Impact of the Appalachian Career Training in Oncology (ACTION) Program on Undergraduate Participants

Abstract

Objective

Kentucky ranks first in the nation in cancer incidence and mortality. The Appalachian region of the state experiences the highest cancer disparities due to inequities in many social determinants of health. As a strategy for addressing cancer and education disparities in the region, the Appalachian Career Training in Oncology (ACTION) Program at the University of Kentucky Markey Cancer Center engaged 16 Appalachian-native undergraduate student participants annually in cancer-related activities.

Methods

Students were recruited on an annual basis for the two-year ACTION program. Entry, mid-point, and exit surveys were administered to participants. Classical test theory and exploratory factor analysis were used to examine the instruments used for program evaluation, whereas repeated measures ANOVA, paired t-tests, chi-squared, and post-hoc analyses were used to analyze six years of survey data.

Results

There was a significant increase in participants’ understanding of cancer-related topics among the entry, midpoint, and exit surveys (p < .001) and a significant increase in comfort with research, clinical, and outreach activities between entry and midpoint and entry and exit (p < .001), but not between midpoint and exit. With respect to research skills and perceptions of the program, increases in mean scores were observed between midpoint and exit, but these increases were not statistically significant (p = .167, p = 223, respectively).

Conclusion

ACTION increased participants’ understanding of cancer-related topics; comfort with research, clinical, and outreach activities; and research skills. These data suggest that ACTION has a significant impact on participants.

Introduction

Cancer results in over 600,000 deaths in the United States, making it the second leading cause of death [1]. There was an estimated 1.8 million new cancer cases in 2021 in the United States of which approximately 30,000 new cases originated in Kentucky; Kentucky experiences the highest cancer incidence and mortality rates in the country [2] .The Appalachian region, which makes up 54 of 120 counties in the state, carries the heaviest cancer burden. In rural, Appalachian Kentucky residents are 8% more likely to develop a preventable, cancerous malignancy than those living outside the region [3]. This disparity is caused by various factors, including obesity and tobacco use [4]. Social determinants of health, such as a lack of access to health care facilities and low socioeconomic status, also contribute to the high cancer rates in Appalachia [5].

Furthermore, Kentucky ranks 45^th in education in the United States [6]. Low education levels in the Appalachian region contribute to decreased cancer literacy, which is defined as a person’s ability to understand and make appropriate decisions based on the advice of a healthcare professional. Cancer literacy is essential to proper cancer treatment and prevention, as a patient’s education level directly correlates to their ability to make informed healthcare decisions and receive timely screenings [7]. Recently, we have demonstrated that cancer education interventions increase cancer literacy in middle and high school students living in Appalachian Kentucky [8]. These students represent a population willing to share cancer-related advice with their friends and family, encouraging them to adopt healthy lifestyle choices [8].

Cancer education programs at the undergraduate level have a longstanding history of impact on students. One such ten-week program at Diné College involving Native American undergraduate students produced significant changes in students’ attitudes and opinions towards cancer [9]. Another program, conducted at the University of Louisville, involved undergraduate students from various underserved populations in cancer research activities. Results demonstrated significant increases in student interests in cancer research following program completion [10]. A third program, which is a National Cancer Institute Youth Enjoy Science (NCI YES) program, involved African American and Hispanic students in an eight-week intensive cancer research program at Case Western University. The data collected at the end of the program showed that most students reported substantially increased cancer knowledge [11]. These studies demonstrate the success of cancer-related experiential programs on undergraduate students.

While each of the above cancer research education programs engaged underserved students, the report herein focuses on a unique group of underserved students from Appalachian Kentucky — students from rural and/or economically distressed counties classified in the bottom 10% nationwide [12]. The Markey Cancer Center at the University of Kentucky (UK) created the Appalachian Career Training in Oncology (ACTION) Program in 2016 and it is now an NCI YES R25 program that engages Appalachian Kentucky students in cancer-related research education activities. The ACTION Program selects a cohort of eight UK undergraduates each year and involves them in cancer research, education, clinical observations, and community outreach activities for two consecutive years, as a means of influencing their preparation for and pursuit of a cancer-related career. Between 2016 and 2021, six cohorts of undergraduate students entered the program. The characteristics and general components of the ACTION Program have recently been described elsewhere [13].

