Student Arm

The study included 3^rd to 8^th-grade students (N = 186) in Michigan, whose school principal is a member of Michigan Elementary and Middle School Principals Association (MEMSPA). The student arm was carried out via an age-appropriate, nine-question online Qualtrics survey. No existing psychometrically tested questionnaires were available. Two surveys were developed consisting of eight quantitative questions assessing cancer knowledge (etiology, prognosis, treatment, side effects, etc.) and one qualitative free response allowing the respondent to express apprehensions regarding a friend with cancer. Readability for the 3^rd to 5^th- grade survey, assessed by the Flesch-Kincaid Grade Level Test, was at a grade level 3.6. Readability for the 6^th to 8^th-grade survey, assessed by the Flesch-Kincaid Grade Level Test, was at a grade level 4.1.

The student surveys were distributed via email sent from the president of MEMSPA; it is estimated that it reached approximately 1,200 members. MEMSPA members were asked to forward the Qualtrics survey links for each age group as well as a parent information sheet to their school’s 3^rd to 8^th-grade parents/students. By distributing the survey link, parental consent for a child to participate in the study was assumed. Oral assent was assumed for children under the age of 14 years old. Assent from the children 14–17 years old was acquired via an embedded mandatory response consent document in the survey.

Patient Arm

This study included patients (N = 29) currently receiving cancer therapy or patients who had completed cancer therapy within the past two years at a Michigan academic medical center. The institution did not have an organized school reentry program, but did have a hospital teacher on staff at the time of the study. Patients under the age of eight (<3^rd grade) were excluded from the study. Patients over the age of 18 were included if they had been diagnosed before age 18.

The patient arm of the study was carried out via an age-corrected, eight-question online Qualtrics survey. No existing, psychometrically tested questionnaires were available. A survey was developed consisting of five quantitative questions assessing age at diagnosis, school absenteeism, return to school, and worries related to returning to school. The additional three questions were qualitative free response, allowing respondents to describe their transition back to school (if they had returned) and to comment on what they thought should be included or excluded from a school reentry program. Readability for the patient survey, assessed by the Flesch-Kincaid Grade Level Test, was at a grade level 3.8.

The survey was distributed in person during a scheduled clinic visit. Qualified patients/families were given a printed cover letter describing the study as well as a printed information sheet. Patient participants completed the survey online either via a personal electronic device utilizing the paper link or a QR code provided with the cover letter or via a tablet in the office. By distributing the survey link (or allowing survey completion via tablet), parental consent for a child to participate in the study was assumed. Oral assent was assumed for children under the age of 14 years old. Assent from the children 14–17 years old was acquired via an embedded mandatory response assent document in the survey. Consent from the patients 18 years old or older was also acquired via an embedded mandatory response consent document in the survey.

Quantitative Analysis

All of the quantitative questions were assessed with answer-choice frequency and percent. All comparisons were analyzed using the chi-square test, of which the assumptions were adequately met. Any P value <0.05 (P <0.05) indicated a statistically significant result. All analysis for the study was done in SAS 9.4 (SAS Institute Inc.).

While similar, the questions “If you have done something wrong you can get cancer?” in the 3^rd to 5^th-grade survey and “What causes cancer?” in the 6^th to 8^th-grade survey could not be combined due to different response choices. All other questions could be combined into one data set for analysis.

Qualitative Analysis

Open-ended questions were reviewed separately by two members of the research team with coding expertise to identify emerging themes. Once coding was complete, disagreements were discussed, and the final results underwent a check for inter-rater reliability using Cohen’s Kappa. The Fisher’s Exact Test was used to compare the qualitative responses of 3^rd to 5^th graders with 6^th to 8^th graders. The Fisher’s Exact Test was used because the assumption of adequate expected counts for the chi-square test was not met. Answer-choice frequency and percent were reported.

Quantitative Assessment

The majority (>70.00%) of 3^rd to 5^th-grade students (Appendix A1) and 6^th to 8^th-grade students (Appendix A2) answered correctly all of the survey questions related to etiology, prognosis, side effects, and treatment of cancer.

As described in Table 1, 10.09% of the 3^rd to 5^th graders indicated that cancer is contagious, as opposed to 0% of the 6^th to 8^th graders (P = 0.0036). There was not enough evidence to conclude that any of the other questions were significantly associated with respondents’ grade level (all P ≥ 0.05).

A total of 18.69% of 3^rd to 5^th-grade students indicated that they had a friend with cancer compared to 27.85% of 6^th to 8^th-grade students. Having a friend with cancer was not associated with respondents’ grade level (P > 0.05, Appendix A3). There was not enough evidence to conclude that any of the quantitative responses of 3^rd to 8^th-grade students were significantly associated with having a friend with cancer (all P > 0.05, Appendix A4).

