Abstract P3-12-12: ATTITUDE: Understanding and reducing ATTrition in longiTUDinal studiEs of cancer survivors

Abstract

Abstract Background: Reduction in response rates is common in longitudinal studies, generating significant attrition bias. ATTITUDE is a comprehensive project built on the CANcer TOxicity (CANTO) cohort, using a mixed-methodology, multi-step approach aimed at reducing attrition: Step 1- understand patterns of attrition; Step 2- understand patients’ experiences in providing data; and Step 3- design and pilot a tailored intervention to reduce attrition. Here we present the results of Step 1. Methods: CANTO (NCT01993498) is a longitudinal cohort study that includes patients with stage I-III breast cancer from 26 cancer centers. CANTO collects detailed clinical and patient-reported outcome (PRO) evaluations at diagnosis of breast cancer, and at year-1, year-2 and year-4 post-diagnosis. Concurrent enrollment in other investigational studies is allowed. Our primary outcome of interest was “non-response”, defined as not responding to pre-planned CANTO evaluations, prematurely exiting the study, withdrawing consent or being lost to follow-up. Patients with a valid reason for study termination, including fatal or recurrence events, were excluded. Multivariable logistic regression with bootstrapped Augmented Backwards Elimination retained factors associated with non-response. Results: Among 5903 patients enrolled in CANTO from 2012-2015 with mature follow-up at year-4 post-diagnosis, mean age was 56.7 years (SD 11.4), 51.5% had stage I breast cancer, and 27.2% were concomitantly enrolled in other investigational studies. The rate of non-response to clinical and PRO evaluations was 0% and 5.7% at diagnosis, 12.3% and 21.2% at year-2, and 20.1% and 43.1% at year-4, respectively. Factors associated with non-response at year-4 included study entry period, CANTO center size, smoke, alcohol consumption behavior, and symptom burden at inclusion. Patients included in larger patient volume centers, with higher socio-economic status at diagnosis and those that co-enrolled in other studies had reduced odds of non-response (Table). Conclusions: In this large cohort of patients with early-stage breast cancer, attrition at year-4 post-diagnosis was substantial, particularly for PRO evaluations, and it was affected by center- and patient-related factors. Results of the Step 1 of ATTITUDE will help us identify patients with higher risk of non-response and inform targeted interventions to minimize attrition, including better risk stratification and personalization of the follow-up during participation in longitudinal research studies. These interventions will be developed and tested during subsequent steps of ATTITUDE. Table. Factors associated with non-response at year-4 by type of evaluation; Adjusted* Odds Ratio (95% CI)Clinical evaluationsPRO evaluationsStudy entry period2012-2013Ref.Ref.2014-20150.74 (0.62 - 0.87)1.77 (1.57 - 2.02)Center size (patients enrolled in CANTO)Small (<100)Ref.Ref.Medium (100-1000)0.54 (0.41 - 0.73)0.27 (0.21 - 0.35)Large (≥1000)0.44 (0.32 - 0.61)0.33 (0.25 - 0.44)Age at diagnosis5-year increase1.08 (1.03 - 1.13)0.98 (0.94 - 1.01)EducationPrimary schoolRef.Not retainedHigh school0.69 (0.55 - 0.88)College or higher degree0.65 (0.49 - 0.86)Monthly household income (Euro)<1500Ref.Ref.≥1500 and <30000.89 (0.71 - 1.13)0.85 (0.71 - 1.03)≥30000.75 (0.58 - 0.97)0.73 (0.60 - 0.88)Tobacco behaviorNever smokerRef.Ref.Current smoker1.72 (1.37 - 2.14)1.39 (1.16 - 1.66)Alcohol consumption frequencyLess than dailyRef.Ref.Daily1.21 (0.96 - 1.52)1.23 (1.02 - 1.47)Fatigue**10-point increase1.05 (1.02 - 1.08)1.04 (1.01 - 1.08)Pain**10-point increaseNot retained1.02 (0.99 - 1.08)Co-enrollment in other studies,NoRef.Ref.Yes0.67 (0.55 - 0.81)0.79 (0.68 - 0.91)*by marital status, physical activity, comorbidities, psychological distress (Anxiety and Depression; HADS), breast cancer stage, and treatment-related factors**EORTC QLQ-C30, assessed at study entry (diagnosis) Citation Format: Pietro Lapidari, Julie Havas, Daniele Presti, Davide Soldato, Elise Martin, Barbara Pistilli, Aurelie Bertaut, Diane Boinon, Guillemette Jacob, Anne-Laure Martin, Sibille Everhard, Alexandra Jacquet, Gwenn Menvielle, Matteo Lambertini, Laurence Vanlemmens, Stefan Michiels, Fabrice André, Ines Vaz-Luis, Antonio Di Meglio. ATTITUDE: Understanding and reducing ATTrition in longiTUDinal studiEs of cancer survivors [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P3-12-12.