In this paper, we provide evidence for the effectiveness of the ACTION Program on the first four cohorts of students. The data analyzed in this paper was collected over six years from 2016 to 2021. Overall, the data suggest that the ACTION Program successfully improves students’ 1) understanding of cancer-related topics; 2) comfort with research, clinical, and outreach activities; 3) research skills; and 4) perceptions of the program. The findings of this report represent a significant contribution to the literature regarding a diversified and comprehensive cancer education program for rural, underserved students.

Methods

Measures

Data were collected by instruments developed by the ACTION Program Director. The entry instrument (Supplemental Document 1) consisted of three demographic-type questions and eight Likert-type questions. The response options for the Likert-type items included 1 (Very knowledgeable or Very comfortable), 2 (Knowledgeable or Comfortable), 3 (Somewhat knowledgeable or Somewhat comfortable), and 4 (Not knowledgeable or Not very comfortable). The midpoint and final instruments (Supplemental Document 2) contained 24 additional items with the five response options 1 (Agree or Very positive), 2 (Somewhat agree or Positive), 3 (Neutral), 4 (Somewhat disagree or Negative), and 5 (Disagree or Very negative).

Procedure

Each year from 2016 to 2021, the ACTION Program recruited eight UK undergraduate students to participate in program activities over the course of the program. Given the two-year nature of the program, 16 undergraduates participate each year after the inaugural year. Participants were recruited from students enrolled at UK who were in or planned to major in a health or science-related field. To facilitate participant recruitment, ACTION Program personnel partnered with academic programs, the honor’s college, the Office of Undergraduate Research and other experiential pipeline programs or student support services on campus.

Participants completed surveys at three points throughout their participation in the program: upon program entry (T1), after the first year of program engagement which represents the midpoint of the program (T2), and upon program completion when students had completed their two years in the program (T3). All instruments were developed and administered electronically through RedCap. It is important to note that only cohorts one through four completed all three rounds of surveys at the time this manuscript was written. This report represents longitudinal analyses of six years of data from 2016 to 2021.

Data Analysis

Classical Test Theory analyses were conducted in SPSS. Inter-item correlations and item-to-total correlations were used to determine strength of relationships among items and the appropriateness of using them as one scale. Item-to-total correlation guidelines of 0.3 or more were considered acceptable. The internal consistency of the scales was assessed by Cronbach alpha (α), which indicates how well the items are measuring the same underlying construct. If Cronbach alpha was improved by removing the item, then the item was removed. Reliabilities of subscales at different time points were calculated. Reliability for the Understanding Cancer-Related Topics upon entry with 32 undergraduates was α = .777, for the Comfort with Research, Clinical and Outreach Activities was α = .715, Research Skills was α = .839 at T2, which was the first administration, and Perceptions of the Program was α = .751 at T2, which was also the first administration.

Exploratory factor analyses (EFA) were conducted with SPSS to explore the measurement instrument’s dimensionality or internal structure. Overall, the EFA established the presence of four underlying constructs. Two factors were extracted from the first eight Likert items, creating two subscales. The first subscale was named Understanding Cancer-Related Topics and was composed of four items, whereas the second subscale was named Comfort with Research, Clinical, and Outreach Activities and contained four items. Six items from the midpoint and final instruments formed a third subscale named Research Skills. The final subscale was also formed from items in the midpoint, and final instrument was Perceptions of the Program and contained six items.

All data were checked for normality, influential points, and outliers. One-way repeated measure ANOVA was used to compare different subscale means across repeated measurements of time. Dependent variables included total scores on various subscales, whereas the independent variable was time. Students’ scores were matched across administrations (i.e., entry, midpoint, final) using majors and cohorts as identifiers, and only complete sets of observations were used as required by repeated-measures ANOVA. A paired-samples t-test was conducted to determine if there was a significant difference in means when there were only two timepoints. ANOVA was also conducted to determine if there were significant differences among student perceptions of program elements (i.e., research, clinical shadowing, education, outreach, mentoring). The Mann-Whitney U test, a rank-based nonparametric test used to determine if there are differences between two groups on a continuous or ordinal dependent variable, was used as the nonparametric alternative to the independent-samples t-test when data are not normally distributed. Post-hoc analyses using Tukey’s Honestly Significant Difference (HSD) were used to compare all possible pairs of means. While an ANOVA indicates if results are significant overall, it will not tell a researcher where differences lie; however, Tukey’s HSD is often run to determine which group means are different. A chi-square test of independence was conducted in R to determine if there were statistically significant associations at different time points between types of majors and career paths.