Qualitative Assessment

Both the student and the patient arm of the study utilized themes that were confirmed with a Cohen’s Kappa inter-rater reliability analysis, described in Table 2.

Both overall and by question, there was Almost Perfect Agreement or Perfect Agreement between the two raters for every question. Raters’ scoring sheets were reconciled for the final coding.

The Michigan 3^rd to 8^th-grade student concerns pertaining to a friend with cancer are highlighted in Appendices B1–B4. There was no difference between the qualitative responses of 3^rd to 5^th-grade students and 6^th to 8^th-grade students on whether or not they had a friend with cancer (all P > 0.05, Appendices B5–B7).

As illustrated in Table 3, among the 3^rd to 5^th-grade students, fewer students who had a friend with cancer (12.50%) were worried that their friend might die than those who did not have a friend with cancer (55.93%, P = 0.0002). There was not enough evidence to determine whether there were any other significant theme differences between these groups (all P > 0.05).

As illustrated in Table 4, among the 6^th to 8^th-grade students, fewer students who had a friend with cancer (5.26%) were worried that their friend might die than those who did not have a friend with cancer (69.05%, P = <0.0001). There was not enough evidence to determine whether here were any other significant theme differences between these older groups (all P > 0.05).

Quantitative Assessment

An overview of age at diagnosis and when they completed the survey, school absenteeism, and what concerns patient respondents had upon returning to school may be found in Appendix C1.

When comparing age to apprehensions about going back to school (Table 5), 70% of patients 11 or younger were worried about their physical appearance going back to school and 20% of patients 12 or older were worried about how they looked going back to school. Respondents who were 11 and younger were significantly more likely to be worried about their appearance going back to school than respondents who were 12 or older (P = 0.0124). There was not enough evidence to conclude that any of the other questions were significantly associated with the age of the respondent (all P > 0.05).

There was not enough evidence to conclude that any of the multiple-choice concerns were significantly associated with the length of time out of school (all P > 0.05) (Appendix C2).

Qualitative Assessment

Patient responses when asked to describe their transition back to school, what should be included in a school reentry program, and what should not be included in a school reentry program are described in Tables 6, 7 and 8.

Quantitative Assessment

Though previous studies (Benner & Marlow, 1991; Canter & Roberts, 2012; DeLong, 1999; Treiber et al., 1986) have demonstrated an increase in peer knowledge after educational interventions, our study demonstrated that most students in both 3^rd to 5^th grade and 6^th to 8^th grade responded correctly on all survey questions related to cancer etiology, prognosis, treatment, and side effects. This suggests that peer interventions should not exclude, but also not continue to focus on disease facts if students already have a good foundation of knowledge. Instead, interventions should emphasize peer relationships and individual student apprehension. Moreover, previous studies (referenced above) have often confirmed that even with an increase in factual information, students may not change their attitudes towards a classmate with cancer. Thus, this results merely in increased student knowledge with little to no effect on the child the intervention is intended to support. Individual classroom knowledge should be considered before interventions are established. Though most students have a strong foundation of basic cancer facts, not all disease-based facts should be excluded from a peer intervention. As an example, significantly more of the 3^rd to 5^th graders (10.09%) in this study indicated that they thought cancer was contagious than the 6^th to 8^th graders (0%, P = 0.0036). Thus, younger students may be more likely to broaden the concept of contagion to noncommunicable diseases; however, only the minority of students responded in this way.

Qualitative Assessment

The three most frequently reported concerns for 3^rd to 5^th-grade students who have had a friend with cancer included (a) watching them get sick/suffer (43.75%), (b) not being able to visit/play (31.25%), and (c) worry that they might die (12.50%). For 3^rd to 5^th-grade students who had not had a friend with cancer, the three most frequently reported responses included (a) worry that they might die (55.93%), (b) watching them get sick/suffer (33.90%), and (c) not being able to visit/play (23.73%). By comparison, for 6^th to 8^th-grade students who have had a friend with cancer, the three most frequently reported concerns included (a) watching them get sick/suffer (36.85%), (b) not being able to visit/play (31.58%), and (c) outlier responses (26.32%). The top three responses for 6^th to 8^th-grade students who had not had a friend with cancer included (a) worry that they might die (69.05%), (b) watching them get sick/suffer (23.81%), and (c) not being able to visit/play (11.90%). Of these common concerns, fewer 3^rd to 8^th-grade students who had a friend with cancer were worried that their friend might die than those who did not have a friend with cancer (3^rd to 5^th graders [P = 0.0002] and 6^th to 8^th graders [P = < 0.0001]).

