Skin cancer prevention practices among malignant melanoma survivors: a systematic review

Abstract

Purpose

This systematic review was conducted to evaluate and summarize the existing literature on prevalence of ultraviolet radiation (UVR) exposure, sun protection, and screening behaviors among individuals diagnosed with malignant melanoma (MM).

Methods

The search was performed in PubMed, CINAHL, PsycINFO, ScienceDirect, EMBASE, and ERIC from inception of each database through July 2014. Studies were included if (1) individuals diagnosed with MM were the primary sample, (2) measured UVR exposure, primary and secondary preventive behaviors, (3) original research communication that constitutes an entire set of empirical data, (4) observational design, and (5) English peer-reviewed. Studies were excluded if (1) all of the inclusion criteria were not met and (2) duplicates, conference abstracts, editorials, news, letters to the editor, comments, reviews, feature articles, white papers, and guidelines.

Results

The search resulted in 255 articles that were screened for relevance; however, only 15 articles met all of the inclusion criteria. Most of the studies were cross-sectional (n = 10), used self-administered surveys (n = 8), and were conducted in North America (n = 10). The sample sizes ranged considerably, but were mostly Caucasian (n = 6) and included a higher proportion of women (n = 8). Evidence demonstrated that individuals with MM still engaged in sunbathing, indoor tanning, and reported sunburns. Moreover, survivors reported inadequate levels of both sun protection and skin self-examinations.

Conclusions

The findings highlight the need for intensifying intervention strategies to reduce the risk of new primary MMs in this group. Future research should increase in rigor and include more diverse populations and regions.

Introduction

Malignant melanoma (MM) accounts for approximately 75 % of all deaths from skin cancer, constituting an important and growing public health problem (American Cancer Society 2014a, b). Nearly 86 % of MM cases are linked with exposure to ultraviolet radiation (UVR) emitted by the sun (Parkin et al. 2011). Adoption of effective sun protection and regular skin surveillance behaviors is of paramount importance, particularly among individuals previously diagnosed with MM, as they are at an increased risk of developing an additional MM in the future. Epidemiologic evidence has demonstrated that for MM survivors, the risk of development of a subsequent MM is nearly nine times greater when compared with the risk of developing first primary MM in the general population (Bradford et al. 2010). Furthermore, these survivors remain at increased risk of developing another MM for over 20 years (Bradford et al. 2010).

Considering the life-threatening nature of MM, and the elevated risk of new primary lesions in these patients, it is imperative to understand the degree to which patients engage in recommended skin cancer risk-reduction behaviors. Moreover, exploration of skin cancer-related knowledge and attitudes would augment the potential for health professionals to design targeted interventions in order to promote methods of skin cancer prevention in this population group. Accordingly, the goal of the present systematic review was to assess relevant existing research studies to address the following specific questions: “What is the prevalence of UVR exposure, sun protection, and screening behaviors among individuals diagnosed with MM?” and “What are knowledge levels and attitudes concerning skin cancer among this potentially vulnerable group?” Additionally, this review will identify gaps in the currently available literature and propose recommendations for future research in this topic area. To the best of our knowledge, no such systematic review has yet been published.

Methods

Eligibility criteria

A priori inclusion and exclusion criteria were specified to select studies for this systematic literature review.

The following inclusion criteria were used for retrieving as well as reviewing the studies:

1. Population: post-MM diagnosis.

2. Measures (at least one of the following): UVR exposure, primary and secondary preventive behaviors.

3. Article type: original research communication that constitutes entire set of empirical data.

4. Study design: observational.

5. Language of publications: English.

6. Journal type: peer-reviewed.

Studies were excluded if

1. All of the inclusion criteria were not met.

2. Did not feature “individuals diagnosed with MM” as the primary sample of the study.

3. Duplicates, conference abstracts, editorials, news, letters to the editor, comments, reviews, feature articles, white papers, and guidelines.

