Missingness mechanisms and generalizability of patient reported outcome measures in colorectal cancer survivors - assessing the reasonableness of the "missing completely at random" assumption.

BACKGROUND: Patient-Reported Outcome Measures (PROM) provide important information, however, missing PROM data threaten the interpretability and generalizability of findings by introducing potential bias. This study aims to provide insight into missingness mechanisms and inform future researchers on generalizability and possible methodological solutions to overcome missing PROM data problems during data collection and statistical analyses. METHODS: We identified 10,236 colorectal cancer survivors (CRCs) above 18y, diagnosed between 2014 and 2018 through the Danish Clinical Registries. We invited a random 20% (2,097) to participate in a national survey in May 2023. We distributed reminder e-mails at day 10 and day 20, and compared Initial Responders (response day 0-9), Subsequent Responders (response day 10-28) and Non-responders (no response after 28 days) in demographic and cancer-related characteristics and PROM-scores using linear regression. RESULTS: Of the 2,097 CRCs, 1,188 responded (57%). Of these, 142 (7%) were excluded leaving 1,955 eligible CRCs. 628 (32%) were categorized as initial responders, 418 (21%) as subsequent responders, and 909 (47%) as non-responders. Differences in demographic and cancer-related characteristics between the three groups were minor and PROM-scores only marginally differed between initial and subsequent responders. CONCLUSION: In this study of long-term colorectal cancer survivors, we showed that initial responders, subsequent responders, and non-responders exhibit comparable demographic and cancer-related characteristics. Among respondents, Patient-Reported Outcome Measures were also similar, indicating generalizability. Assuming Patient-Reported Outcome Measures of subsequent responders represent answers by the non-responders (would they be available), it may be reasonable to judge the missingness mechanism as Missing Completely At Random.

Guided versus non-guided digital psychological interventions for cancer patients: A systematic review and meta-analysis of engagement and efficacy.

OBJECTIVE: To evaluate engagement with and efficacy of guided versus non-guided digital interventions targeting psychological symptoms of cancer via a systematic review of current evidence. METHODS: PubMed, Scopus, PsychINFO, MEDLINE, and CINAHL databases were searched. Eligible publications were randomised controlled trials of guided or non-guided digital psychological interventions used in cancer settings reporting intervention efficacy and/or engagement. Study methodological quality was assessed using the Cochrane risk of bias (RoB) tool. Random effects meta-analyses were performed on outcomes with sufficient data, with sub-group analyses of intervention type and follow-up period. RESULTS: Forty-three studies were included. Studies varied by level of guidance, type of technology used, duration, and outcomes assessed. Most studies had a high overall RoB. Meta-analysis indicated that guided interventions significantly reduced distress, anxiety, and fatigue, while non-guided did not. For depression and quality of life, both guided and non-guided interventions produced significant improvements. Guided interventions reported higher levels of intervention engagement than non-guided. CONCLUSIONS: Guided digital psychological interventions were likely to be more effective than non-guided ones for cancer patients, particularly in reducing distress, anxiety, and fatigue. Whilst both types were found to improve depression and life quality, guided interventions were associated with higher patient engagement. These findings suggest digital interventions could supplement traditional cancer care, warranting further research concerning long-term effects and cost-efficiency.

Validity and screening capacity of the FCR-1r for fear of cancer recurrence in long-term colorectal cancer survivors.

