Psychological Correlates of Having Advance Care Planning in Patients with Hematological Malignancies.

Previous studies in patients with hematologic malignancies who underwent hematopoietic cell transplantation show only 50% have participated in advance care planning (ACP) at the time of transplant. Moreover, the group that lacked ACP had a higher mortality rate, suggesting that the group least likely to have ACP is the group most likely to need it. Having ACP is thought to be an indicator of quality of care, thus efforts to increase in engagement of ACP are becoming part of clinical practice. However, many argue that discussions about possible death can potentially add to the stress cancer patients already face. We therefore studied the psychological correlates of having ACP to help understand how best to design intervention(s) that may increase engagement in ACP.

Data are derived from the HEMA-COMM study cohort – an observational study of patients with hematologic malignancies evaluating doctor-patient communication; 293 patients had self-reported information about ACP and were categorized as to presence or absence of ACP. ACP+ was defined as both having a designated health care proxy and completion of a living will. ACP- lacked one or both. Variables compared include: socio-demographic characteristics, diagnosis, previous treatments, patient and physicians estimate of cure and life expectancy, discussions of wishes for life support, and various measures for coping (BRIEF-COPE), depression/anxiety (HADS), denial, social support (MOS) and quality of life (SF-36). Multivariate logistic regression was used to evaluate the association between the different psychosocial measures and having ACP while adjusting for patient characteristics.

Of the 293 patients included, 149 (51%) had ACP. Patients with ACP were more likely to be older (56y v 52y), have higher income, have leukemia or MDS, and have received prior cancer treatment. Of the patients with ACP, only 30% had discussed them with physicians with or without family, 62% had discussed with family only, while 8% had not discussed wishes for life support at all. Patients with ACP had worse estimated prognosis as measured by lower cure rate and life expectancy estimates provided by their physicians. In the multivariate analysis after adjustment for age, diagnosis, income, prior treatment, and patients' perceived life expectancy, patients who use the following coping styles were more likely to have ACP: 1) use of instrumental support (advice, assistance or information), p<0.001; 2) active coping (doing something), p=0.006; 3) use of emotional support (seek moral support, discuss feelings), p=0.009; 4) planning (strategize a plan of action), p=0.04; and 5) positive reframing (positive outlook), p=0.04. These coping styles are all consistent with problem-focused coping as opposed to emotion-focused coping. Discussing ACP wishes with physicians was associated with general health (p=0.007) and overall physical composite SF-36 score (p=0.03). Measures of depression, anxiety, social support, and level of denial were not associated with having or discussing ACP.

Our results suggest that patients who engage in ACP have problem-focused coping patterns and may perceive a need based on worse prognosis. Our findings could also suggest that engagement in ACP may be reflective of a patient's realistic or pragmatic view of clinical outcomes. Therefore, efforts to increase engagement in ACP should emphasize the practical benefits of ACP regardless of prognosis and de-emphasize the emotional content of ACP. These approaches may be more appealing to patients regardless of coping style and help increase ACP participation.

No relevant conflicts of interest to declare.