Literature sources and search strategy

The steps outlined by internationally established guidelines were followed to direct the search strategy for this systematic review (Moher et al. 2009). In order to identify potentially relevant articles, an exhaustive search was conducted in July 2014 on six bibliographic electronic databases (i.e., PubMed, CINAHL, PsycINFO, ScienceDirect, EMBASE, and ERIC). To ensure that no studies were missed, additional searches were conducted in the University of Mississippi library and Google Scholar. The searches were not restricted by the date of publication. Furthermore, key dermatology journals were hand searched to supplement database searching. The search process did not include any strategy to access gray literature including unpublished or other difficult-to-access works. All search strategies were executed by two independent reviewers of this research.

To capture every possible study, keywords were derived through scanning previous literature related to skin cancer. A list of synonyms of the identified keywords was created for other search terms. Boolean operators (“AND” and “OR”) were used to construct the search strings, which were pilot tested and further modified to assure that they locate available significant literature to address the review objective. The final search strings entered were as follows:(Skin Cancer OR Melanoma) AND (Survivors OR Diagnosed OR History) AND (Primary OR Secondary) AND (Protect* OR Prevent*) AND (Knowledge OR Attitudes) AND (Sun* OR UV*) AND (Tan* OR Expos* OR Risk) AND (Behavior OR Habits). All retrieved references were then manually examined, and duplicates were removed. Titles, abstracts, and full texts were reviewed for inclusion or exclusion based on the aforementioned criteria. The reference lists of primary articles were checked to obtain additional pertinent studies. Finally, researchers screened all the included studies to ascertain whether the studies met the predefined eligibility criteria. There were no disagreements between reviewers over the eligibility assessment.

Quality appraisal of studies

To evaluate methodological strength of the included studies, questions were specifically adapted from instruments previously used in systematic reviews (Estabrooks et al. 2009; Kajermo et al. 2010). The tool consisted of six items covering aspects of sampling, measurement, and statistical analysis. A full list of assessment questions is presented in Table 1. All items were scored with yes (1) or no (0). The overall quality assessment score of a study was calculated by summing the scores of each item and then dividing the total points scored by the total possible points (6). The final maximum score for each study that could be obtained was 1. Studies were then rated as weak quality (<0.50), moderate quality (0.50–0.74), and strong quality (0.75–1). The rating results were not used to determine eligibility for inclusion, but to provide information about the quality of the selected studies, and to aid in identifying factors that might affect the findings of this research. The first reviewer and second reviewer independently assessed the methodological quality of all the included studies. Using SPSS (Statistical Package for Social Sciences) version 22, Cohen’s kappa coefficient (K) was calculated to establish inter-rater reliability. Any conflicts that emerged were resolved by consensus discussion between the reviewers.

Table 1.

Quality assessment of included studies

	Yes	No
Sampling:
Was probability sampling used?	2	13
Was sample drawn from more than one site?	7	8
Was the response rate more than 60 %?	4	11
Are the participants likely to be representative of the target population? (very likely = 2; somewhat likely = 3)	5	10
Measurement:
Was validity or reliability obtained?	6	9
Statistical analysis:
Were the results reported based on inferential statistical analyses?	12	3
Study quality rating:
Weak quality (<0.50): 9 studies
Moderate quality (0.50–0.74): 4 studies
Strong quality (0.75–1): 2 studies

Data extraction

A data extraction table was predetermined by the reviewers. After the first reviewer extracted the data independently, the second checked for accuracy. Again, discrepancies between both the reviewers were discussed and reconciled by consensus.

Data synthesis

Data were presented descriptively. A formal meta-analysis was prohibited by methodological shortcomings of the studies, insufficient data for statistical pooling, and wide variability in operationalization of outcome measures. As such, a narrative synthesis was performed on results extracted from the set of studies included in this review.

Results

Search outcome

The electronic and manual searches identified a total of 410 references. After removing duplicates, titles and abstracts of 255 articles were screened for relevance, after which 53 articles remained for full-text reading. Of the latter, 40 articles were eliminated based on preceding eligibility criteria. The reference list scanning uncovered two other potential articles. This searching process resulted in 15 articles meeting all inclusion criteria. Figure 1 depicts the flow diagram of the literature search procedure. The outcome of data extraction from each included study is found in Table 2, chronologically arranged.

