Importance
Melanoma has a high survival rate if it is detected early. Training patients with early-stage melanoma who are at risk of developing new melanomas to perform skin self-examination (SSE) may improve survival.
Objective
To examine for whom the intervention works best in a sample composed of dyads of patients with melanoma and skin-check partners who received an SSE intervention vs customary care.
Design, Setting, and Participants
For 494 patients with stage 0 to IIB melanoma (mean age, 55 years; 253 [51.2%] females) and their skin-check partners (mean age, 55 years; 280 [56.7%] females), a randomized clinical trial was conducted in ambulatory care dermatologic offices from June 6, 2011, to April 14, 2014. Follow-up assessments were performed at 12 months. Analysis was performed between March 23 and June 25, 2015.
Methods
Dyads of 494 patients and their partners were randomly assigned to receive the intervention (395 patients) or customary care (control) (99 patients). The main outcome was patient SSE self-efficacy. Partner motivation to assist with SSE and relationship quality (eg, agreeability, activities with partner, and happiness) were assessed for moderation of the influence of the intervention’s effect on SSE self-efficacy.
Results
Relationship quality, defined by activities with the partner (β = –0.892, SE = 0.432, t = –2.066; P = .001) and happiness (β = –4.586, SE = 2.044, t = –2.24; P = .001), significantly moderated the intervention effects on patients’ SSE self-efficacy. In contrast, patient-partner agreeability (β = –0.262, SE = 0.148, t = –1.773; P = .09) and partner motivation (β = –0.328, SE = 1.024, t = –0.320; P = .10) did not significantly moderate the intervention effects on patients’ SSE self-efficacy. Differences between the conditions were highest when activities performed with the partner were below average (mean difference, 6.652; P = .001) and when happiness was below average (mean difference, 7.000; P = .001). Although everyone receiving the intervention experienced some benefit, the findings indicate the greatest increases in self-efficacy were observed for those with below-average activities performed with the partner and happiness.
Conclusions and Relevance
The training of patients with melanoma and their partners in early-detection SSE benefited some more than others. Pairs who have low relationship quality, as determined by activities performed with the partner and happiness, may have received the greatest benefits from the training intervention because they were given an activity to perform together.
Trial Registration
clinicaltrials.gov Identifier: NCT01432860
With an estimated 73 000 new cases of invasive melanoma and more than 9900 deaths expected to occur in 2015, melanoma remains a significant public health concern.1 Melanoma is a treatable cancer and has a high survival rate if it is detected in its early stages. However, increased complications and less optimal prognoses are associated with late-stage melanoma, potentiating the need for early detection.2 Individuals who were previously diagnosed with melanoma are 10 times more likely to develop additional melanomas in their lifetimes, making them an important at-risk population on which to focus early detection.3,4
Research has shown that patients with a history of melanoma can be successfully trained to perform skin self-examination (SSE) to identify potentially malignant skin lesions5,6 and that melanomas detected during SSE are more likely to have favorable outcomes.7 However, there are many areas of the body that are difficult to examine by oneself (ie, the scalp and back), indicating that the aid of a skin-check partner may be most beneficial. Robinson et al8 found that in-person SSE training of patients with partners, relative to patients alone, resulted in significantly more SSEs at 4 months after the intervention.
This study uses a different cohort than in previous work and consists solely of dyads (patient and partner) who received either the training intervention or customary care (control) and were evaluated 4, 8, and 12 months after the intervention.9 This analysis seeks to identify moderators of a partner-assisted SSE training intervention to identify for whom the intervention worked best (or least). When a construct (eg, partner motivation) changes the association between a predictor (eg, intervention) and outcome (self-efficacy or the confidence in the ability to perform SSE), it is considered a moderator of the association.
