Awareness of diagnosis, treatment and risk of late effects in Chinese survivors of childhood cancer in Hong Kong

Abstract Background For survivors of childhood cancer, awareness of personal health risks is a critical component of long‐term health management. Objective To evaluate the awareness of the diagnosis, treatment and risk of late effects among survivors of childhood cancer in Hong Kong. Methods Between June 2019 and March 2020, this cross‐sectional study recruited 155 adult survivors (mean age = 26.9, standard deviation [SD] = 6.4 years) and 45 parents of paediatric survivors (mean age = 11.1, SD = 3.6 years) from a long‐term follow‐up clinic. At >10 years post‐treatment (mean = 13.4, SD = 7.6 years), they completed a structured questionnaire to report their cancer‐specific knowledge. Multiple linear regression analysis was conducted to identify clinical, socioeconomic and behavioural factors associated with poor awareness. Results The majority of participants accurately recalled their diagnoses (73.5%) and major treatment modalities (chemotherapy 92.4%, radiation 82.9% and surgery 88.2%). However, less than half (45%) of the participants recognized more than 25% of the total late effects for which they were at risk. The highest levels of awareness were reported for endocrine problems (49%), neurocognitive impairment (44%) and secondary cancers (43%), and the lowest for peripheral neuropathy (21%) and vision problems (23%). Compared with survivors of haematological malignancies, those of central nervous system (CNS) tumours (standardized estimate [B] = −9.33, 95% confidence interval [95% CI]: −13.41 to −5.26) and non‐CNS solid tumours (B = −8.47, 95% CI: −12.39 to −4.94) had less knowledge about their diagnosis. Retaining medical records (P < .0001) and better medical information‐seeking habits (P = .048) were associated with better awareness. Conclusions Survivors of childhood cancer in Hong Kong have deficient awareness of their personal health risks. They may benefit from the provision of a survivorship care plan and personalized education regarding treatment‐related late effects. Patient Contribution Patients contributed in designing the study tools. Results were presented at a non‐governmental organization.


| INTRODUC TI ON
One goal of quality survivorship care is to improve the health literacy of survivors by helping them understand the nature, screening, prevention and management of possible late effects. 6 The World Health Organization defines health literacy as 'the cognitive and social skills which determine the motivation and ability of individuals to gain access to, understand and use information in ways which promote and maintain good health'. 7 Studies comparing general and cancer populations have revealed associations of a low level of health literacy with increased medical costs, poor medication adherence, the misinterpretation of health information, a lower level of physical activity and a higher risk of developing health-damaging habits, such as smoking. 8,9 Conversely, increased awareness of the symptoms of potential late effects in cancer survivors can prompt preventive action and timely management. 10 For instance, a Chinese study observed that more health-literate survivors of cancer tend to have better survivorship outcomes, including cognitive and social functioning, mental well-being and financial ability. 11 Empowering the health literacy and cancer-specific knowledge of survivors is therefore beneficial in terms of improving their quality of life.
Several studies in Western countries have revealed that a younger age at diagnosis, a lower socioeconomic status (SES) and the specific cancer type are the major factors influencing health literacy among survivors and caregivers. 12,13 However, these risk factors may not be generalizable to the Asian population due to differences in cultural, lifestyle and health practices. 14  The primary objective of this study was to examine the extent of knowledge regarding the diagnosis, treatment and risk of late effects in survivors of childhood cancer in Hong Kong and to identify the predictors of cancer-specific health literacy in this population. Hospital in Hong Kong. This regional tertiary care public hospital serves as a major hub to provide LTFU care to survivors of childhood cancer. Written informed consent was obtained from all participants.

| Subjects
Survivors were eligible for inclusion if they were diagnosed with cancer by a paediatric oncologist/haematologist before 18 years of age, received cancer treatment at the study institution and had been in remission for ≥5 years post-diagnosis. Patients who had a cancer relapse or secondary malignancy were also

| Data collection method
Adult survivors and parents of paediatric survivors (referred as 'participants' hereafter) completed a structured questionnaire to evaluate their cancer-specific knowledge and medical informationseeking habits (Appendix S1). To assess the accuracy of the participants' reports, their responses were verified against information abstracted from the Clinical Management System (CMS), an electronic health data repository of the public health-care system in Hong Kong that is considered a reliable data source for epidemiological research in Hong Kong. 19 The CMS includes cancer-related variables (eg diagnoses and age at diagnosis) and treatment-related variables (eg treatment start and end dates, treatment protocols, chemotherapy drugs, surgical sites and radiation sites).
The responses were then graded according to the scoring system presented in Appendix S2. Three subscores (diagnosis awareness score, treatment awareness score and late effects awareness score) were computed (range: 0 to 100), with higher scores indicative of better knowledge. professionals regarding the quality of the health information. Each item was rated on a 4-point Likert scale from 1 (strongly disagree) to 4 (strongly agree), and the item scores were summed to yield a total score ranging from 1 to 20 points. A higher score indicates a better ability to acquire and interpret health-related information.