Fig. 1.

Literature search procedure

Table 2.

Summary of included studies

First author, year, location, design, and quality	Data collection method and sample size	Gender, ethnicity, and age	UVR exposure	Primary preventive behaviors	Secondary preventive behaviors
Freiman A, 2004, Canada, cross-sectional, and weak	Self-administered survey and n = 217	52 % male and $\overline{x}=56\text{years}$	Sunbathing: 21 % (at least sometimes) Indoor tanning: 2 %	Sunscreen: 72 % (often or always), 18 % (sometimes), and 7 % (never) Hat and clothing: 85 % Sun avoidance: 82 %	–
Loescher LJ, 2006, USA, cross-sectional, and weak	Self-administered survey and n = 70	53 % male, 97 % Caucasians, and $\overline{x}=65\text{years}$	–	–	Thorough skin self-examination: 33 %
Manne S, 2006, USA, cross-sectional, and strong	Self-administered mailed survey and n = 229	57.2 % female, 99.1 % Caucasians, and $\overline{x}=53.8\text{years}$	–	Sunscreen: 59.4 % (often or always) Hat: 44.5 % (often or always) Long-sleeved shirt: 44.9 % (often or always) Shade: 53.2 % (often or always) Sunglasses: 70.7 % (often or always)	Thorough skin self-examination: 13.7 %
Lee TK, 2007, Canada, case–control, and moderate	Telephone survey and n = 35/70	51 % female, 100 % Caucasians, and 50+ years = 54 %	Time spent in outdoor recreational activities per month: $\overline{x}=23\text{h}$ Time spent outdoors during the working week: $\overline{x}=17.1\text{h}$ Time spent outdoor during the weekend: $\overline{x}=8.7\text{h}$ Sunburns: $\overline{x}=0.7/\text{person}$	Mean monthly protected episodes: 26.1 Mean monthly unprotected episodes: 32.2	–
Mujumdar, UJ, 2009, USA, cross-sectional, and moderate	Telephone survey and n = 115	55 % female, 99 % Caucasians, and $\overline{x}=60\text{years}$	–	Sunscreen: 57 % (always or nearly always) and 7 % (never) Hat: 32 % (always or nearly always) and 32 % (never) Long-sleeved shirt: 13 % (always or nearly always) and 41 % (never) Shade: 43 % (always or nearly always) and 3 % (never)	Thorough skin self-examination: 17 %
Gomez-Moyano E, 2010, Spain, cross-sectional, and weak	Self-administered survey and n = 195	63 % female	Sunbathing: 66.1 % (at least sometimes)	Sunscreen use: 49.2 % (often or always) and 38.5 % (never) Hat and clothing: 90.8 % Sun avoidance: 75.4 % (often or always) and 20 % (never)	–
McMeniman E, 2010, Australia, cross-sectional, and weak	Telephone survey and n = 52	50 % female and $\overline{x}=59\text{years}$	–	Sunscreen: 84.6 % Hat and clothing: 96.2 %	Skin self-examination: 86.6 %
Soto E, 2010, USA, cross-sectional, and weak	Telephone survey and n = 68	95 % ≥ 40 years	Sunbathing: ≈1 % (daily), ≈8 % (weekly), ≈5 % (monthly), ≈1 % (yearly)	Sunscreen: 69 % (most of time or always) Hat: 66 % Clothing: 67 % Sun avoidance: 74 % (most of time or always)	–