These data show that the majority of peers’ concerns are consistent through 3^rd to 8^th-grade students and may allow for standardizing interventions that focus on these concerns for these age groups. However, within each age group, students who had a friend with cancer were significantly less likely to be concerned about their friend dying than those who did not have a friend with cancer. This may be a result of additional education or interaction with their friend and/or their friend’s family easing this concern. These findings suggest, therefore, that when developing an intervention for children who have not previously had a friend with cancer, that prognosis should be addressed with an emphasis on the overall encouraging data that support high survivorship for most childhood cancer patients. Students in the 6^th to 8^th-grade age group with a friend with cancer also seemed to be more concerned about the inability to visit/play than in those who did not have a friend with cancer. Though this was not statistically significant, it may demonstrate that interaction with their friend and/or their friend’s family may diminish some concerns but increase others, especially within this age range where socialization and peer connection are highly valued.

Quantitative Assessment

All patient respondents (100%) had returned to school since beginning treatment at the time they completed the survey. The mean age of diagnosis for these respondents was 11.63 years old. At completion of the survey, 30.95% of respondent were 18 years old or older, 38.10% were 14–17 years old, and 30.95% were <14 years old. The majority had missed between 2–6 months (38.71%) and 6–12 months (22.58%) of school, confirming the existence of a significant amount of absenteeism in the pediatric cancer population.

The top three worries that patients experienced included (a) not being able to catch up with classwork (65.52%), (b) feeling sick or tired at school (58.62%), and (c) being treated differently (51.72%). A significant number of respondents also acknowledged other worries, including receiving insufficient or too much attention from teachers and classmates, how they looked, and their classmates not understanding their situation. Of these respondents, 10.34% acknowledged not having worries. Respondents who were 11 and younger were significantly more likely to be worried about their appearance (70.00%) when going back to school than respondents who were 12 or older (20.00%, P = 0.0124). This discrepancy may be another reason to consider including education about side effects to peers in the younger age groups. It also is an indication of the need to ensure younger patients have a support system and/or counseling in place in the event that they face bullying and/or psychosocial consequences of their anxieties in this area.

Qualitative Assessment

When asked to describe their transition back to school, the two most frequently responses by patients indicated that it was both academically (26.92%) and medically (19.23%) challenging. At the same time, 19.23% acknowledged that they had a smooth transition back to school and 15.38% acknowledged that their teachers were supportive. Other patients noted that stigma played a significant role in their transition (15.38%), transitioning overall was challenging but got better (7.69%), the logistics with the school were difficult (7.69%), absences were difficult (7.69%), and that it was overall challenging (3.85%). For example, many patients mentioned how it was medically challenging to return to school, “Some days I got to school and within an hour I’d feel sick from chemo, and it was very difficult to focus. Some days I felt like I was going to throw up, and I just wanted to leave and go home so that it wouldn’t happen in the middle of class.” These responses are verifying evidence of the many challenges this population faces when returning to school.

When asked to describe what should be included in a program, patients’ top three responses included (a) academic assistance (34.48%), (b) support groups (20.69%), and (c) education (20.69%). For example, many patients wished that they had received more academic assistance before returning to the classroom, “[Include] a way for the students to meet their teachers prior to their return to school so they can get to know one another and devise a plan for the school year” and once back in school, “Help with work missed and time to get back energy.” Some patients noted a desire for support groups, “I would like to be able to communicate with other teens my age that have gone through this. We could help each other and transition together.” These responses indicate the need for future robust school reentry programs to include both peer intervention and coordination with school and teachers to provide a safe, healthy, accommodating, and stimulating environment for these patients to continue to learn and grow. Additionally, hospitals and healthcare teams should consider incorporating support groups for patients in the same age groups to facilitate encouragement from each other as they transition back to school together.

When asked to describe what should not be included in a program, patients’ top response was not to isolate or differentiate them from others when giving assistance (33.33%). Many patients were unsure what should be excluded (47.62%). The most prominent response was an emphasis on the desire to limit the differentiation from other students while still giving them help, “[Exclude] anything that would make other students feel like they weren’t getting treated fairly compared to the child with the sickness. I never liked the attention because I wanted to be treated like every other student.” While cancer patients who return to school overwhelmingly desire assistance, they do not want to be set apart from their peers in a way that gives them noticeable special treatment or labels them as different. This was also evident, though not statistically significant, in the multiple-choice question about reintegration concerns.