PURPOSE: Existing fear of cancer recurrence (FCR) screening measures is being shortened to facilitate clinical use. This study aimed to evaluate the validity and screening capacity of a single-item FCR screening measure (FCR-1r) in long-term colorectal cancer (CRC) survivors with no recurrence and assess whether it performs as well in older as in younger survivors. METHODS: All Danish CRC survivors above 18, diagnosed and treated with curative intent between 2014 and 2018, were located through a national patient registry. A questionnaire including the FCR-1r, which measures FCR on a 0-10 visual analog scale, alongside the validated Fear of Cancer Recurrence Inventory Short Form (FCRI-SF) as a reference standard was distributed between November 2021 and May 2023. Screening capacity and cut-offs were evaluated with a receiver-operating characteristic analysis (ROC) in older (≥ 65 years) compared to younger (< 65 years) CRC survivors. Hypotheses regarding associations with other psychological variables were tested as indicators of convergent and divergent validity. RESULTS: Of the CRC survivors, 2,128/4,483 (47.5%) responded; 1,654 (36.9%) questionnaires were eligible for analyses (median age 76 (range 38-98), 47% female). Of the responders, 85.2% were aged ≥ 65. Ninety-two participants (5.6%) reported FCRI-SF scores ≥ 22 indicating clinically significant FCR. A FCR-1r cut-off ≥ 5/10 had 93.5% sensitivity and 80.4% specificity for detecting clinically significant FCR (AUC = 0.93, 95% CI 0.91-0.94) in the overall sample. The discrimination ability was significantly better in older (AUC = 0.93, 95% CI 0.91-0.95) compared to younger (0.87, 95% (0.82-0.92), p = 0.04) CRC survivors. The FCR-1r demonstrated concurrent validity against the FCRI-SF (r = 0.71, p < 0.0001) and convergent validity against the short-versions of the Symptom Checklist-90-R subscales for anxiety (r = 0.38, p < 0.0001), depression (r = 0.27, p < 0.0001), and emotional distress (r = 0.37, p < 0.0001). The FCR-1r correlated weakly with employment status (r = - 0.09, p < 0.0001) and not with marital status (r = 0.01, p = 0.66) indicating divergent validity. CONCLUSIONS: The FCR-1r is a valid tool for FCR screening in CRC survivors with excellent ability to discriminate between clinical and non-clinical FCR, particularly in older CRC survivors.

Study protocol: a randomized controlled trial comparing the efficacy of therapist guided internet-delivered cognitive therapy (TG-iConquerFear) with augmented treatment as usual in reducing fear of cancer recurrence in Danish colorectal cancer survivors.

BACKGROUND: Cognitive therapy has been shown to reduce fear of cancer recurrence (FCR), mainly in breast cancer survivors. The accessibility of cognitive behavioural interventions could be further improved by Internet delivery, but self-guided interventions have shown limited efficacy. The aim of this study is to test the efficacy of a therapist guided internet-delivered intervention (TG-iConquerFear) vs. augmented treatment as usual (aTAU) in Danish colorectal cancer survivors. METHODS/DESIGN: A population-based randomized controlled trial (RCT) comparing TG-iConquerFear with aTAU (1:1) in n = 246 colorectal cancer survivors who suffer from clinically significant FCR (Fear of Cancer Recurrence Inventory Short Form (FCRI-SF) ≥ 22 and semi-structured interview). Evaluation will be conducted at 2 weeks, 3 and 6 months post-treatment and between-group differences will be evaluated. Long-term effects will be evaluated after one year. Primary outcome will be post-treatment FCR (FCRI-SF). Secondary outcomes are global overall health and global quality of life (Visual Analogue Scales 0-100), bodily distress syndrome (BDS checklist), health anxiety (Whiteley-6), anxiety (SCL4-anx), depression (SCL6-dep) and sickness absence and health expenditure (register data). Explanatory outcomes include: Uncertainty in illness (Mishels uncertainty of illness scale, short form, MUIS), metacognitions (MCQ-30 negative beliefs about worry subscale), and perceived risk of cancer recurrence (Visual analogue Scale 1-100). DISCUSSION: This RCT will provide valuable information on the clinical and cost-effectiveness of TG-iConquerFear vs. aTAU for CRC survivors with clinical FCR, as well as explanatory variables that may act as outcome moderators or mediators. TRIAL REGISTRATION: ClinicalTrials.gov; NCT04287218, registered 25.02.2020. https://clinicaltrials.gov/ct2/results?cond=&term=NCT04287218&cntry=&state=&city=&dist=.