Bowen D, 2012, USA, cross-sectional, and moderate	Telephone survey and n = 313	55.9 % female, 98.7 % Caucasians, and $\overline{x}=56\text{years}$	–	Sunscreen: 45 % (often or always) and 35.1 % (never) Wide brimmed hat: 15.6 % (often or always) and 67.1 % (never) Something on head: 34.8 % (often or always) and 34.8 % (never) Long-sleeved shirt or blouse: 59.1 % (often or always) and 14.7 % (never) Long pants or long skirts: 79.9 % (often or always) and 5.4 % (never) Shade: 35.5 % (often or always) and 27.5 % (never) Sunglasses: 46 % (often or always) and 26.8 % (never) Sun avoidance: 19.5 % (often or always) and 54.6 % (never)	Thorough skin self-examination: 22 % (self-perception) and 37.7–72.5 % (body parts) Clinical skin screening: 88 %
Mayer D, 2012, USA, case–control, and strong	Self-administered mailed survey/telephone survey and n = 156/11,564	52.5 % female, 91 % Caucasians, and 50+ years = 86.5 %	Tanning bed: 6.4 %	Sunscreen: 50.9 % (often or always) and 29.6 % (rarely or never) Clothing: 74.3 % % (often or always) and 8.1 % (rarely or never) Shade: 42.9 % (often or always) and 26.1 % (rarely or never)	–
Zivkovic MV, 2012, Croatia, case–control, and moderate	Self-administered survey and n = 120/240	58.3 % male and $\overline{x}=51.11\text{years}$	Natural sunlight sunbathing: 22.5 % (during the whole year) and 13.3 % (during the whole day) Artificial sunbathing: 1 % (1–2 times a month)	Sunscreen: 41.6 % (summer holidays), 28.3 % (spring to autumn), 10 % (whole year), and 16.6 % (never)	–
Idorn LW, 2013, Denmark, prospective case–control, weak	Personal electronic dosimeter, sun exposure diary, and n = 53/104	64 % female and median = 37 years (28–70) (recently diagnosed patients); 45 years (26–66) (patients diagnosed in the past)	Patients diagnosed in the past had higher UVR dose than recently diagnosed patients Sunburns: $\overline{x}=2\pm 2$ (recently diagnosed patients) and 1 ± 1 (patients diagnosed in the past)	Patients diagnosed in the past had significantly lower number of days wearing sunscreen compared to newly diagnosed patients	–
Korner A, 2013, Canada, cross-sectional, and weak	Self-administered survey and n = 47	51.1 % male and $\overline{x}=55.39\text{years}$	–	–	Skin self-examination: 87.8 %
Palesh O, 2014, USA, cross-sectional, and weak	Self-administered electronic survey and n = 160	51 % male, 94 % Caucasians, and $\overline{x}=61.9\text{years}$	After diagnosis decreased use of tanning beds (27.5 %), time spent in the sun (73 %) and seeking a tan (63 %)	After diagnosis increased use of sunscreen (71 %) and protective clothing (73.8 %)	Clinical skin screening: 94 %
Idorn LW, 2014, Denmark, prospective case–control, weak	Personal electronic dosimeter, sun exposure diary, and n = 20/40	65 % female and $\overline{x}=43\text{years}$	Increase in daily UVR dose across years Sunburns: 60 % at least once during the 3 years	Patients had more number of days using sunscreen on the upper extremities compared with controls only in the first year of follow-up	–