Results

Demographics

This study examined the impact of the ACTION Program on 32 undergraduate students at UK. Data were collected from four cohorts of undergraduates from 2016 to 2021. Most students were sophomores (n = 16, 50.0%) at the beginning of the program, and 72% were female. Furthermore, 91% were White, 28% were first-generation college students (that is, first in their family to have postsecondary education experience), 19% were from low-income families (designated using the poverty thresholds defined by the US Census Bureau), and 97% were from rural counties (designated using the Health Resources and Services Administration’s definition of rural) (Table 1). It is important to note that ACTION undergraduate participant race/ethnicity demographics generally align with those of the Appalachian Kentucky region, especially with regard to the overrepresentation of White participants. The region’s population is 94.4% White, 1.9% Black, 1.9% American Indian/Alaskan, Asian, Native Hawaiian and other Pacific Islander or two or more races, and 1.8% Hispanic or Latino [14]. All participants were natives of Appalachian Kentucky counties that are defined as rural, low-income, and/or health professional shortage areas.

Table 1.

ACTION Participant Demographics.

Parameter	Frequency N= 32	Percent
Academic Level
Freshman (at entry – T1)	5	15.6%
Sophomore (at entry – T1)	16	50.0%
Junior (at entry – T1)	11	34.4%
Senior (at entry – T1)	0	0%
Gender
Male	9	28.1%
Female	23	71.9%
Race / Ethnicity
American Indian / Alaska Native	0	0%
African American/Black	0	0%
Asian	3	9.4%
More than one race	0	0%
White	29	90.6%
Hispanic or Latino	0	0%
Not Hispanic or Latino	32	100%
Disparity Status
1st generation student	9	28.1%
Low income*	6	18.8%
Rural resident**	31	97.0%

^*Low income students were identified through self-reported taxable income. Low-income levels are those defined by the US Census Bureau.

^**Rural areas are as designated by the Health Resources and Services Administration.

Career Paths and Majors

Upon entry to the program, most students were biology majors (44%) followed by those majoring in agriculture and medical biotechnology (19%), human nutrition and public health (13% each), and neuroscience (9%) (Supplemental Table 1). Medicine was the most prominent career path identified by students (69%) at the beginning of the program, whereas a combination of medicine and research was the second most popular career path identified by students (25%). Research and outreach/education were selected by only one student each (3%) (Supplemental Table 2). Chi-squared tests of independence were conducted in R to determine if the frequencies of majors and career paths identified by students at the beginning of the program were significantly different from those identified at the end of the program. A correction was made for small sample size. Neither test showed statistically significant results (χ²(2) _majors = 1.285, p = .77327; χ²(6)_{career paths} = 1.9608, p = .9233). Nonetheless, as students progressed through the program, six changed majors, but not enough to change the order of popularity except in the case of public health, which became only 6% of the students’ majors. In addition, a new major, human health science, was selected by one student (Supplemental Table 1). In terms of career paths, medicine remained the most popular career path throughout the program (71%) followed by medicine and research (21%) (Supplemental Table 2).

Understanding Cancer-Related Topics Subscale

A one-way repeated measures ANOVA was conducted to determine whether there were statistically significant differences in students’ Understanding of Cancer-Related Topics throughout the two-year program measured at three time points, entry, midpoint, and endpoint. There were no outliers, and the data for entry (T1) and midpoint (T2) were normally distributed, as assessed by boxplots, Q-Q plots, and a Shapiro-Wilk test (p > .05). Mauchly's Test of Sphericity indicated the assumption of sphericity had not been violated, χ²(2) = 2.833, p = .243. Students’ Understanding of Cancer-Related Topics was statistically significantly different at different time points during the program, F(2, 54) = 21.827, p < .001. The sample effect size based on within-subjects factor variability was measured by partial eta squared (η_p² = .447) and was considered moderate to large. There was a significant increase in mean student Understanding of Cancer-Related Topics from entry (T1) to midpoint (T2) (differences in means = 2.893, p < .001) and entry (T1) to final (T3) (differences in means = 4.214, p < .001) but not from midpoint (T2) to final (T3) (differences in means = 1.321, p = .082). See Table 2, Figure 1A.

