Understanding the impact of early onset colorectal cancer on quality of life: a qualitative analysis of online forum data

Using online forum analysis this study highlighted a range of HRQOL issues amongst individuals with EOCRC. These impacts overlap with those reported in a recent systematic review of the HRQOL impacts of EOCRC which included 14 quantitative studies and one qualitative interview study [14, 17]. Waddell et al. [17] identified career and financial impacts, body image, intimate relationships, sexual dysfunction, emotional distress (including anxiety), social functioning and effect on personal relationships and family functioning (including caring for children) as important issues, all of which are encompassed within the themes identified in this study. Notably, sexual dysfunction and impact on intimate relationships were rarely identified in messages. This likely reflects individual’s reluctance to discuss such private issues within an open internet forum, as sexual dysfunction in adolescent and young adult (AYA) cancer survivors [37, 38] and colorectal cancer patients is widely reported [31, 39‐41]. More traditional methodologies may be required to explore these domains.

Except for stoma issues, many of the impacts identified in this study are not disease-specific and are applicable to a range of cancers in AYAs [42, 43]. Accordingly, the HRQOL impacts reported here overlap with eight of nine overarching domains identified in a systematic review of HRQOL issues facing AYAs with cancer, which included 58 quantitative studies and 11 qualitative utilising interviews [18]. The domain not reflected in our study was impact on cognitive function which is potentially an issue individuals may not volunteer without direct questioning or disclose in an open forum.

The overlap with impacts identified through systematic review both specifically for EOCRC [17] and AYA cancers [18] support our findings and highlight the utility of this novel methodology of using forum data to identify HRQOL issues. Furthermore, this study offers new insights into the HRQOL issues in EOCRC not reported by Waddell et al. [17] including the specific impacts of isolation and early menopause. We also expand on the impact of parenthood - particularly the positive benefit of parenthood and psychological impact on children – and describe the psychological impact of potential infertility and practical implications of fertility treatments.

In addition to describing the practical impacts of cancer on child-caring abilities, we identified the psychological impacts on children and demand for tailored children’s support. A mixed methods study of unmet needs in EOCRC also reported individuals unable to find such services [31]. Emotional and behavioural problems in children of parents with cancer and the positive impact of adaptive coping strategies have also been highlighted, supporting a drive for dedicated children’s services [44]. Providing intensive interventions for all families is not practically possible but our findings show some already utilise resources provided by charitable organisations and advocate the use of these to peers within forum messages. Understanding existing resources and signposting appropriately may be an impactful first step for clinicians in providing support. Further research is needed to investigate how to identify families most likely to benefit from more intensive interventions.

Themes of isolation and difficulties relating to family, friends and other cancer patients by virtue of diagnosis at a young age, are echoed in other studies [29, 31]. Whilst it is not surprising that the positive impact of connecting with other EOCRC patients is highlighted in a study analysing data from a forum specifically designed for this purpose, the value of interactions with other EOCRC survivors as a source of support is emphasised in studies using more traditional qualitative methodologies [15]. Sodergren et al. [18] describe isolation amongst AYA patients and highlight a study of patients with acute lymphoblastic leukaemia (ALL) reporting the benefit of opportunities to make new connections with fellow patients [45]. Compared to ALL which is usually diagnosed in younger people [46], the typical colorectal cancer patient is still over 50, despite the rising incidence of early onset disease [2]. Therefore, the opportunities for EOCRC patients to makes links with others going through the same situation may be more limited, and the need for access to dedicated support groups to forge these connections may be especially pertinent amongst EOCRC patients.

The impact of infertility after treatment for EOCRC was frequently discussed. Whilst risk of gonadal failure is reportedly low following adjuvant chemotherapy for colon cancer [47, 48], gonadal failure post pelvic irradiation for rectal cancer is an expected side effect [47, 49, 50]. Fertility preservation guidelines recommend sperm cryopreservation for males, and embryo or oocyte cryopreservation and/or ovarian transposition pre-radiotherapy in females, with a caveat of variable success rates in the latter [50, 51]. Discussions about fertility and referrals to reproductive specialists are recommended to occur as early as possible [51]. Users described waits for fertility treatments leading to cancer treatment delays, and the additional stress of decision-making surrounding this. Other studies report a lack healthcare provider recommendations for post-treatment fertility issues [15]. This may reflect a knowledge gap and lack of established fertility preservation referral pathways in colorectal cancer services initially designed to treat older patients. Recommendations state all practitioners involved in treating young cancer patients should be able to address these issues [51]. Improvements likely require education of healthcare professional to assist in complex decision making and facilitate fertility discussions early in the cancer journey.