Appraisal of the quality of studies

Of the final 15 included studies, only two studies indicated strong methodological quality, four studies indicated moderate quality, and the remaining nine studies indicated weak quality. Inter-rater reliability between assessors on the quality measurements of the studies was very good (K = 0.81).

Study characteristics

All studies included were performed over the past decade. Studies were conducted in the USA (n = 7), Canada (n = 3), Denmark (n = 2), Australia (n = 1), Croatia (n = 1), and Spain (n = 1). Over the years, studies in this area have been predominantly cross-sectional research designs (n = 10). Five studies found were case–control; of these, the methodology of three studies was in essence cross-sectional (Lee et al. 2007; Mayer et al. 2012; Živkovic et al. 2012), and other two were performed prospectively (Idorn et al. 2013, 2014). Self-administered survey (n = 8) was the most frequently used method of data collection.

Participant characteristics

The number of MM patients in the studies ranged from 20 to 313. The majority of the studies (n = 8) noted a higher proportion of females (range 51–65 %). Seven studies provided data on ethnicity, of which six reported that >90 % of the respondents were Caucasians, and one study’s sample was composed of all Caucasians. The mean age of the participants varied from 43 to 65 years across the 10 studies that presented this information.

UVR exposure

Studies indicated that MM patients are still involved in summer outdoor activities (Lee et al. 2007), sunbathing (Freiman et al. 2004; Gómez-Moyano et al. 2010; Soto et al. 2010; Živkovic et al. 2012), and indoor tanning (Freiman et al. 2004; Mayer et al. 2012; Živkovic et al. 2012). Moreover, findings of the studies based on personal UVR dosimeters revealed that patients increased their amount of time spent under the sun following diagnosis (Idorn et al. 2013, 2014). Even more surprisingly, survivors reported experiencing sunburns after their diagnosis (Lee et al. 2007; Idorn et al. 2013, 2014).

Primary preventive behaviors

Between 7 and 38 % of the MM patients reported that they “never” apply sunscreen when outside on a sunny days (Freiman et al. 2004; Mujumdar et al. 2009; Gómez-Moyano et al. 2010; Živkovic et al. 2012; Mayer et al. 2012). Some research groups elicited hat and clothing use in combination, and their findings regarding engaging in this practice were strikingly high (85–96 %) (Freiman et al. 2004; McMeniman et al. 2010). Contrarily, Mujumdar et al. (2009) reported that 41 % “never” wore a long-sleeved shirt when out in the sun. Furthermore, in another study, a 7-day recall showed that 67 % of the respondents “never” wore a wide brimmed hat and 35 % “never” use something to cover their head (Bowen et al. 2012). In addition, 55 % of the participants in the latter study indicated that they “never” avoid outdoors during the hottest hours of the day (Bowen et al. 2012). A population-based study noted that one-fourth of the MM survivors “rarely or never” stay in the available shade when out in the sun (Mayer et al. 2012).

Secondary preventive behaviors

With regard to skin screening, relatively fewer studies were identified. Overall, 14–33 % of MM patients acknowledged engaging in thorough skin self-examinations (SSE) (Loescher et al. 2006; Manne and Lessin 2006; Mujumdar et al. 2009). Recent surveys indicated that a high majority (88–94 %) of MM patients had received clinical skin examination (Bowen et al. 2012; Palesh et al. 2014).

Skin cancer-related knowledge and attitudes

Moderate to high levels of knowledge about skin cancer and risk factors were documented in the reviewed studies (Manne and Lessin 2006; Lee et al. 2007), although these data are noticeably sparse.

Furthermore, studies showed that MM survivors still possess positive attitudes toward tanning (e.g., tan looks healthier) (Freiman et al. 2004; Lee et al. 2007; Gómez-Moyano et al. 2010).

Discussion

Findings in the literature clearly illustrated that certain segments of MM survivors do not limit their exposure to UVR, which is regarded as a primary risk factor for skin cancers (Parkin et al. 2011). Episodes of sunburn were reported by patients after MM diagnosis (Lee et al. 2007; Idorn et al. 2013, 2014), indicating that they intermittently expose themselves to erythema-inducing levels of UVR. This is of pressing concern because frequency of sunburn, even in adulthood, elevates individuals’ chances of developing MM (Pfahlberg et al. 2001). Moreover, some evidence was found that over a period of time (Idorn et al. 2013, 2014), MM patients increased their UVR exposure after diagnosis. These results suggest that periodic reinforcement is requisite to reduce the possibility of developing subsequent skin cancer later in life. Qualified professionals, such as physicians and nurses, have the potential to be a powerful channel to routinely discuss risks associated with UVR exposure and sunburn. During follow-up visits, health care providers should emphasize and discuss the need for continually practicing sun protective behaviors and the importance of adhering to these behaviors in an effort to prevent MM recurrence. Medical care providers could also play a vital role in periodically measuring sunburn prevalence among survivors to monitor adherence to sun safety recommendations (Saraiya et al. 2002). Moreover, efforts should focus on collecting rigorous longitudinal data after diagnosis to shed light on changes over time in psychosocial determinants of UVR exposure behaviors. Such studies may aid health care professionals to design and implement intervention strategies that have long-term behavior change potential.