Table 2.

Descriptive Statistics for Subscales at Entry, Midpoint and Final Administrations.

Subscale	Mean T1 (SD)	Mean T2 (SD)	Mean T3 (SD)	F statistic	U Statistic	p-value	Partial Eta Square ηp2
Understanding Cancer-Related Topics n = 28	12.07 (2.36)	14.96 (1.73)	16.29 (2.93)	21.827	NA	p < .001	.447
Comfort with Research, Clinical and Outreach Activities n = 18	13.93 (2.51)	16.48 (1.87)	16.79 (2.47)	11.459	NA	p < .001	.306
Research Skills n = 27	NA	28.47 (2.36)	29.26 (1.40)	NA	293	p = .167	NA
Perceptions of the Program n = 28	NA	27.72 (2.37)	28.46 (1.92)	NA	319	p = .223	NA

NA = Not applicable.

Fig 1.

Changes in participants’ understanding, comfort, research skills and perceptions over time.

Comfort with Research, Clinical, and Outreach Activities Subscale

The same procedures were followed for the construct Comfort with Research, Clinical, and Outreach Activities. There was one outlier in the midpoint data (T2) detected by boxplots and stem & leaf plots, and the entry and midpoint data (T1 and T2) were normally distributed, as assessed by boxplots, Q-Q plots, and a Shapiro-Wilk test (p > .05). Mauchly’s Test of Sphericity indicated a violation of the sphericity assumption, χ²(2) = 7.267, p = 0.026, therefore, Huynh-Feldt (ε = 0.843) corrected results are reported. Results for students’ Comfort with Research, Clinical and Outreach Activities mean scores were statistically significantly different at different time points during the program (F(1.685, 43.814) = 11.459, p < .001) with a large effect size (η_p² = .306). There was a significant increase in mean student Comfort with Research, Clinical and Outreach Activities from entry (T1) to midpoint (T2) (differences in means = 2.556, p < .001) and entry (T1) to final (T3) (differences in means = 2.852, p = .004) but not from midpoint (T2) to final (T3) (differences in means = 0.296, p = 1). See Table 2, Figure 1B.

Research Skills Subscale

Students’ ability to perform specific Research Skills, including scientific thinking, experimentation, scientific communication, critical thinking, teamwork, and troubleshooting, was assessed at the midpoint (T2) and final (T3) data collection points. Two outliers were detected in the midpoint (T2) dataset by boxplots and stem & leaf plots, and one with a high Cook’s Distance (Cook’s D) value was removed. Cook’s D is used for assessing influential observations in regression models usually with parametric data – in this paper it is used solely as a rough guide. There were also two outliers in the final dataset (T3); however, those cases did not have a high Cook’s D value and were kept in the dataset. The entry and midpoint datasets (T1 and T2) were not normally distributed, as assessed by boxplots, Q-Q plots, and a Shapiro-Wilk test (p < .05). Log transformations were attempted but were unsuccessful; therefore, a Mann-Whitney U test was run to determine if there were differences in Research Skills score between midpoint (T2) and final (T3) datasets. Results showed that although the mean scores increased from the midpoint (28.47) to final (29.26) there were no statistically significant differences between midpoint (T2) and final (T3) time points (U = 293, z = −1.393, p = .167), using an exact sampling distribution for U. See Table 2, Figure 1C.

Perceptions of the Program Subscale

The final ACTION construct evaluated was students’ Perceptions of the Program instructional strategies, including research, clinical shadowing, education, outreach, mentoring, and overall perceptions. Perceptions were measured at midpoint (T2) and final (T3), with one item per instructional strategy. One extreme outlier in the midpoint dataset (T2) with a high Cook’s D value was removed. Two outliers existed in the final (T3) dataset and were detected by boxplots and stem & leaf plots. Only one had a moderately high Cook’s D value; however, neither was removed. The midpoint (T2) and final datasets (T3) were not normally distributed, as assessed by boxplots, Q-Q plots, and a Shapiro-Wilk test (p < .05); therefore, a Mann-Whitney U test was run to determine if there were differences in Perceptions of the Program score between midpoint (T2) and final datasets (T3). Although mean scores increased from 27.72 to 28.46, there was not a statistically significant difference between midpoint (T2) and final (T3), U = 319, z = −1.230, p = .223 using an exact sampling distribution for U. Moreover, the Perceptions of the Program scores were already positively skewed at the midpoint with a skewness values of −1.107 and only became more skewed at final administration (−1.544). See Table 2, Figure 1D.