This study expands on the previous work in 3 ways. First, past studies with a different cohort compared SSE training of patients and partners vs patients alone,8,10,11 whereas this analysis examines a sample composed of patient-partner dyads randomized to receive the training intervention or customary care. Customary care consists of the information (if any) provided by a physician regarding the likelihood of developing another melanoma plus any information about finding a melanoma that is widely available on the Internet or the didactic training intervention (MoleScore). Second, this study examined partner motivation and relationship quality as moderators of the SSE training intervention on SSE self-efficacy. These 2 variables have empirical and theoretical associations with the adoption and maintenance of preventive health behaviors and have been shown to moderate SSE self-efficacy.12-14 Finally, this analysis extends the follow-up assessment to 1 year after the intervention, permitting an assessment of the moderating effects of relationship quality for a longer duration.
Patients were recruited from a large Midwestern city from June 6, 2011, to April 14, 2013. In the first recruitment phase, electronic medical records of patients in the Northwestern Medicine health care system were reviewed, and individuals deemed potentially eligible were sent letters inviting them to participate in the study. The second phase included the placement of advertisements in the health sections of 2 regional newspapers with large circulations (together, >1 200 000 weekly) for 12 weeks. Eligible patients had a previous diagnosis of stage 0 to IIB melanoma and at least 6 weeks having elapsed since surgical treatment, aged 21 through 80 years, and had a relationship for more than 1 year with a person who was willing to participate as a skin-check partner. The skin-check partner was identified as a cohabitating spouse or committed partner or a noncohabitating friend, child, parent, sibling, or other relative. Additional information about inclusion criteria, details about the study design, and the study flowchart are available in previous work.9 The Northwestern University Institutional Review Board approved the study. Written consent was obtained from the participants. Analysis was performed between March 23 and June 25, 2015. The full study protocol synopsis can be found in Supplement 1.
After enrollment, the first 150 participants were randomized to 1 of 3 groups: in-person intervention, intervention with a workbook read in the office and taken home, and controls (who received customary care). During the early phase of enrollment, technological advancements in small tablet personal computers (tablets) made it possible to examine the potential benefit of a tablet version of the intervention. The remaining patients were randomized among 4 groups: controls, intervention with a workbook, in-person intervention, and intervention with the tablet version. At each data collection point, the patient and the partner were separated from each other in different rooms to complete the surveys. Patients and partners in the control and intervention arms completed a preintervention assessment of demographic, relationship, and partner variables and postintervention assessments of SSE frequency and self-efficacy in SSE at 12 months.
The intervention consisted of a didactic skills training session at baseline and clinical reinforcement of the skills at 4-month intervals at clinical visits with the dermatologist. The content of the didactic skills training intervention provided patients and their skin-check partners with information on the ABCDE rule (Assess Border, Color, Diameter, and Evolution of pigmented lesions) to assess moles during SSE. Each aspect of the ABCDE rule was explained with color picture examples.15 Participants in the intervention arms (in-person, workbook, and tablet) received a ruler and a lighted magnifying lens to aid in assessing moles, a laminated card with the ABCDE rules, and a body map that included drawings of the regions of the body. Last, the dyads in the intervention arms received a scorecard on which they were asked to record the scores based on the ABCDE rules. For example, border, color, and diameter each received a score from 1 to 3, with 1 indicating normal; 2, not sure; and 3, abnormal.
The first clinical visit with the dermatologist (J.K.R.) occurred 4 months after the didactic intervention and was a structured visit. The patient was completely disrobed, gowned, and seated on an examining table in preparation for a total body skin check. The partner was seated in the examination room. The dermatologist, who started every examination of pairs in the training intervention group by observing the back, invited the skin-check partner to join her in looking at the patient’s back. The partner was asked to point out any spots he or she wanted to be checked and to explain why he or she was concerned. The dermatologist reassured the partner about the correctness of the decision or suggested minor adjustments and informed the patient and the partner that the partner was doing a “good job.”
Dyads in the control group received no didactic training. Before the clinical care visit with the dermatologist, the medical record was flagged to indicate that the dyad was part of the control group. Partners in the control group were not invited to look at the patient’s back with the dermatologist. The dermatologist did not engage in verbal discussions of the features of possibly concerning lesions. While dyads in the control group may have discussed how to check their skin with other dermatologists or searched for such guidance on the Internet, this information was not included in the surveys in order to limit cueing participants.
