Determinants of poor quality of life in older patients with cancer: A longitudinal mixed methods study (PROGNOSIS-Q).

INTRODUCTION

Patient-reported outcomes are becoming more employed in oncologic research because many older patients with cancer prioritize preserved health-related quality of life (HRQoL) over prolonged survival. However, few studies have examined the determinants of poor HRQoL in older patients with cancer. This study aims to determine whether HRQoL findings are truly reflective of cancer disease and treatment, as opposed to external factors.

MATERIALS AND METHODS

This longitudinal, mixed-methods study included outpatients, age 70 years or more, with a solid cancer, who reported poor HRQoL (EORTC QLQ-C30 Global health status/QoL (GHS) score ≤ 33.3), at treatment initiation. A convergent design was employed, in which HRQoL survey data and telephone interview data was collected in parallel at baseline and three-months follow-up. Survey and interview data was analyzed separately and subsequently compared. Thematic analysis of interview data was conducted according to Braun & Clarke, and changes in patients GHS score were calculated using mixed model regression.

RESULTS

Twenty-one patients with a mean age of 74.7 years were included (12 men and 9 women) and data saturation was achieved at both time intervals. Baseline interviews (n = 21) showed that poor HRQoL at cancer treatment initiation was primarily reflective of participants' initial shock upon receiving their cancer diagnosis and their change in circumstance and sudden functional independence. At three months, three participants were lost to follow-up and two provided only partial data. Most participants experienced an increase in HRQoL, with 60% showing a clinically significant improvement in GHS scores. Interviews showed that this was due to lessening functional dependency and disease acceptance achieved by mental and physical adjustment. HRQoL measures were less reflective of cancer disease and treatment in older patients with preexisting highly disabling comorbidity.

DISCUSSION

This study showed good alignment between survey responses and in-depth interviews, demonstrating that both methodologies are highly relevant measures during oncologic treatment. However, for patients with severe comorbidity, HRQoL findings are often more reflective of the steady state of their disabling comorbidity. Response shift may play a part in how participants adjusted to their new circumstances. Promoting caregiver involvement from the time of diagnosis may increase patients´ coping strategies.