour	Group 1 (n=116)	Group 2 (n=97)	Group 3 (n=95)
9	How often do you sunbathe with the intention to tan during the summer in Sweden?	0.12	−0.46***	−0.28 *
10	How often do you usually go on "sun vacation" abroad?	−0.04	−0.01	−0.04
11	Do you usually use a sunscreen when sunbathing?	0.14	−0.26	−0.04
12	If you use sunscreens, which sun protection factor do you choose?	−0.27	−0.24	−0.45*
13	When in the sun, without intention to tan, how often do you use any of the
	following ways to protect from the sun: a) sun screens	0.37	−0.41*	−0.21
	b) shirt or sweater with short sleeves	0.05	−0.04	−0.12
	c) shirt or sweater with long sleeves	0.24	−0.06	−0.13
	d) sun hat or cap	−0.12	−0.25	0.01
	e) trousers with long legs	−0.18	−0.34*	−0.05
	f) staying in the shade	−0.22	−0.12	−0.19
14	How many times have you been sunburnt during the past year?	−0.09	−0.19*	−0.21*
15	How often do you use a sun-bed?	−0.07	−0.12	−0.18*
16	How long do you usually stay in the sun between 11 am and 3 pm (June–Aug)?	−0.25*	−0.34**	−0.13

*p < 0.05, **p<0.005, ***p<0.0005. A negative value indicates change towards a lowered risk behaviour. Group 1: letter group; group 2: doctor's consultation group; group 3: phototest group.

Stages of change

The changes in mean value for the five scored stages of change (question 17a–d), between the initial and the final questionnaire, are presented in Table II. Significantly lowered risk behaviour was only seen in groups 2 and 3.

Table II.  Change in mean values, between the initial and the follow-up questionnaires in each group, for the five stages of change scored as 1–5 (from maintenance to pre-contemplation stage) and for each of the four behavioural items.

		Change of mean value
Question no.	Stages of change Behavioural item	Group 1 (n=116)	Group 2 (n=97)	Group 3 (n=95)
17 a)	Giving up sunbathing	−0.19	−0.57**	−0.42*
b)	Using clothes for sun protection	−0.12	0.08	−0.27
c)	Using sunscreens	−0.19	−0.04	−0.09
d)	Staying in the shade	−0.25	−0.72***	−0.31

*p<0.05, **p<0.005, ***p<0.0005. A negative value indicates change towards a loweredrisk behaviour. Group 1: letter group; group 2: doctor's consultation group; group 3: phototest group.

Attitudes towards sunbathing

Table III displays the results of the questions on attitudes towards sunbathing (questions 18–22), following the same principle as for sun habits in Table I. The greatest effect on attitudes was seen in group 2.

Table III.  Change in mean values of the five-point Likert scale scorings, between the initial and the follow-up questionnaires in each group, for the questions on attitudes towards sunbathing (questions 18–22 in the questionnaire).

		Change of mean value
Question no.	Questions on attitudes towards sunbathing	Group. 1 (n = 116)	Group. 2 (n = 97)	Group. 3 (n = 95)
18	How do you like sun bathing?	−0.14	−0.21 *	−0.06
19	Do you think that the advantages of sun bathing outweighs the disadvantages?	−0.20	−0.29	0.08
20	How extensive do you consider the health risks with sun bathing?	−0.15	−0.38 **	−0.04
21	How extensive do you consider the risk for you to develop skin cancer?	−0.20 *	−0.24 *	−0.09
22	How important is it for you to get tanned during the summer?	−0.09	−0.12	−0.09

*p<0.05, **p<0.005, ***p<0.0005. A negative value indicates change towards a lowered risk behaviour. Group 1: letter group; group 2: doctor's consultation group; group 3: phototest group.

Role of phototesting

Group 3 (the phototest group) was at analysis divided into two subgroups according to the number of erythemal reactions reported from the phototest. This gave a group with low UV sensitivity (0–2 reactions) and a group with high UV sensitivity (3–5 reactions). No subject reported six erythematous reactions. In Table IV, only questions for which a significant change in scoring mean value (p < 0.05) occurred are presented, following the same principle as in Tables I and III. Statistically significant changes in sun habits/sun protection behaviour and attitudes were seen in both subgroups. Lowered risk behaviour, as determined by the Transtheoretical Model of Behaviour Change, was only seen in the high UV sensitivity group.

Table IV.  Change of mean values of the five-point Likert scale scorings (upper part) and for the five stages of change scored as 1–5, from maintenance to pre-contemplation stage (lower part), between the initial and the follow-up questionnaire in group 3 (phototest group), subdivided to show UV sensitivity according to the number of reported reactions by the phototest.

		Phototest result
Question no.	Sun habits/sun protection behaviour and attitudes towards sunbathing	Low UV sensitivity (0–2 reactions) n=41	High UV sensitivity (3–5 reactions) n=54
9	How often do you sunbathe with the intention to tan during the summer in Sweden?		−0.32*
11	Do you usually use a sunscreen when sunbathing?	−0.30**
12	If you use sunscreens, which sun protection factor do you choose?		−0.60***
13b	When in the sun, without int. to tan, how often do you use shirt with short sleeves?		−0.30*
14	How many times have you been sunburnt during the past year?	−0.37*
15	How often do you use a sun-bed?	−0.13*
21	How extensive do you consider the risk for you to develop skin cancer?	0.23**
Stage of change for
17a	Giving up sunbathing		−0.46*
17b	Using clothes for sun protection		−0.55*
17d	Staying in the shade		−0.38*

A negative value indicates change towards a lowered risk behaviour. Only statistically significant changes are shown (*p<0.05, **p<0.005, ***p<0.0005).

