Hepatocellular carcinoma surveillance may be associated with potential psychological harms in patients with cirrhosis

Abstract

Background and aims: The value of HCC surveillance is determined by the balance between benefits and harms; however, no studies have enumerated psychological harms.

Approach and results: We fielded surveys measuring psychological harms to patients with cirrhosis in a multicenter randomized trial of HCC surveillance outreach. All patients with positive or indeterminate surveillance results and matched patients with negative results were invited to complete surveys measuring (1) depression through the Patient Health Questionnaire-ninth version, (2) anxiety through State-Trait Anxiety Inventory, (3) HCC-specific worry through Psychological Consequences Questionnaire, and (4) decisional regret. Patients were classified into 4 groups: true positive (TP), false positive (FP), indeterminate, and true negative (TN). Multivariable longitudinal regression analysis using the generalized estimating equation method was performed to compare the means of measures across groups. We conducted 89 semistructured interviews in a subset of patients stratified by health system and test results. Of 2872 patients in the trial, 311 completed 1+ follow-up survey (63 FP, 77 indeterminate, 38 TP, and 133 TN). Moderate depression decreased in TN patients, increased in TP, and had intermittent but mild increases in those with FP and indeterminate results. High anxiety temporarily increased in patients with TP results but resolved over time and was stable in those with FP and indeterminate results. Decisional regret was low and did not differ across groups. In semistructured interviews, patients reported apprehension, anxiety, emotional distress, and coping related to HCC surveillance.

Conclusions: Psychological harms of HCC surveillance appear mild but differ by test result. Future research should determine the impact of psychological harms on the value of HCC surveillance programs.

Figure 1A:

Adjusted mean proportion of patients with moderate depression over time

Figure 1B:

Adjusted mean proportion of patients with high anxiety over time Models were adjusted for demographic and clinical characteristics (e.g., age, sex, language, race/ethnicity, liver disease etiology, Child-Pugh-Turcotte score, and site).

Figure 2.

Changes in proportion of patients with fear of dying from HCC, one of the measures of HCC-specific distress/worry