Melanoma toolkit for early detection for primary care providers: A pilot study

INTRODUCTION
Primary care providers have greater access to patients despite often lacking the appropriate training or time to implement effective skin cancer screenings in their busy practices.1 Through collaborative efforts with Oregon's War on Melanoma public health campaign and other primary care trainings, we created "Melanoma Toolkit for Early Detection" (MTED): a training curriculum and resource repository for primary care providers.


METHODS AND RESULTS
MTED consisted of three self-paced modules. Each participant completed a pre- and post-test consisting of demographic, confidence, and image identification questions. A subsequent optional 6-month follow-up image identification test was also provided. Of the 96 participants, 40 completed all the modules, the pre, and post-test. On average, scores increased by 6.0 (95% CI: 3.5 to 8.6) percentage points (P<0.001, paired t-test). The percent of participants reporting confidence with melanoma identification increased significantly from the pre- (23.3%) to the post-test (67.4%), an increase of 44.2 (95% CI: 29.3 to 59.0,) percentage points (P<0.001, McNemar's test).


DISCUSSION
This study shows that an online curriculum such as MTED has the potential to increase PCPs confidence and knowledge. This could subsequently lead to improved early melanoma diagnosis by primary care providers.


| INTRODUC TI ON
The visible nature of cutaneous melanoma (CM) provides an opportunity to detect and treat suspicious lesions early, thereby potentially reducing mortality. However, one study of 216 patients with melanoma found that only 20% had an established dermatologist whereas 63% received primary care provider (PCP) evaluation within a year prior to diagnosis (Geller et al., 1992). Another study using photographs of lesions found diagnostic accuracy and management of pigmented lesions to be greater by dermatologists than PCPs (Chen et al., 2006).
PCPs report that skin cancer early detection is important, but they often lack the appropriate training or time to implement effective skin screenings in their busy practices (Goulart et al., 2011). Given greater patient access to PCPs than dermatologists, training PCPs to detect and triage concerning pigmented lesions may provide crucial support to the early detection of melanoma. Through Oregon's War on Melanoma public health campaign, we created the "Melanoma Toolkit for Early Detection" (MTED) training curriculum and resource repository for medical providers outside of dermatology. Collaborative efforts from other primary care trainings and an assessment of community needs informed the creation of MTED (Jiang et al., 2017). A novel aspect of MTED is the "Toolkit" design that allows PCPs to choose all or part of a curriculum of evidencebased training, resources, and patient educational materials. This model supports the adaptability of the MTED intervention, as PCPs can select the materials most meaningful for their practice and capability. During the training development, PCP stakeholders vocalized that they supported this learner-centered approach as they valued freedom to engage only with the material most relevant to their practice. Educational theory informed both the intervention presentation and the instrument content creation: Mayer's cognitive theory of multimedia learning suggests learning involves actively integrating, attending to, and filtering information (Mayer, 2010). We hypothesized that execution of this learning theory (which involved reducing extraneous processing of irrelevant material, aligning content with learning objectives, and providing goal-directed practice with feedback) would lead to increased knowledge and confidence in identifying benign and malignant lesions.

| ME THODS
The state of Oregon and Oregon Health and Science University (OHSU) has robust PCP networking resources that include the fol- practice change. Participants were offered continuing medical education (CME) credit upon completion of both the pre-and post-tests. selected "not confident" with melanoma identification on the pretest selected "confident" on the post-test (Figure 2).

| D ISCUSS I ON
In this pilot study, we found that MTED completion by PCPs was associated with increased confidence in diagnosing melanoma and improved ability to correctly identify melanoma. These findings suggest that an online curriculum has the potential to improve early melanoma diagnosis in the primary care setting. However, further work is needed to ensure that this training does not lead to increased PCP confidence without concomitant increased knowledge, especially surrounding when it is appropriate to biopsy a suspicious lesion.
The relatively small percentage-point score improvement in knowledge may have resulted from high baseline test scores, which suggests that a more nuanced test with more challenging lesions may be needed to reflect the knowledge gained. A prior study applying visual perception training for melanoma detection in teaching medical students with lower levels of baseline knowledge demonstrated greater effect in the post-test with sustained knowledge on repeat testing 4 weeks later (Choi et al., 2019). Our findings suggest similar results.
The primary limitation of this study is its small sample size of participants completing all three modules and tests. Reasons for this are inherent to research with PCPs who have many competing education F I G U R E 1 Melanoma toolkit for early detection (MTED) completion was associated with improvement in image classification pre-post test scores. The percentage of pigmented lesions correctly identified as benign or malignant is shown. An average of 6 percentage point improvement was seen overall (p < .001), and greater improvement was seen in the ability to identify malignant lesions F I G U R E 2 Melanoma toolkit for early detection (MTED) completion was associated with increased melanoma identification confidence. Reported confidence revealed a dramatic increase in "confident" and decrease in "neutral" and "not confident" responses, representing an overall positive shift in confidence following completion of the curriculum goals and limited time to pursue them outside their busy clinical practices. This is demonstrated in similar online trainings for PCPs. For example, Samuelson et al. published a self-paced PTSD course that only had 33% of enrolled participants completed all aspects of the study despite the duration being shorter (2 hr) and a $25 incentive in addition to CME credit (Samuelson et al., 2014). Furthermore, our learner-centered approach encouraged participants to engage in the modules most meaningful to them rather than completion of all the trainings. This resulted in a self-selected sample of highly motivated participants who were interested in engaging with the entirety of the curriculum. Future goals include improving follow-up response to measure provider long-term retention of knowledge, continued confidence, and lasting behavioral changes. Additionally, we plan to host future iterations of MTED in a learning management system which is capable of collecting more precise data on frequency and length of time participants are engaging with each module. This study shows that MTED has the potential to increase PCP confidence and knowledge, which could lead to improved PCP detection of melanoma.