All measures were drawn from previous literature on SSEs.16,17 Relationship and partner variables were measured before the intervention, and SSE self-efficacy was measured 12 months after the intervention.
Patient SSE Self-efficacy
Patient self-efficacy in performing the SSE was measured with 16 items that assessed confidence in checking different skin areas (eg, “I am very confident I can carefully examine the skin on my face”), confidence in being able to have their partner’s assistance (eg, “I am very confident that I can have my partner help check the places that I cannot see myself [such as my back]”), and the correct way to perform an SSE (eg, “I am very confident that I know how to examine a mole for an irregular border”). Response options were on a 5-point Likert scale that ranged from –2 (strongly disagree) to 2 (strongly agree). Items were summed for a composite score of self-efficacy (α = .90).
The Spouse Treatment Mediation Inventory18 was adapted to measure various aspects of partner support. Partners reported on 2 items used to assess motivation in the program: “I am highly motivated to make changes through participating in this skin exam program,” and “I believe that participating in this skin exam program will help me.” Partners responded on a 5-point Likert scale that ranged from –2 (strongly disagree) to 2 (strongly agree). The 2 items were summed to create a single index of motivation. The 2 items were significantly correlated (Pearson r = 0.68; P < .01).
The Dyadic Adjustment Scale19,20 was adapted to assess the quality of the relationship between the patients and partners. Three aspects of the patient-partner relationship were assessed: agreeability between the patient and partner, activities performed together, and happiness in the relationship. Agreeability was measured using 14 items that asked participants how often they agree on various topics, such as handling financial matters, amount of time spent together, and career decisions, using a 5-point Likert scale (–2 [never agree] to 2 [always agree]). The 14 items were measured for a single index of agreeability, and the items had good reliability (α = .91). Participants were asked to report using a 5-point Likert scale ranging from –2 (never) to 2 (always) on the frequency with which they perform 4 activities with their partner (eg, “Have a stimulating exchange of ideas”). These 4 items were summed for a composite score of activities (α = .78). Happiness was assessed by agreement with the statement, “I am happy in my relationship with my partner.” Participants were asked to respond using a 5-point Likert scale that ranged from –2 (strongly disagree) to 2 (strongly agree).
Based on an estimated 20% attrition throughout the study, a sample size of 430 patients and their partners was chosen.9 The present study used moderated regression modeling described by Jaccard and Turrisi.21 No statistically significant differences in the SSE outcomes were found between the intervention arms9; therefore, all individuals in the training intervention groups (ie, workbook, in-person, and tablet groups; n = 395) were merged into 1 intervention group and compared with the control group (n = 99). Patient SSE self-efficacy was the outcome measure, with 3 independent variables in the regression model that included the intervention variable (control or training intervention), the mean centered moderator variable (eg, partner motivation or relationship quality), and the product term of the moderator variable × the intervention variable. These product terms, or interaction effects, were computed using mean centered moderator variables (partner motivation or relationship quality, consisting of agreeability between the patient and partner, activities performed together, and happiness). The regression coefficient for the product term provides an estimate of the magnitude of the moderator effect. Significance was determined by observing a t value associated with the regression coefficient greater than 2. Post hoc analyses were conducted to understand the nature of the moderator effects that were found to be significant. These analyses involved comparing the treatment and control groups on self-efficacy at 3 different levels of the moderator using Tukey tests (eg, at below average agreeability, average agreeability, and above average agreeability).
At baseline, no significant differences in the demographic characteristics of the 494 patients were observed between the intervention arms and the control group (P > .05 for all) (Table 1). The patient population had a mean age of 55 years, and 253 (51.2%) were female. Partners were also a mean of 55 years, and 280 (56.7%) were female (Table 2).