Additional Student Outcomes

In addition to improvements in cancer knowledge, comfort, and skills, the ACTION Program provided students with opportunities that helped to shape their future careers. For example, eleven students in the first four cohorts of ACTION were coauthors of 15 peer-reviewed publications. Some students also wrote essays detailing their personal experiences with cancer in their families and communities. These stories were included in a book entitled, The Cancer Crisis in Appalachia: Kentucky Students Take ACTION, which was published by the University of Kentucky Press [15]. Many students are also matriculating to professional and graduate programs. Of the 32 students from the first four cohorts, 17 are in medical school, two are in pharmacy school, three are in a physician assistant program, and one is enrolled in a cancer biology Ph.D. program. Three other students work in STEM-related jobs, one student joined the military, and five have not yet graduated.

Discussion

This study examined the impact of the ACTION Program on 32 undergraduate students at UK. Data were examined using chi-squared tests to determine if there was a significant change with respect to students majors and career paths; no significant change was found, but the data showed that a few students changed majors and vacillated between career paths designated as “medicine” and “medicine and research” at the midpoint and final time points. The ACTION program has the right programmatic elements to maintain students’ interest in STEM-related career paths. Feist [16] reported that many talented students know early, sometimes at a young age, they want to enter a STEM career, which may be reflected in these results. Challenging hands-on, minds-on academic experiences, especially at the secondary and post-secondary levels, influence the choice of a science major for men and women [17].

Next, student data were examined to determine significant changes regarding students’ Understanding of Cancer-Related Topics across all time points. Results indicated that the mean Understanding of Cancer-Related Topics scores steadily increased from entry to the final administration of the instrument. In addition, a multiple comparison analysis showed that significant differences existed between the means at all time points. The effect size (η_p² = .447) was considered large. These are important results, indicating that the ACTION program had a large impact on the students’ self-reported understanding of cancer topics, and the impact occurred throughout the program.

Students provided information on their levels of Comfort with Research, Clinical, and Outreach Activities. Once again, results showed that the means for student Comfort with Research, Clinical, and Outreach Activities increased from the beginning of the program to the end, and there were significant differences in the means across time with a large effect size (η_p² = .632). A multiple comparison analysis showed significant differences existed between the means at entry and midpoint and entry and final administrations but not midpoint and final administrations. These contrasts seem logical because most students are becoming accustomed to the environment in the first few months of the program and are adjusting to the atmosphere of the cancer-related world. After this initial period, that environment is no longer new, and students may feel more at ease; thus, their comfort level stabilizes.

The assessment of students’ Research Skills is an essential feature of many program evaluation plans and is often used to determine the attributes of undergraduate research programs, including students’ ability to “think like a scientist” and improved oral and written communication [18]. By engaging with different scientific community members in undergraduate research programs, students’ skill development increases, and they construct their identity as scientists [19]. Thiry et al. [18] explored research skill outcomes between novice and experienced undergraduate researchers and found that almost all students described gains in research skills such as collecting scientific data.

In this study, the research skills of scientific thinking, experimentation, scientific communication, critical thinking, teamwork, and troubleshooting were assessed at the midpoint and final administrations. Although scores increased, there was no significant difference between the means at these two time points. One potential cause for the lack of non-significant results is students’ inability to distinguish what they know from what they do not yet know in terms of research skills. Serra and DeMarree [20] claimed that students’ perceived abilities are often misaligned with their actual knowledge. If this is the case, students’ self-report data on their research may be over-estimated at early time points in the program.

Nonetheless, the fact that these cancer research students reported increases in research skills is important and consistent with Thiry et al.’s [18] findings in which he found that novice researchers reported developing basic skills, such as gaining a greater understanding of the process of scientific research. Many ACTION students can be classified as novice researchers because they came from underserved, rural Appalachian high schools with little to no laboratory experience.

The implications of improved student research skills may be substantial. Bowman and Holmes [21] claimed that first-year participation in research does not significantly affect students’ first-year G.P.A., but does have an apparent effect on fourth-year G.P.A. This is because cognitive skills learned during research participation in the first-year benefit students in more advanced coursework, which is more cognitively demanding. Previous literature has shown that gains in critical thinking, writing, and communication skills due to research participation are often related to higher-level coursework [22]. Thus, the potential of the ACTION program to increase students’ research skills will impact their academic careers far beyond their participation in the two-year program.