Discussion

The basic prerequisite for all kinds of preventive endeavours in healthcare is to identify the most appropriate and effective way of presenting the message. This paper clearly demonstrates a higher impact for personally presented information/prevention connected to a doctor's consultation than by solely written feedback. This is of course both encouraging and motivating for all healthcare personnel, and strongly indicates that written information alone has difficulty in reaching its aims in competition with other information and messages aimed at the individual often in similar ways for a variety of purposes. A patient's individual consultation with a doctor or with other healthcare professionals constitutes a unique occasion not only for delivering a preventive message, but also for adjusting the message according to the personality, individual risk factors, and receptive communication abilities of the patient.

An interesting, and to some extent surprising, observation is that the addition of a phototest did not contribute to increased readiness to change behaviour in the phototest group as a whole. On the contrary, this actually appears to be a little lower compared to the doctor's consultation group. This may, of course, be due to several reasons. The most conceivable explanation is that the “subgroup” of individuals reacting with the lowest UV sensitivity according to the phototest were less likely to change their behaviour, finding in the phototest no reinforcement for increased sun protection. This is partly confirmed by the finding that individuals with higher UV sensitivity had a markedly higher tendency to change their behaviour, even concerning questions for which the doctor's consultation group showed no change. Another possible confirmation of this explanation is provided by the outcome of the question on considered risk for getting skin cancer (question 21), which subjects with low UV sensitivity actually appeared to be significantly less concerned about at follow-up. This might indicate that the phototest outcome resulted in a false sense of security in this group. Interestingly, at the same time, they reported an increased use of sunscreens (question 11) and also a reduced use of sun-beds, a phenomenon not present for the high UV sensitivity subgroup. A possible explanation for this might be that individuals with higher UV sensitivity already had more extensive use of sunscreens, and a lower frequency of sun-bed use, as a result of a tendency to become sunburnt, which they would have noted previously.

Is there, then, a place for, or benefit in, performing a phototest in association with the questionnaire intervention in the manner here described? Indeed, the answer is not fully clear. On the one hand, since the demonstrated subgroup of individuals with higher UV sensitivity, according to the phototest, also showed higher propensity to improve sun protection behaviour, the phototest might be viewed as a tool for more efficient targeting. In this subgroup, with a constitutionally higher risk for getting skin cancer, improvement of sun protection behaviour would be of high preventive value. On the other hand, the question remains as to whether or not it is realistic to believe in a more extensive use of phototesting in primary care. In using the patient self-report approach here described, neither phototest administration nor interpretation of test results is difficult in the primary care setting. The financial cost for the phototest equipment is low. The question is rather whether the prevention area itself is of sufficient priority and whether the phototest actually can contribute to better outcome. Further study to address the issues here raised is warranted.

It is important to comment on the possibility of selection bias. Although we targeted an unselected material of adults visiting a primary healthcare centre, it is likely that certain personalities have a higher tendency to participate in such studies, to complete health questionnaires, and to engage in health-promoting activities, and that these might also have a higher tendency to change behaviour than those who desist. In the present study, the larger number of female subjects may be an expression of this phenomenon.

Another important question is whether self-reported behaviour correlates to actual behaviour. A discrepancy between these has been reported [28]. It cannot be excluded that the change of a reported behaviour in reality reflects the will to change behaviour rather than an actual change. This change would, however, not be likely to be so clearly group dependent, and should perhaps also have been more marked in the letter group. Bränström et al. [21] investigated the stability of questionnaire answers concerning sun protection behaviour, and found that, for an individual subject, this may differ between two answering occasions (3 weeks apart), indicating that the individual does not remember what he/she answered on the first occasion. In previous studies on sun habits/sun protection behaviour, various follow-up intervals have been utilized, from 2 months up to 1 or 2 years. Some of these have been able to demonstrate a lasting behavioural change over time [5], [14]. In the present study, we chose a half-year assessment in order to be able to study behaviour change over one summer season.

Prevention of sun-related skin disease only represents a small part of the primary healthcare prevention workload, which is increasingly dominated by lifestyle-related measures in the cardiovascular field, with diet, exercise, tobacco smoking, and stress constituting important behaviour parameters for the development of ischemic heart disease, hyperlipidaemia, and diabetes. Although the present study focuses on skin cancer, the outcome is likely also to be relevant to a broad range of conditions, such as those mentioned above. The study emphasizes the importance of the personal doctor's consultation, and the fact that what is communicated from doctor to patient in terms of prevention does have a concrete consequence for the patient. Information from large media campaigns and from focused preventive efforts probably play important roles in the prevention panorama, but are naturally restricted to isolated points in time or campaigns. Prevention in primary healthcare, on the other hand, has a more marked requirement for continuity and the ability to proceed appropriately at each relevant occasion of patient care. Studies such as the present study may help to identify new modalities and methods for bringing home the preventive message, and to find new ways to target specific risk groups. A phototest with patient self-reading could, for example, be the equivalent of a risk profile questionnaire, a laboratory test, or any other kind of measure-of-risk indicator. In an increasingly challenged health economic climate, it is undoubtedly of major importance that preventive measures are being used effectively and with forethought.

Conclusion

We have demonstrated that, after a questionnaire focusing on issues of relevance for skin cancer prevention, a personally communicated prevention message delivered during a doctor's consultation had a clearly better impact on the recipient than a corresponding prevention message presented solely in letter form. The addition of a phototest to the consultation improved behavioural outcome in individuals with high UV sensitivity, but not for the phototest group as a whole. Since high UV sensitivity individuals constitute a specific risk group for skin cancer, enhancement of behavioural change in this group would be of high preventive value, and future studies on the role of phototesting in skin cancer prevention should therefore focus more specifically on this group of susceptible, at-risk individuals.

Acknowledgements

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.