Measuring symptom burden in patients with cancer during a pandemic: the MD Anderson symptom inventory for COVID-19 (MDASI-COVID)

Abstract

Background Symptom expression in SARS-CoV-2 infection (COVID-19) may affect patients already symptomatic with cancer. Patient-reported outcomes (PROs) can describe symptom burden during the acute and postacute stages of COVID-19 and support risk stratification for levels of care. At the start of the COVID-19 pandemic, our purpose was to rapidly develop, launch through an electronic patient portal, and provide initial validation for a PRO measure of COVID-19 symptom burden in patients with cancer. Methods We conducted a CDC/WHO web-based scan for COVID-19 symptoms and a relevance review of symptoms by an expert panel of clinicians treating cancer patients with COVID-19 to create a provisional MD Anderson Symptom Inventory for COVID-19 (MDASI-COVID). English-speaking adults with cancer who tested positive for COVID-19 participated in the psychometric testing phase. Patients completed longitudinal assessments of the MDASI-COVID and the EuroQOL 5 Dimensions 5 Levels (EQ-5D-5L) utility index and visual analog scale, which were presented through an electronic health record patient portal. To test the validity of the MDASI-COVID to distinguish between known groups of patients, we hypothesized that patients hospitalized, including having a hospitalization extended, for COVID-19 versus those not hospitalized would experience higher symptom burden. Correlation of mean symptom severity and interference scores with relevant EQ-5D-5L scores tested concurrent validity. The reliability of the MDASI-COVID was evaluated by calculating Cronbach alpha coefficients and test-retest reliability was evaluated by calculating Pearson correlation coefficients between the initial assessment and a second assessment no more than 14 days later. Results The web-based scan found 31 COVID-19-related symptoms; rankings of a 14-clinician expert panel reduced this list to 11 COVID-specific items to be added to the core MDASI. Time from literature scan start in March 2020 to instrument launch in May 2020 was 2 months. Psychometric analysis established the MDASI-COVID’s reliability, known-group validity, and concurrent validity. Conclusions We were able to rapidly develop and electronically launch a PRO measure of COVID-19 symptom burden in patients with cancer. Additional research is needed to confirm the content domain and predictive validity of the MDASI-COVID and define the symptom burden trajectory of COVID-19.

The advantages, disadvantages, threats, and opportunities of electronic patient-reported outcome systems in cancer: A systematic review

“…The most prevalent disadvantage was “Small sample size,” cited in 19 studies (15.70%), reflecting a common challenge in research methodologies that may limit the generalizability of findings, 1 , 17 , 33 , 34 , 38 , 47 , 54 , 56 , 58 , 63 , 65 , 70 , 72 , 74 , 83 , 85 , 87 , 90 , 91 followed by “Technical issues” with 16 references (13.22%), highlighting the challenges posed by technology in the implementation of electronic health platforms 18 , 19 , 24 , 34 , 39 , 47 , 50 , 60 , 66 – 68 , 74 , 76 , 81 , 92 , 93 and “Potential bias” with 12 references (9.92%), emphasizing the need for researchers to address and mitigate bias in participant selection, data extraction and reporting. 16 , 22 , 40 , 41 , 43 , 57 , 65 , 74 , 84 , 85 , 94 , …”

“…“Potential bias” emerged as a notable disadvantage, highlighting the need for researchers to address and mitigate bias in participant selection. 16 , 22 , 40 , 41 , 43 , 57 , 65 , 74 , 85 , 94 , 95 This indicated the critical need for researchers to address and mitigate participant selection bias, aligning with a broader emphasis on enhancing the reliability and validity of study findings. The findings contributed substantially to a comprehensive understanding of biases in ePRO studies, aligning with the overarching goal of thoroughly exploring drawbacks and providing insights for refining study designs.…”