Correlation between Quality of Life and Burden in Caregivers of Advanced Stage Cancer Patients on Best Supportive Care

INTRODUCTION

Cancer ranks as one of the leading causes of death globally.^[1] Given the late presentation to centres for treatment, locoregionally advanced or metastatic nature of the disease,^[2] a sizable proportion of the patients become unfit for curative therapies. They are often started on palliative therapies which can be chemotherapy, radiation therapy and rarely surgery to prolong life with a good performance status for as long as possible. At some point in time, the patient may no longer be a candidate for further therapy. These patients are referred to as on Best Supportive Care in this study. Moreover, given inadequate access to hospice facilities, most of the patients are cared for at homes by families, which can impact the family at large, primarily their family caregivers. There is paucity of data regarding Quality of Life (QOL) and caregiver burden among family caregivers of cancer patients in this subgroup. We hypothesised that the burden felt by the caregiver might adversely impact their QOL. Hence, we did this study to see the correlation between the QOL of the patient’s caregiver and the burden of caregivers for advanced-stage cancer patients who are on Best Supportive Care in our hospital.

We also assessed the patient’s QOL and tried to evaluate its relationship with the burden felt by the caregiver.

MATERIAL AND METHODS

This is a prospective and cross-sectional study done in the Palliative Care Unit of a tertiary care centre in India. This study was started after the Institutional Ethics Board approval. The patients were with advanced cancer no longer suitable for any oncologic therapy, as decided by their treating oncologist. These patients are referred to be on Best Supportive Care and are on OPD follow-up in the Palliative Care Unit of our centre. We defined the family caregiver as the person most involved in the patient’s care without receiving financial reimbursement for the care that they provided. We included all consecutive advanced stage cancer patients on Best Supportive Care on OPD follow-up and their respective caregivers. We excluded patients <18 years of age, those on any form of oncologic treatment and those unable to complete the questionnaires. After the patient has finished the follow-up with the Palliative Care Clinic and after informed consent from both Patient and Caregiver, the principal investigator collected demographic and clinical details, then assessed the patient’s QOL from the patient using the EORTC QLQ C15 PAL questionnaire, local language version and the Caregiver Burden and Caregiver QOL from the Caregiver using the Zarit Burden Interview and the WHO QOL BREF questionnaire, respectively, local language version, in a single setting.

Before the start of the study, local language versions were obtained from the respective organisations for use in the study. Scoring was done per the scoring manuals of the respective questionnaires. Of note, the WHO QOL BREF domain scores were transformed in the 4–20 range. Higher score in the ZBI total score and symptom scales of EORTC C-15 PAL represents higher burden and higher symptoms. Higher scores in the WHO QOL BREF domains, EORTC C-15 PAL functional scores and global QOL score represent better QOL/functioning.

RESULTS

In this cross-sectional study, 220 patients and their respective 220 family caregivers were enrolled in this study. Data from 440 subjects were used for statistical analysis.

Mean age of the patients was 53.8 years (SD – 14.3). Age of the patients ranged widely from 18 to 90 years of age. One hundred and eight (49.1%) were males. Majority were married (n = 170, 73.5%), Hindu, illiterate or educated up to primary school. The most common primary cancer diagnosis in our study was gastrointestinal malignancies (n = 89, 40.4%). The subsite of gastrointestinal malignancies included in the study in order of decreasing frequency was carcinoma of the stomach, oesophagus, gallbladder, colon and pancreas. Most patients in our study were with metastatic disease, with Eastern Cooperative Oncology Group performance status 3. Among the patients with metastatic disease or distal recurrence, lung metastasis (73.3%) was the most common, followed by liver/peritoneal metastasis (51.8%). Majority of the patients (n = 96) were on active oncologic interventions for < 6 months in our study. Some patients (n = 39) were never on any active oncologic therapy previously. In our study, 42.3% (n = 93) had previous surgery, 64.1% (n = 141) had previous chemotherapy and 48.6% (n = 107) had previous radiotherapy. Patient characteristics are summarised in [Table 1].

The mean age of the family caregivers was 45.5 years (SD-14). Age of the family caregivers ranged widely from 18 to 78 years. A larger proportion of caregivers were female (n = 119, 54.1%), homemakers, residing in rural areas (n = 128, 58.2%) and spouses of the patient. Caregivers were generally better educated compared to the patients’ group. The characteristics of family caregivers are summarised in [Table 2].