Patient SSE Self-Efficacy
The intervention increased patient SSE self-efficacy scores from baseline to 4 months (mean difference, 7.87). Individuals in the intervention group also had an increase in patient self-efficacy scores from 4 to 12 months (mean difference, 2.14; F2,340 = 68.72; P < .001) (Table 3).
We observed no significant moderated effect of partner motivation on SSE self-efficacy (β = –0.328, SE = 1.024, t = –0.320; P > .05), which indicates that the effect of the intervention on self-efficacy did not change across varying levels of partner motivation. No follow-up post hoc tests were needed. There was no significant moderator effect of agreeability with the partner (β = –0.262, SE = 0.148, t = –1.773; P > .05), which indicates that the effect of the intervention on self-efficacy did not change across varying levels of agreeability. No follow-up post hoc tests were needed.
In contrast, we observed significant moderator effects for activities performed with the partner (β = –0.892, SE = 0.432, t = –2.066; P < .05) and happiness in the relationship (β = –4.586, SE = 2.044, t = –2.24; P < .05). Results of tests using Tukey post hoc analysis revealed that the intervention group had significantly higher levels of SSE self-efficacy relative to those in the control group at all 3 levels of activities performed with the partner (eg, below average, average, and above average; P > .05 for all). However, the mean difference between the intervention and control groups was the smallest when levels of activities performed with the partner were above average (mean difference, 2.489) and greatest when levels of activities performed with the partner were below average (mean difference, 6.652) (Figure 1).
For the nature of the moderator effects of happiness, results of tests using Tukey post hoc analysis revealed the mean difference in patient SSE self-efficacy between the intervention and control groups was smallest when happiness was above average (mean difference, 1.957) and greatest when happiness was below average (mean difference, 7.000; both P < .05) (Figure 2). All main effects were significant, positive, and in the expected direction. The intervention was associated with increased self-efficacy, and, as partner motivation and relationship quality increased, patient self-efficacy increased (eTable in Supplement 2). All pairs, even those who reported less happy relationships, benefited from the SSE intervention.
This study demonstrates the sustainable effects of the SSE intervention on patient SSE self-efficacy with cohabitating and noncohabitating partners. Because many patients who did not have a cohabitating partner expressed a desire to benefit from the training intervention with a lifelong friend or relative as the partner performing skin checks, eligible partners were expanded to include a noncohabitating skin-check partner trained with the patient. Secondary findings establish the effect of variables of relationship quality (activities performed with the partner and happiness) on how the intervention affects patient SSE self-efficacy. The greatest benefit of SSE education was identified in those who spend the least time together and have the least happiness in their relationship. Partner motivation to assist with SSE did not affect patient SSE self-efficacy.
Previous work with a different cohort confirmed that the SSE intervention increased patient SSE self-efficacy at 4 months after the intervention,8,10,11 and this study showed that increased patient SSE self-efficacy was maintained in the intervention group 1 year after the intervention. One recent study has shown that SSE performance was also maintained in the treatment group 1 year after the intervention.9 This finding suggests that the SSE intervention provided skills that patients with melanoma and their partners continued to use long after the didactic training was completed. Future studies should continue to examine how the SSE intervention affects patient SSE self-efficacy as well as SSE behavior beyond 1 year to determine the duration of the intervention’s effects.