Furthermore, engagement in intentional tanning behaviors among MM survivors was demonstrated in the studies (Freiman et al. 2004; Gómez-Moyano et al. 2010; Soto et al. 2010; Mayer et al. 2012; Živkovic et al. 2012). This might be attributed to patients’ desire to have a healthy and attractive appearance; findings from previous research on tanning suggest that the persistent self-perceived belief that tan skin enhances body appearance is a driving psychological force of high-risk tanning habits (Robinson et al. 1997; Rhainds et al. 1999; Pagoto et al. 2009). Individuals who highly value the benefits of tanning are, for the most part, resistant to skin cancer prevention educational messages (Borland et al. 1990, 1991; Detweiler et al. 1999; Pagoto et al. 2009). Consequently, special preventive programs in the last decade were developed that focused on damaging effects of UVR exposure on appearance (e.g., sagging, wrinkles, and brown spots) to counteract positive tanning attitudinal beliefs in order to increase sun protection practices (Williams et al. 2013). A recent systematic review and meta-analysis showed that this intervention technique was consistently effective in a number of studies (Williams et al. 2013). Future research should evaluate the utility of an appearance-based educational intervention to minimize survivors’ artificial and solar UVR exposure. Along with this, extensive work is greatly needed to examine the underlying psychological mechanism of the decision-making process regarding the adoption of UVR-seeking behaviors among this specific population group.

Virtually, all studies herein indicated that some MM patients fail to practice one or more sun safety methods. Sunscreen use was the most commonly measured sun protection strategy in the literature. However, only few studies have attempted to measure the appropriateness of sunscreen application. In these studies, the results revealed that a significant proportion of patients did not reapply sunscreen, failed to apply it to all exposed areas, and applied it for very short period of time during the year (Gómez-Moyano et al. 2010; Soto et al. 2010; Živkovic et al. 2012). It was also found that a substantial portion of MM survivors used sunscreen with low SPF levels (Mujumdar et al. 2009; Živkovic et al. 2012). Interestingly, in one of the studies, over 90 % of the patients were advised by their doctors to utilize sunscreen, but many reported that they did not recall any form of advice on appropriate SPF values (Soto et al. 2010). Collectively, these findings highlight the need to practitioner–patient communication by incorporating specific instructions to guide patients about correct sunscreen use to increase effectiveness of this practice. Moreover, in future research, attention should be given to in-depth assessment of sunscreen-use behavior among MM patients.

Gender disparities in sun protection behaviors were also evident. For example, male MM survivors were less likely to use sunscreen and more likely to use hats than female MM survivors (Manne and Lessin 2006; Bowen et al. 2012). Moreover, studies noted that being female was associated with both regular sun protection and sun protection behavior change after MM diagnosis (Mujumdar et al. 2009; Gómez-Moyano et al. 2010). This could be partially explained by higher levels of skin cancer risk perceptions, greater positive sun protection attitudes, and fewer perceived sun protective barriers in females, compared to their male counterparts (Kasparian et al. 2009). However, higher sunscreen use among female survivors is of particular concern because it is related with a decrease in other sun protection strategies (e.g., protective clothing and sun avoidance) and an increase in exposure to sunlight (Autier et al. 1999; McCarthy et al. 1999; Cho et al. 2010), which is generally referred to as “sunscreen paradox” (Bränström et al. 2010). Health care practitioners need to be cognizant that these differences exist and tailor psychoeducational programs in a manner to efficiently target their patients’ particular sun protection needs.