The median score for ZBI in the study population (n = 220 Caregivers) was 39 (IQR – 10). The median scores for the caregiver QOL (n = 220 Caregivers) as assessed by the domains of the WHO QOL BREF questionnaire were for the physical, psychological, social and environmental domains 16 (IQR-2), 13 (IQR-3), 13 (IQR-3) and 12 (IQR-3), respectively. The patient QOL as measured by the EORTC C15 PAL questionnaire (n = 220 patients) showed median scores of 55.6 (IQR-22.2), 33.3 (IQR-33.3) and 16.7 (IQR-16.7), respectively, for physical functioning, emotional functioning and global QOL scores.

Correlation between Caregiver Burden and Caregiver and Patient QOL was done using the Spearman test. The Spearman Rho values for the correlation of ZBI total scores with the WHO QOL BREF Domain scores are mentioned in the [Table 3].

Our study tried to find clinicosocial parameters associated with statistically significant changes in the caregiver burden. The ZBI total score was found to significantly vary with the age and gender of the patient being cared for, different cancer diagnoses, ECOG performance statuses. Caregiver occupation, caregiver educational qualifications, relation of caregiver with the patient and family income were also factors with statistically significant differences in ZBI Total Scores, as shown in [Table 4].

DISCUSSION

Results of this study from 440 study participants, that is, 220 patients and their respective 220 family caregivers, showed a statistically significant negative (Spearman) moderate correlation between the caregiver burden as assessed by ZBI total score and the social domain of the WHO QOL BREF, and low correlation between ZBI total score and the psychological and environmental domains of WHO QOL BREF Questionnaire. A similar result has been reported by Rha et al.,^[3] in their study among caregivers of patients on oncologic therapy in South Korea, had achieved a statistically significant correlation between caregiver burden (ZBI scores) and caregiver QOL (WHO QOL BREF Domains). However, our study did not report a statistically significant correlation of burden with the physical domain of WHO QOL BREF, possibly due to our different caregiver study population, which is caregivers of patients who are not on any active oncologic therapy. A similar significant negative correlation was discussed in another study,^[4] but they used different questionnaires to assess caregiver burden. Our study suggests that worse patient QOL or higher patient symptoms will negatively impact the caregiver by increasing the severity of the burden perceived by the caregiver. This is similar to the conclusions drawn by Lee et al.^[5] and Tang et al.,^[6] although with different questionnaires. Other studies have used cutoffs in ZBI total scores to classify caregivers as those with low burden, high burden and severe burden, like Lukhmana et al.^[7] from Delhi. They had reported the majority of their patients with no burden or mild-to-moderate burden, with their ZBI total scores being less than 40 for >90% of their study population. The median score here, however, was 39. Another study^[8] done on caregivers of patients on chemotherapy in India has also shown around 70% of the caregivers reporting mild-to-moderate burden and around 21% reporting moderate-to-severe burden with ZBI. In this study, however, we noticed 37.2% of caregivers reported moderate-to-severe burden and 58.6% reported mild-to-moderate burden. This can be because they had chosen caregivers of patients on treatment, while, in this study, we chose caregivers and patients who are not on any active oncologic therapy and in advanced disease stages.

We also tried to assess clincosocial parameters with caregiver burden. Due to the scarcity of data, comparisons were drawn from studies on patients on treatment instead of our population, which are patients not on any oncologic therapy. Caregiver burden scores were on average significantly higher for caregivers of younger patients (<55 years) as compared to older patients. This finding was also noted by Choi et al.,^[9] but they suggested that though family caregivers of younger patients felt a higher burden, they were more likely to adapt positively.

The statistically significant differences in the ZBI caregiver burden scores based on cancer diagnosis were noted on subgroup analysis due to melanoma patients versus other groups. This could be because these patients had low volume metastatic disease with better ECOG performance status than others. Caregiver burden scores were on average lower with better ECOG Performance status of the patient, similar to previous studies in cancer patients.^[3,10,11]

Spouses reported on average a statistically significant higher caregiving burden than other relations in our study. Other studies have also shown similar results with co-resident caregivers in the general population.^[12] Illiterate caregivers were noted to have, on average, a higher burden compared to well-educated caregivers, as noted in a previous study.^[13] On average, it was noted that salaried employees had lower caregiver burden, similar to a study done previously.^[3] Caregiver burden scores were significantly different between different income groups. The general trend was towards lower caregiving burden scores for higher-income groups, similar to a previous study.^[3]

CONCLUSION

This study shows that a higher caregiving burden is associated with a decrease in the caregiver’s QOL. This study also shows that a decline in patient QOL or increase in the symptom scores of patient can be associated with a higher perceived caregiving burden.

Many socioeconomic and clinical factors were associated with the caregiving burden scores or patient QOL and symptom scores, or caregiver QOL scores, and these factors require further analysis and studies to target these factors with interventions or with longitudinal follow-up to aim to decrease perceived caregiver burden and improve the QOL of the patient and caregiver.