| Statistical analysis
Descriptive statistics are used to summarize the demographic and clinical characteristics of the participants. First, a univariate analysis was conducted to identify the factors associated with the following primary outcomes: (1) diagnosis awareness score, (2) treatment awareness score and (3) late effects awareness score. The Mann-Whitney U test was used to evaluate independent variables that are categorical in nature (eg education level, cancer diagnosis, presence of chronic health condition). The Spearman correlation test was applied to continuous variables (eg age at diagnosis, time since treatment and HLQ score). Factors that were significantly associated (P < .10) with any of the three subscales in the univariate analyses were included in subsequent multivariable analyses.
Multiple regression models were used to identify factors associated with the primary outcomes after adjustment for the covariates of sex and age at evaluation. These associations are presented using standardized estimates (B) and 95% confidence intervals (CIs).
The significance threshold was set at P < .05. As the responses for paediatric patients (<16 years of age) were proxy-reported by their parents, a sensitivity analysis was conducted by repeating the multivariable analysis in adult survivors only (ie on self-reported data only).
All statistical analyses were performed using SAS (SAS 9.4, SAS Institute, Cary, NC, USA) and were two-tailed.  A slight majority of survivors reported that they kept cancerrelated medical records (n = 106, 53%). The mean (SD) medical information-seeking ability score was 13.8 (2.7) points (range: 5 to 20 points).

| Awareness of primary cancer diagnosis
The most common type of cancer diagnosis was haematological malignancy (n = 106, 53.0%), followed by non-CNS tumour (n = 80, 40%) and CNS tumour (n = 14, 7%) ( Around half of the participants who were diagnosed with haematological malignancies could accurately and precisely report their diagnoses (n = 49, 46.2%) (Figure 1), whereas smaller proportions of participants with CNS tumour (n = 4, 28.5%) and non-CNS tumour malignancies (n = 15, 18.8%) were able to do so. The majority of participants in the latter groups were unable to name the tumour sites and subtypes accurately.

| Awareness of treatment exposure
The vast majority of survivors were treated with chemotherapy (n = 185, 92.5%) (

| Awareness of potential late effects
Based on the treatment exposures, the study cohort was at risk for a median of 6 (interquartile range: 4 to 7) of the 12 late effects Only 9% (n = 18) of the participants could identify more than three-quarters of the late effects for which they were at risk ( Figure 3). More than half (n = 110, 55%) could only recognize less than a quarter of the potential late effect risks, and nearly a quarter (n = 44, 22%) could not identify any of their health risks (Figure 3).

| Factors associated with cancer-related health literacy
The   For these survivors, parents may often be the primary sources of in-

| Limitations
This study is subject to the following limitations. First, only survivors who attended the clinic's LTFU consultations were interviewed. It is reasonable to assume that our study participants might be more health conscious than survivors who failed to visit the LTFU clinic. though a post hoc analysis revealed that higher household income and education attainment were associated with better (non-cancerspecific) medical information-seeking abilities (Appendix S9). The few indicators of SES adopted in our study might not have been sensitive enough to detect differences in the awareness outcomes.
We did not conduct a comprehensive measure of SES in the context of Hong Kong; such an investigation should also include information about the housing tenure, occupation, monthly expenditures and economic hardship. 42 The actual rates of chronic health conditions may be underestimated due to the high rates of undiagnosed disease and inconsistent screening of health conditions in survivors.
However, we tried to capture chronic health conditions comprehensively by including both clinical records and self-reported data.
Admittedly, the latter approach might be subjected to inaccurate reporting but it is still recognized to be effective in facilitating collection of health outcomes data in epidemiological studies involving paediatric cancer population. 43,44 Despite these limitations, the current study provides evidence to support the development of targeted interventions and facilitate larger-scale research agendas intended to improve the health and psychosocial outcomes of survivors of childhood cancer in Hong Kong.

| CON CLUS ION
The results of our study reveal a deficiency in cancer-related knowledge among long-term survivors of childhood cancer in Hong Kong, especially those diagnosed with solid tumours and/or at a younger age. The awareness of treatment-related late effects was generally poor in this population. Our findings highlight the need to facilitate communication with survivors about the potential treatment-related late effects and develop tailored educational interventions and behavioural interventions to improve their health statuses. As the majority of survivors are still adolescents and young adults in the early period of survivorship, helping these individuals to advocate for themselves could potentially improve their health and psychosocial outcomes and will eventually translate to a substantial positive impact on society.

ACK N OWLED G EM ENT
We would like to acknowledge Professor Smita Bhatia and Professor

CO N FLI C T S O F I NTE R E S T
All authors have disclosed no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.