Survey population.

The survey population will be survivors of pubertal, adolescent, and young adult cancer diagnosed between 2007 and 2018. Participants eligible for the survey must comply with all the following at selection from the registry: current age ≥ 18 years, age at cancer diagnosis (0–25 years), alive at the time of last contact, and with complete registry information, including a traceable phone contact. We included individuals 0–25 to encompass childhood and young adult cancer survivors, using a young adult definition of 18–25 years. Although there is significant variation between societies for appropriate age definition of young adulthood, the Society for Adolescent Health and Medicine defines young adulthood as a period ranging from 18–25 years of age [14]. We assume, among others [3,14,15], that this transition represents youth [16] and is associated with several unique physiological, social, and psychological phenomena that render them vulnerable to receiving inadequate health care. In addition, for many within this age range, their reproductive expectations are not concrete, and their interest in reproductive survivorship is high.

To estimate the sample size for the cross-sectional survey, we used the results of the Childhood Cancer Survivor Study (CCSS) cohort [17]. In this study, the proportion of survivors that had been pregnant or fathered a pregnancy was 38% compared to 62% of sibling controls. Therefore, using the ability to have or father a pregnancy as a crude indicator of reproductive function, we applied 38% as a prevalence of normal reproductive function among childhood cancer survivors in the statistical formula described by J Charan and T Biswas [18] and determined the minimum sample size for the survey to be 362 childhood and young adult cancer survivors.

Survey instrument.

An extensive literature review of the MEDLINE database was performed to find relevant structured questionnaires that could answer the study objectives. Priority was given to questionnaires administered by an interviewer and used the CATI method. The Furthering Understanding of Cancer, Health, and Survivorship in Adult (FUCHSIA) Women’s Study [19] questionnaire was identified as the most suitable to answer the research question. The Understanding Reproductive Health among Survivors of pediatric and young adult (URHSPY) CANCER questionnaire was developed by the study team from the FUCHSIA questionnaire. This was achieved by modification, deletion, and addition of questions that are context-sensitive and addressing the current study objectives. In addition, a male version was designed and made suitable to understand reproductive health among male cancer survivors. The study team reviewed the draft URHSPY questionnaire to ensure conciseness, relevance, and completeness. The URHSPY cancers CATI will address these areas: (1) cancer treatment history, (2) demographics, (3) menstrual health, (4) infertility history, (5) pregnancy history, and (6) lifestyle. The final version of the URHSPY questionnaire was programmed into REDCap data capture software [20].

Further pre-testing of complex skip patterns, typing errors, non-offensiveness, question flow, and completeness were done by an expert on the study team. Two research assistants were trained on how to conduct telephone-assisted interviews. A field pre-test of the CATI-based survey was done on a sample of 10 cancer survivors accessed conveniently with a similar eligibility profile as the intended survey subjects. After the pre-test, the questionnaire was modified before its final rollout based on any feedback concerning the appropriateness, question interpretation, question intrusiveness, and interview time.

Survey data collection, management, and analysis.

All eligible subjects were contacted to participate in the survey. In addition, a text message containing summarised information about the ongoing survey were sent to the registered KCR contacts of all potential subjects before the first call by the study team.Participants were contacted by telephone using the phone contact provided in the registry. If contact was successful, the interviewer introduced themselves, the survey, and its relevance. The interviewer ascertained the identity of the contact and confirmed their vital status. All efforts have been made to agree on the most convenient time and day of the week for the interview. The interviewer contacted the participants on the agreed time and day, which may or may not be the same day as the first contact. Upon making contact, the interviewer re-introduced themselves and provide more information about the survey. The interviewer then administered and guided the participant through the computer-assisted verbal consent process and the survey. The interviewer read the questions directly from the interactive computer screen on the survey instrument and recorded the verbal answers of the respondent. The research assistants were trained to observe emotional cues during the telephone survey.

In addition, the participants have the option to decline to answer questions that are sensitive or halt the interview. In case of emotional distress, the participant is referred to the study counselor for further care. To define a failed contact attempt, the investigators agreed that ten call attempts made over five working days with at least two calls per day (morning and afternoon) were exhaustive considering available resources and would be the desired criterion All participants get a digital or USSD airtime gift card worth UGX 20,000, to compensate for their time at the end of the interview. In addition, the principal investigator does a data check of 10% of the interviews performed off-site via online REDCap project administrator access. This is a data quality check, after which a weekly data quality report is made and uploaded onto a secure, password-protected research purpose-only computer. When reporting this protocol, 50% of the intended participants have been contacted.

