The Assessment of Social Isolation and Loneliness in Cancer Patients and Survivors in the Pre-COVID-19 Period: A Systematic Review
Authors:
Marziliano, A., Byakova, A., Patel, P., Herman, S. W., and Diefenbach, M. A.
Abstract:
BACKGROUND: In the context of cancer research, identifying social isolation and loneliness is a priority given how both exacerbate poor outcomes and lead to increased mortality in oncological populations. The purpose of this systematic review is to identify all quantitative instruments that have been used to assess either social isolation or loneliness in patients previously or currently diagnosed with cancer in the pre-COVID-19 period.
METHOD: PubMed (Web), Scopus, CINAHL, and PsycINFO were searched on August 22, 2019. All databases were searched from inception with no filters applied. The search strategies included terms that captured the following concepts: instruments/tools, social isolation or loneliness, and cancer.
RESULTS: A total of 289 titles/abstracts were returned. Upon review, 114 titles/abstracts were deemed to be potentially eligible and the full text was retrieved. Of the 114 full texts, 69 articles met inclusion criteria and comprised the final sample. Publications span years 1980 through 2019, with the majority (71%) occurring in the last decade prior to this review, between 2009 and 2019. Average age of the study samples, with few exceptions, was often over 50 years old. Many studies used all-female samples, while only one study used an all-male sample. The most common cancer diagnosis of participants was breast cancer. The most common measure was the UCLA Loneliness Scale, used in 22 studies. Most measures we identified were used only once, and 11 measures were used 2-3 times. When the information was given, response ranges were always Likert-type scales most often ranging from 1-4 or 1-5, and sometimes from 1-10 possible response options. In terms of psychometrics, test-retest reliability and validity were rarely reported; by contrast, internal consistency (Cronbach's alpha) was reported more than half of the time (60.9%).
CONCLUSION: When selecting a measure to assess loneliness in cancer populations, the UCLA Loneliness Scale is both psychometrically strong and versatile across patients with different cancers, ages, and racial backgrounds. When selecting a measure to assess social isolation in cancer populations, both the PROMIS-SF V 2.0 social isolation and the Berkman-Syme Network Index are brief and have been used in patients with non-White racial backgrounds.