This study identified 2 significant moderators of the intervention—activities patients performed with their partners and happiness in the relationship—as well as 2 variables that did not moderate the intervention’s effect on patient SSE self-efficacy—partner motivation and patient-partner agreeability. These findings have implications regarding SSE training recommendations. First, individuals receiving the training intervention consistently had increased patient SSE self-efficacy compared with the control group, suggesting that all patients may benefit from SSE education. Second, dermatologists and health care professionals may be comfortable recommending SSE training for patients and their skin-check partners, even if there are reservations about either the partner’s motivation to assist with SSEs or the agreeability between the patient and the partner. Since neither of these variables moderated the intervention’s effect on patient SSE self-efficacy, it indicates that patients benefited equally from the intervention at all levels of partner motivation and agreeability between themselves and their partners. Thus, older adult patients with melanoma may benefit from engaging their child to assist with checking locations that are hard to see. Last, our findings suggest that patients with the lowest levels of reported activities performed together and/or the lowest levels of happiness in their relationship actually seem to benefit the most from SSE training. For example, the dermatologist may observe that the patient and the accompanying person verbally correct each other and seem to have a contentious relationship. In this study, the dermatologist invited the skin-check partner to participate in helping the patient, gave the partner a task, and empowered the partner by informing the pair that the partner was doing a “good job.” While dermatologists are most likely not going to be able to change the relationship quality, pairs who were given an activity—partner-assisted SSE—to perform together did so; thus, pairs with low relationship quality increased their activities performed with their partner. Since these individuals showed the largest increase in patient SSE self-efficacy after they received the SSE training, dermatologists or health care professionals should consider recommending SSE training for these individuals and empowering their partners.
We believe that these findings are of specific importance and may have implications for other clinical training interventions outside of dermatology. By nature, physicians and other health care professionals may be wary of recommending patient-partner training if the patient reports low levels of relationship quality. However, our findings suggest that for efficacious programs, relationship quality may not influence the intervention’s effect when compared with customary care and that pairs with lower relationship quality may benefit most from the patient-partner interventions.
The study is not without some limitations. First, the research relies on self-reports; however, previous work has shown that individuals are, for the most part, reliable reporters of their skin-related preventive behaviors.10 Furthermore, the informed consent procedures made the dyads aware that they were involved in a study examining SSEs. Last, previous work found that partner motivation was a significant moderator of SSE self-efficacy at 4 months.11 This finding is contrary to the 12-month findings in this cohort; however, this study included only dyads and educated the patient and partner together, whereas previous work compared the SSE training intervention with patients only (solo treatment or control) vs the patients with their spousal partner (dyadic training). Thus, the skin-check partners in this sample may have higher motivation, specifically because part of the inclusion criteria of this study was to have a partner willing to participate in the study. Future work should attempt to get information from partners who are not willing to help assist with SSE in order to identify barriers. Further research examining the efficacy of the intervention under real-world conditions can ask whether patients will maintain their SSE self-efficacy if they are not completing regular surveys and receiving reinforcement by the dermatologist.
This analysis provides insight on how patient and skin-check partner characteristics and perceptions can affect partner-assisted SSE training intervention. Our findings highlight the sustainability of the intervention’s effects both across time and with varying levels of partner motivation to engage in SSE and relationship quality. In addition, our findings have important clinical implications for recommending SSE training. Specifically, patients with lower levels of partner motivation or relationship quality may benefit from training in SSE with their partners.
Accepted for Publication: July 3, 2015.
Corresponding Author: June K. Robinson, MD, Department of Dermatology, Northwestern University Feinberg School of Medicine, 132 E Delaware Pl, Ste 5806, Chicago, IL 60611 (june-robinson@northwestern.edu).
Published Online: September 30, 2015. doi:10.1001/jamadermatol.2015.2819.
Author Contributions: Drs Turrisi and Robinson had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Turrisi, Mallett, Robinson.
Acquisition, analysis, or interpretation of data: Hultgren, Turrisi, Ackerman.
Drafting of the manuscript: Hultgren, Turrisi, Ackerman.
Critical revision of the manuscript for important intellectual content: Hultgren, Turrisi, Mallett, Robinson.
Statistical analysis: Hultgren, Turrisi.
Obtained funding: Turrisi, Mallett, Robinson.
Administrative, technical, or material support: Ackerman, Robinson.
Study supervision: Mallett, Robinson.
Funding/Support: This study was supported by grant R01 CA154908 from the National Cancer Institute (Dr Robinson).
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: Dr Robinson is the editor of JAMA Dermatology but was not involved in the editorial review or the decision to accept the manuscript for publication.
2.Schneider
JS, Moore
DH
II, Mendelsohn
ML. Screening program reduced melanoma mortality at the Lawrence Livermore National Laboratory, 1984 to 1996.