There are currently far too little studies that have examined SSE in MM diagnosed patients. This is likely due to the fact that no single randomized controlled trial has explicitly evaluated the potential efficacy of screening on reducing MM mortality (Manne and Lessin 2006). Nevertheless, a case–control study indicated that performance of SSE is associated with a 63 % decrease in MM-related mortality (Berwick et al. 1996). Previous studies showed that 40–55 % of diagnosed MMs were initially detected by the patients incidentally or while conducting a deliberate SSE (Hamidi et al. 2008). In addition, various professional organizations suggest routine skin screening among high-risk groups, either by their health care provider or SSE (American Cancer Society 2014a, b; American Academy of Dermatology 2014; Skin Cancer Foundation 2014). This approach is efficacious in earlier detection of occurrence as well as reoccurrence of skin cancer, and hence improving the chances of survival (Berwick et al. 1996). With respect to SSE, the pattern of findings seen in this review accords with previous promising studies by Weinstock and colleagues which showed that prevalence estimates of SSE are subject to vary broadly due to differences between measurements of SSE performance (Hamidi et al. 2008). For example, when studies queried about SSE practice without specific definition, high rates (59–88 %) were reported among MM patients (Loescher et al. 2006; McMeniman et al. 2010; Körner et al. 2013). On the other hand, when MM patients were questioned with specific definition whether they practiced thorough SSE (Berwick et al. 1996; Weinstock et al. 1999, 2004), the rates (14–33 %) were disappointingly much lower (Loescher et al. 2006; Manne and Lessin 2006; Mujumdar et al. 2009), and these results were fairly consistent with thorough skin examination performance (9–18 %) among the general population (Weinstock et al. 1999, 2004). According to Weinstock et al. (2004), the most accurate measures are possibly those that inquire the most explicit information about examination of each of the specific areas of skin surface. To what extent individuals examine their body surface comprehensively is critical, as substantial numbers of MM occur on body sites that can be harder for patients to visualize (e.g., scalp, ears, neck, back, and back of legs) (Aitken et al. 2004). This review not only highlights the significance of performing a thorough SEE, but also encourages investigators to take thorough SSE into consideration to report more accurate estimation of the overall SSE performance.

Turning now to rates of clinical skin examination, studies that assessed this information showed that an overwhelming percentage of MM patients (88–94 %) were being screened during their routine clinical care (Bowen et al. 2012; Palesh et al. 2014). This may imply that a large number of patients frequently interface with health care providers; hence, these encouraging findings showcase a viable opportunity to promote skin cancer prevention. It has been found that performance of SSE and sun protection behaviors in MM survivors were associated with physicians recommending these practices (Manne and Lessin 2006). During a visit, medical practitioners should provide instructions to survivors on how to efficaciously conduct SSE in order to improve proficiency and possibly increase self-efficacy of performing SSE. They should also convey information about the frequency of SSE and specific tools to be used for this procedure, for example wall mirror, hand mirror, and MM illustrations. In addition, counseling about sun protection should be incorporated into all routine visits of MM survivors. Robinson et al. (2007) showed that SSE skills training involving partners could also be an optimal strategy for clinicians to enhance SSE self-efficacy and SSE performance among MM patients. Moreover, a body of research in the area of other chronic conditions documented that interventions targeting patients and partners are promising in terms of managing illness and improving treatment adherence (Martire et al. 2010). The couple-oriented interventions seem therefore valuable for increasing compliance with skin cancer prevention practices. Health professionals should explore how to best practically integrate this approach into survivors’ care plan.

The potential limitations should be considered while interpreting the results of the present research. Although literature searches were performed in a wide range of electronic databases, some journals of interest were possibly missed because they may not have been indexed in these databases. Thus, it is plausible that some of the studies were not identified in the retrieval process that would otherwise have been qualified for inclusion. The search strategy included only English-language scientific studies, which may introduce a degree of bias; however, the influence of language bias on conclusions drawn from systematic reviews is minimal (Wright et al. 2007).

Research studies covered in this review were based in western countries with roughly half being undertaken in the USA; therefore, data are not representative of MM patient populations worldwide. It is worth noting that data presented in this review came from affluent countries. These countries rank high on the human development index; yet, MM patients in these countries continue to put themselves at risk by engaging in unsafe sun practices. In order to build a comprehensive body of knowledge on this subject, it will be necessary to conduct studies in diverse geographical locations. It is possible that sun protection behaviors may vary in relation to economic conditions across nations (Stanton et al. 2004). A potential avenue for future studies is examining preventative behaviors of patients in non-affluent countries with low levels of human development. Such studies will shed light on whether engaging in sun protective practice among MM patients goes beyond socioeconomic conditions. The current research suggests that despite the accumulated wealth and human development opportunities in geographic locations where much of the research has taken place, MM patient-tailored interventions targeting multiple factors are still necessary to promote and encourage the adoption of both primary and secondary skin cancer prevention behaviors.

