Concordance of Patient and Caregiver Reports on the Quality of Colorectal Cancer Care

Cancer symptoms: do patients, family caregivers and professionals agree?

OBJECTIVES

To assess the level of agreement among patients, their family caregivers and health professionals regarding the symptoms presented by patients with cancer.

METHODS

This is a cross-sectional study carried out in patients with cancer admitted to a hospital in Brazil from December 2019 to July 2020. One family caregiver for each patient was included in the study. In addition, nurses and physicians involved in patient care were included. Patients, family caregivers and health professionals responded to the Palliative Outcome Scale (POS).

RESULTS

Sixty-one patients with their family caregivers, 18 nurses and 8 physicians were included. Physicians reported a lower mean POS score than patients (p=0.008). In general, physicians and nurses underestimated the anxiety of the patient and the anxiety of the family caregivers and how much the patient feels that his or her life is worthwhile. Intraclass correlation coefficient showed moderate and good levels of agreement between patients' and family caregivers' responses (0.61). Agreement between patient and nurse (0.02) and physician (0.21) responses was poor.

CONCLUSIONS

The level of agreement between patients and healthcare professionals was very poor regarding patients' symptoms. However, the caregiver's perception of the patient's symptoms can be a reliable source of information.

The potential impact of proxy reports for symptom experience and care quality and experience in advanced cancer

OBJECTIVES

As the US tests models of care for the seriously ill, patient perceptions of the quality of care are important. Proxies are often needed for this group. We sought to understand the potential impact of proxy reports for the assessment of care quality and experience in cancer.

METHODS

Secondary data analysis of a deidentified prospective study that included surveys of perceived care quality, including symptom management, from patients with advanced cancer receiving chemotherapy and their caregivers. Surveys were administered at diagnosis (time 1) and treatment (time 2), with top-box scoring used for analysis. Overall concordance was assessed using metrics including Gwet's AC1. The proportion of the highest scores by respondent type within 2 subgroups were examined: (1) symptom burden and (2) practice setting.

RESULTS

Data from 83 dyads were analyzed. Proxies and patients frequently reported the highest scores for quality (time 1: proxies: 77% and patients: 80%). At time 1, 14% of proxies and 10% of patients reported an unmet need for symptom palliation. Most patients reporting an unmet need gave the top score for quality (75%), but fewer proxies did so (45%). Proxy and patient reports were similar within practice settings. Concordance was at least moderate (nearly all outcomes >0.5 and some >0.8) by Gwet's AC1.

SIGNIFICANCE OF RESULTS

Findings of at least moderate concordance and similar experience outcomes within subgroups suggest the use of proxies may not change estimates substantially. However, consideration should be taken when evaluating symptom management, particularly if such evaluations inform assessment of provider performance.

Need differences by treatment phases between patients with colorectal cancer and their caregivers: A text mining analysis

Objective

Methods

Results

Conclusions