J Am Acad Dermatol. 2008;58(5):741-749.
PubMedGoogle ScholarCrossref 3.Bradford
PT, Freedman
DM, Goldstein
AM, Tucker
MA. Increased risk of second primary cancers after a diagnosis of melanoma.
Arch Dermatol. 2010;146(3):265-272.
PubMedGoogle ScholarCrossref 4.DiFronzo
LA, Wanek
LA, Elashoff
R, Morton
DL. Increased incidence of second primary melanoma in patients with a previous cutaneous melanoma.
Ann Surg Oncol. 1999;6(7):705-711.
PubMedGoogle ScholarCrossref 5.Robinson
JK, Turrisi
R. Skills training to learn discrimination of ABCDE criteria by those at risk of developing melanoma.
Arch Dermatol. 2006;142(4):447-452.
PubMedGoogle ScholarCrossref 6.Berwick
M, Oliveria
S, Luo
ST, Headley
A, Bolognia
JL. A pilot study using nurse education as an intervention to increase skin self-examination for melanoma.
J Cancer Educ. 2000;15(1):38-40.
PubMedGoogle Scholar 7.McPherson
M, Elwood
M, English
DR, Baade
PD, Youl
PH, Aitken
JF. Presentation and detection of invasive melanoma in a high-risk population.
J Am Acad Dermatol. 2006;54(5):783-792.
PubMedGoogle ScholarCrossref 8.Robinson
JK, Turrisi
R, Stapleton
J. Efficacy of a partner assistance intervention designed to increase skin self-examination performance.
Arch Dermatol. 2007;143(1):37-41.
PubMedGoogle ScholarCrossref 9.Turrisi
R, Hultgren
B, Mallett
KA, Martini
M, Robinson
JK. Comparison of efficacy of differing partner-assisted skin examination interventions for melanoma patients: a randomized clinical trial [published online June 7, 2015].
JAMA Dermatol. doi:
10.1001/jamadermatol.2015.0690.
Google Scholar 10.Robinson
JK, Turrisi
R, Stapleton
J. Examination of mediating variables in a partner assistance intervention designed to increase performance of skin self-examination.
J Am Acad Dermatol. 2007;56(3):391-397.
PubMedGoogle ScholarCrossref 11.Robinson
JK, Stapleton
J, Turrisi
R. Relationship and partner moderator variables increase self-efficacy of performing skin self-examination.
J Am Acad Dermatol. 2008;58(5):755-762.
PubMedGoogle ScholarCrossref 12.Bovbjerg
VE, McCann
BS, Brief
DJ,
et al. Spouse support and long-term adherence to lipid-lowering diets.
Am J Epidemiol. 1995;141(5):451-460.
PubMedGoogle Scholar 13.Garay-Sevilla
ME, Nava
LE, Malacara
JM,
et al. Adherence to treatment and social support in patients with non-insulin dependent diabetes mellitus.
J Diabetes Complications. 1995;9(2):81-86.
PubMedGoogle ScholarCrossref 14.Trief
PM, Ploutz-Snyder
R, Britton
KD, Weinstock
RS. The relationship between marital quality and adherence to the diabetes care regimen.
Ann Behav Med. 2004;27(3):148-154.
PubMedGoogle ScholarCrossref 15.Robinson
JK, Gaber
R, Hultgren
B,
et al. Skin self-examination education for early detection of melanoma: a randomized controlled trial of Internet, workbook, and in-person interventions.
J Med Internet Res. 2014;16(1):e7.
PubMedGoogle ScholarCrossref 18.Ager
RD, Thomas
EJ. The Spouse Treatment Mediation Inventory (STMI). In: Fischer
J, Corcoran
K, eds. Measures for Clinical Practice.2nd ed. New York, NY: Free Press; 1994.
21.Jaccard
J, Turrisi
R. Interaction Effects in Multiple Regressions. 2nd ed. Newbury Park, CA: Sage Publications; 2003.