An additional important limitation encountered was that most of the reviewed studies were of weak quality with methodological shortcomings such as non-randomization, single-site data collection, and low response rate. As a consequence, results of this systematic review are subject to risk of selection bias. More attention should be paid to enhancing the research design, by conducting prospective, longitudinal studies that draw on diverse samples.

Furthermore, all studies that reported ethnicity breakdown had a large proportion of Caucasian participants, thereby limiting the generalization of the findings to other racial and ethnic groups. This suggests that in future research, efforts should be made to target a more varied population of survivors to determine racial and ethnic differences in sun protection and skin cancer screening behaviors.

Another possible limitation lies in the fact that sun protection behavioral data in the studies exclusively rely on participants’ self-reports, which may be influenced by recall and social desirability biases. However, previous studies have validated self-reports of sun protection practices against direct observation measures (Oh et al. 2004; Glanz et al. 2009).

Finally, the majority of the researchers have failed to provide psychometric properties of the tools used in their studies, which may have influenced the validity and reliability of the results reported. The current review is in accordance with Kasparian et al. (2009) that a strong need exists for valid, reliable, and standardized instruments that can be used in a global context. This possibly will expand the potential for comparable assessment of skin cancer preventive parameters between MM patients’ populations at the national and international levels.

Conclusions

Despite these limitations, this systematic review constitutes an important contribution toward understanding the surprisingly limited efforts by many patients with MM to engage in effective behaviors to prevent skin cancer. Taken together, the sun exposure habits and inadequate sun protection measures reported highlight the need for intensified psychoeducational strategies focused on their specific needs. While the majority of a physician’s responsibility and time is with ongoing and immediate medical problems, the amount of time spent on reinforcing preventative sun protection behaviors is crucial to reduce the likelihood of MM occurrences. We believe that clinicians should communicate with MM patients about their increased risk of new primary lesions at every encounter. This should be followed by the provision of verbal and written communication regarding regular sunscreen application, wearing hats, covering up, and avoiding the sun when possible. Within clinics and hospitals, we recommend that MM prevention information should be integrated within each patient’s care plan. Additionally, electronic medical record reminders could be utilized to help ensure that patients are scheduled for appropriate follow-up visits.

More research is also necessary to thoroughly understand gaps in survivors’ knowledge and personal attitudes regarding skin cancer and UVR exposure. We must better understand cognitive barriers to sun protection that impact changes in behavior. Ultimately, however, adhering to sun protective measures requires enhanced motivation that is not solely rooted in knowledge about skin cancer. Perhaps an emphasis on the damaging effects of UVR on one’s personal appearance may prove to be useful. Of particular importance, these patients should be taught the serious risks associated with a “healthy tan.” Finally, self-screening programs should encourage patients to routinely and thoroughly monitor their entire cutaneous surface for suspicious skin lesions.

Compliance with ethical standards

Conflict of interest

None of the researchers will be benefitting financially from the study or the publication. Vinayak K. Nahar, M. Allison Ford, Javier F. Boyas, Stephanie K. Jacks, Rizwana Biviji-Sharma, Mary A. Haskins, and Martha A. Bass have no conflict of interests to report. Robert T. Brodell discloses the following potential conflicts of interest: Honoraria have been received from presentations for Allergan, Galderma, and PharmaDerm, a division of Nycomed US Inc. Consultant fees have been received from Galderma Laboratories, L.P. Clinical trials have been performed for Genentech and Janssen Biotech Inc. The material in this article is not believed to be relevant to any of these reported conflicts.

Informed consent

Since this was a systematic review, informed consent was not required.

Ethical approval

We have observed appropriate ethical guidelines and legislation in conducting the study. Since this was a systematic review, approval was not required by the IRB at the University of Mississippi.