All REDCap survey data will be exported into up-to-date STATA 17, StataCorp. 2021. Stata Statistical Software: Release 17. College Station, TX: StataCorp LLC. Descriptive statistics will be performed on the data to estimate various proportions and means of survey responses. Further analysis to draw inferences from the responses in each survey section will be done using the chi-squared test for categorical responses and the T-test for continuous responses. We will use regression methods with appropriate interaction terms for subgroup analyses between men and women. We will calculate Relative Risk (RR) and RR Reductions (RRR) with corresponding 95% confidence intervals to compare dichotomous responses. The difference in means will be used for additional analysis of continuous responses. For all tests, we will use 2-sided p-values with alpha ≤0.05 level of significance.

In-depth interview population.

Two categories of participants will be interviewed: (1) Guardians older than 18 years and attending to children (0–18 yrs) diagnosed with cancer and who are out of the induction and consolidation phases of treatment. (2) the oncology care team includes oncology physicians, medical officers, and nurses.

The sample size will be determined by theoretical saturation [23] i.e., when no new meaning can be obtained from additional interviews. Based on GT methodology, we shall use theoretical sampling on a varied purposeful sample of participants. Deviating from the tenets of GT [23], we estimated sample size priori using the assumption by MM Hennink, et al. [25] that it would take 16–24 interviews for inducted or deducted codes to reach meaning saturation. We plan to enrol 20 participants as guardians of children. Comprehensive data analysis will inform the composition of the subsequent sample. If theoretical saturation is not reached, additional interviews will be performed until at least three further interviews do not generate new meaningful codes. We shall interview all ten health workers attached to the pediatric oncology department. This sample includes oncology fellows, oncologists, oncology nurses, and medical officers.

Data collection.

An in-depth interview is the best way to add depth and context to the population-level data. Therefore, separate guides have been developed for guardians and health workers. The interviews will be conducted with an interview guide (S1 and S2 Files) with inherent flexibility. The guide is systematic and semi-structured, with open-ended questions, and not limiting the participant’s expression of their experience. The guide will be developed with significant input from the quantitative survey results [11]. In addition, the interviewers will follow probes during the dialogue with participants to encourage the articulation of experiences, perceptions, and attitudes. The interview guides were pre-tested on a group of five eligible participants to address question clarity, acceptability, and understanding.

After enrolment, face-to-face interviews will be held in a comfortable and private setting. All participants must consent to the study before the interview can start. At all times, an effort will be made to maintain comfort, and alleviate stress and privacy during the interview process, as described by R Elmir, et al. [26]. The interviewers are research nurses certified in good clinical practice and basic human research subject protection, with training in conducting in-depth interviews for qualitative research. The interview will be recorded with a Sony® IC recorder (model: ICD-PX470). Field notes will be taken during the interview to document observed nonverbal behaviours. All participants will get compensation worth UGX 40,000 to compensate for their time in the study. For cases where there is recognizable emotional fallout from the interview, the interview will be halted and postponed. The participant will then be connected to a study counselor to follow up with the debrief.

Data management and analysis.

All the recordings will be transcribed verbatim and translated into English in Microsoft word® version 16.41. Each transcription will be made within 24 hours of ending the interview. A professional research transcription service will do this. The resulting transcripts will be analyzed in qualitative data analysis software NVivo version 12 by QRS international. Using theoretical sensitivity and inductively coding for whichever meaning arises from the data [23], each transcript will be coded thrice by a team of three investigators, using the participant’s own words, as GT requires [23]. A synthesized method of GT will be utilized in a multi-step systematic manner, as described by [24]. In sequential order starting with line-by-line in vivo coding using the participant’s own words, shorter word phrases will then be developed that capture the meaning of the participant’s words while grouping similar coded phrases; concepts are named; concepts are grouped and refined into categories; theoretical coding families will be used to isolate concepts and connect them in different ways; new connections of categories will be established; subcategories will be defined; linkage among categories will be developed; the core category defined; the core category leading to the discovery of the basic social process identified; and lastly explanatory frameworks leading to the derivation of a substantive theory developed.

To reach an inter-coder consensus, we shall use methods described by MA Cascio, et al. [27]. In addition, to ensure rigor in the study process, we will have reflexivity among authors, critical feedback among authors, peer examination, peer briefing, using participant’s own words, consistency in the interview guide across interviews, and an audit trail including data collection process, coding, consensus guidelines, and memos throughout the research process [28].