Concerning post-treatment menopause, which follows pelvic irradiation [52], some users reported poor information giving. A lack of regard for addressing menopausal symptoms and consequences, such as osteoporosis and cardiovascular disease, due to prioritisation of cancer surveillance is also reported in the literature and the importance of multidisciplinary team management with a vital role for primary care providers is emphasised [52]. Ensuring good communication between specialist cancer services and primary care may be the most important step in ensuring adequate follow-up for menopause issues.

Regarding the diagnostic pathway, a considerable number of emergency diagnoses was not an unexpected finding. An observational study of EOCRC diagnoses in England found younger patients were more commonly diagnosed as an emergency and less likely to be diagnosed via two-week wait than older counter parts (p < 0.001) [13]. The perception of barriers to diagnosis in both primary and secondary care is reflected in existing literature and age bias affecting clinical assessment is a common theme [15, 25, 29, 53]. Guilt associated with overburdening strained secondary care services and age-related guideline constraints have been highlighted as barriers for general practitioner’s in referring suspected EOCRC [53]. In August 2023, UK referral guidance was updated to include wider use of faecal immunochemical testing, offering an easily accessible quantitative measure for risk stratifying patients in primary care to enable better prioritisation of secondary care investigations, which may assist in overcoming these barriers [5, 54, 55]. The impact of these guideline changes and campaigns to increase awareness of EOCRC [12] is an important future research question.

The key strength of this work is the use of forum data to reveal key HRQOL impacts of EOCRC. Using this time and cost-efficient methodology, this study captured the issues described through systematic review of existing literature [17] and identified additional impacts. It also provided additional qualitative content which is comparatively lacking within this literature. Analysis of online forum data offers unique advantages over traditional qualitative methodologies. All content is user-generated with no interaction between researcher and participant, eliminating the potential influence of interviewer bias or restrictions of an interview structure and enabling emergence of unexpected themes [23]. This is supported by the emergence of multiple topics of interest through repeat inductive extraction. Users described difficulty talking openly and relating to others given their relatively uncommon diagnosis but stressed the value of the forum in facilitating honest discussion. Forum analysis likely enabled exposure of important experiences such as feelings of isolation which may not have been discussed in interviews. The efficiency of this methodology in terms of time and cost savings is a major strength. Use of forum data allowed a vast quantity of easy to access data, from many individuals, available from the public domain to be retrieved and analysed at reduced cost, with no need for participant recruitment, lengthy ethics delays or conduction and transcription of interviews as required in other studies [14, 15, 23, 31]. This enabled the research to be conducted over a few months, rather than a longer time period usually required for qualitative methodologies, whilst still maintaining high quality data outcomes and encapsulating a significantly greater number of views than is usually possible with qualitative work.

The inclusion of a co-researcher with lived experience in analysis of the results offered an invaluable perspective when interpreting forum message meaning. They highlighted the frequency with which messages offer peer-to-peer advice and advocated for existing services, which raises future research questions surrounding better utilisation of existing services along with designing of new interventions to tackle the HRQOL impacts in EOCRC. The concept of analysis of forum data to explore patient-to-patient advice giving statements has been explored in a recent study and presents further utility of this methodology [26].

This methodology has limitations. The study aims to analyse experiences of a specific group but cannot control characteristics of individuals posting within the forum. Where messages offered sufficient information, strict exclusion criteria were applied, and the forum board clearly stated it was a space for those with EOCRC. It is possible some messages included in analysis do not represent individuals with EOCRC, but this likely represents a very small proportion of analysed data. Potential sample bias with over-representation of certain genders, ethnicities or socio-economic groups must also be acknowledged. Project resource constraints precluded access to alternative analysis software or analysis of multiple forums, the forum used is registered to one country limiting cross-cultural validity. To facilitate analysis of sufficient data regarding HRQOL within the study timeframe, messages including symptom-only data were excluded however comparison of treatment and disease-related physical side-effects in early versus late onset disease presents an interesting separate research question as more aggressive treatment in younger patients is reported [9, 56]. Issues relating to sexual dysfunction, widely reported in existing literature [31, 37‐41], were not revealed through forum analysis which may be an inappropriate methodology for exploring this sensitive issue. The ethical requirement to protect user anonymity requires data anonymisation and paraphrasing of published quotes, however all paraphrased quotes were reviewed by co-researchers ensuring meaning was retained. Data anonymisation precludes user tracking, meaning data stratification by demographics is not possible. One user describing the same experience in multiple messages represents multiple separate data points, as thematic rather than content analysis was applied this has less influence but should be considered in interpretation of the results.