The final construct measured was students’ Perceptions of the Program instructional strategies, including research, clinical shadowing, education, outreach, mentoring, and overall perceptions, measured at the midpoint and final administrations. Analysis showed there were no significant differences between the means from midpoint to final administrations; however, mean scores increased for all instructional strategies, and there was an increase in the overall mean. However, results indicated there was not a significant increase in mean scores between the midpoint and final on the Perceptions of the Program subscale.

Few programs explore students’ perceptions of program instructional elements, but their findings were similar to those of the ACTION program when they did. Coronado and colleagues [23] explored minority undergraduate interns’ perceptions of a cancer research program. Students identified mentoring as a key component of the training program and rated informal time with one’s mentor highly. Evaluations by the National Research Council [24] showed that students who participated in the NIH T34*STAR program highly rated acquiring research skills, the opportunity to travel to conferences and present their research, mentoring, and networking. Abraham et al. [25] explored middle and high school students’ preferences for cancer education. Their research showed that students preferred interactive, relatable, and engaging educational content delivered online videos, personal presentations, and educational games. ACTION students did not state preferences for specific modalities, but different instructional delivery modalities were used throughout the ACTION program.

Findings from the ACTION yearly evaluation informed design improvements during each year’s program, and modifications were continuously made. In addition, the ACTION leadership team hopes that the findings from this study may inform the design of other programs that aim to increase the cancer knowledge and research skills of undergraduate students. Further and expanded evaluations of the ACTION program are needed to discern which program components have the greatest influence on undergraduate knowledge and research skills.

Limitations

While the current investigation makes a valuable contribution to the body of research on cancer outreach to students from rural, underserved and impoverished regions, some limitations should be noted. First, the study included a relatively small sample size. Using a small sample increases the chance of incorrectly labeling a false premise as true, and smaller sample sizes are not as representative of the entire population as larger samples. A future study that builds off these findings with additional students will improve the internal and external validity of the findings. Second, ACTION participant demographics generally align with the demographics of the Appalachian Kentucky region and thus the study focuses on a unique, mostly White population of students. Care should therefore be taken when generalizing these findings to other populations. Third, the study did not have a control group, which would have isolated the effect of the treatment. Fourth, this study used students’ self-reported questionnaires as the primary data source, which often carries associated risks based on variable respondent knowledge, comprehension, and interpretation of scale items. Future studies should contain items that limit the variability in personal interpretation to improve the dependability of the responses. Finally, the fifth limitation is the double-barreled nature of some items on the questionnaires. A double-barreled question is an item that asks about more than one construct within a single survey question. In this situation, there is no way of discovering the true intentions of the respondent from the data afterward. Additional studies should build off these results by taking measures to avoid double-barreled questions to improve the reliability of the findings.

Conclusion

Kentucky produces some of the highest cancer statistics in the country, and the eastern, Appalachian region sees a disproportionate percentage of cancer incidence and mortality. This cancer disparity is the result of several health behaviors and inequities in social determinants of health. The ACTION Program strives to prepare Appalachian Kentucky undergraduate students for a potential career in oncology through cancer-related activities. This study demonstrates how the ACTION Program improved students’ understanding, comfort, and research skills regarding cancer. ACTION students have contributed to numerous scientific publications. Many program graduates have continued to medical school, graduate school, or pursued other professional degrees.

Lofty goals were set when the ACTION program was created: to increase Appalachian students’ matriculation into post-secondary, professional, or graduate science degree programs; instill in students’ lifelong commitments to research, outreach, or health care careers focused on the cancer burden and unique challenges in Appalachian Kentucky; and enhance Appalachian Kentucky communities’ understanding of cancer research and clinical care, linking both to the benefits of science education. The most important, however, was to establish a new paradigm for cancer-focused education programs that serve Appalachian students, emphasizing the importance of science education in rural, low-socioeconomic, and low-education-attainment areas. This paper provides evidence that the ACTION program is on its way toward meeting those goals. If students return to their home region with their acquired knowledge and skills, they will have the opportunity to reduce the cancer disparity in Appalachian Kentucky.