Lymphedema and Patient Reported Outcomes - A Propensity Score Matched Analysis
Leslie Kim, MD1, Michelle Coriddi, MD1, Leslie McGrath, NP1, Meghana Shamsunder, MPH1, Kathryn Haglich, MS1, Jacqueline Chu, BA1, Babak Mehrara, MD1, Joseph H Dayan, MD1, Jonas A Nelson, MD, MPH1
1Plastic and Reconstructive Surgery Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, New York, NY
Background: Lymphedema is known to have a negative effect on patient-reported outcomes measures (PROMs) such as BREAST-Q, but it is unclear which domains are affected and to what degree. We conducted a propensity matched analysis comparing BREAST-Q scores of patients with lymphedema to those of patients who did not to clarify the impact of a lymphedema diagnosis on BREAST-Q scores.
Methods: Propensity score matching analysis (1:1 matching, no replacement) was performed for patients who received either autologous or implant reconstruction with diagnosed lymphedema (ICD-9 or 10 codes) compared to those without. Matched covariates included age, BMI, race, smoking history, any radiation, any chemotherapy, postoperative infection, and reconstruction modality and laterality. Outcomes of interest included BREAST-Q scores for satisfaction with breast and physical well-being (WB) of chest, sexual-WB, and psychosocial-WB domains. The minimum clinically important difference (MCID) for BREAST-Q scores is 4. A p-value of 0.05 was considered significant Results: The overall cohort included 3,268 patients, with a lymphedema incidence noted to be 9.9%. The matched cohorts included 322 patients in each group. Lymphedema versus non-lymphedema patients did not differ significantly with respect to age, BMI, race, smoking status, any radiation or chemotherapy exposure, postoperative infection, or reconstruction modality and laterality. Preoperative BREAST-Q scores did not differ significantly between lymphedema and non-lymphedema groups. Postoperative BREAST-Q scores significantly differed for domains regarding physical-WB of the chest (from 6 months to 3 years post-operatively), breast satisfaction (1- and 2-years post-operatively), sexual well-being (2 years post-operatively) and psychosocial-WB (2- and 3-years post-operatively) (Table 1). For all domains with significant differences, the difference in average score was greater than the MCID. Conclusion:: A diagnosis of lymphedema significantly affects physical well-being of the chest, breast satisfaction, sexual well-being and psychosocial well-being of patients who have undergone breast cancer treatment. These differences manifest at different time points and have a significant clinical impact on patients. Knowing the impact of lymphedema on BREAST-Q scores helps physicians specifically address these domains at appropriate times when advising and interacting with patients.
Back to 2021 Abstracts