Family Mechanisms and Readiness for Advance Care Planning Engagement among Elderly Patients with Cancer in Songkhla Province, Thailand

Abstract

BackgroundAging and multiple comorbidities make the elderly vulnerable to poor health outcomes. In elderly people with cancer, treatment such as chemotherapy or radiotherapy can cause more adverse effects, and sometimes the elderly’s response to such treatment can also be limited. Decision-making with elderly cancer patients when such undesirable effects arise might not be possible due to a lack of mental capacity. Advance care planning (ACP) is integrated into cancer care as a process of communication among clinicians, patients, and family members to ensure respect for patients’ preferences. However, Thai people prioritize family and sometimes their planning is hindered by family mechanisms, including a conspiracy of silence. The current study aims to: 1) understand perceptions of elderly with cancer and their caregivers pertaining to advance care planning, 2) examine whether their perspectives were congruent, 3) describe the elderly’s readiness for advance care planning engagements, and 4) examine the relationships between individual and family factors in advance care planning engagements of the elderly. MethodsPhase 1 study (January – April 2023) aimed to explore perceptions of advance care planning among elderly cancer patients and their caregivers. Participants were recruited from the oncology radiotherapy clinic at Songklanagarind Hospital. Inclusion criteria for patients included being aged 60 years or more, diagnosed with non-hematological cancer, and able to communicate in Thai, while their primary caregivers needed to be adult Thai speakers who provided unpaid care. Those with severe mental or physical conditions, brain tumors, or communication barriers were excluded. After obtaining ethics approval, we approached potential informants through registration staff informing participants about the study. The study employed maximum variation sampling to ensure a diverse participant pool. Data collection involved semi-structured interviews using an interview guide, covering topics such as illness perceptions, benefits and barriers to advance care planning, and related values. Caregivers were interviewed before patients to safeguard elderly participants from sensitive issues potentially introduced during the interviews. The interviews lasted approximately 45 minutes, with permission for audio recording and note-taking. Thematic analysis was applied, guided by the Health Beliefs Model, to identify perceptions influencing readiness for advance care planning engagement. Trustworthiness and transferability were ensured through measures such as investigator triangulation and thick data descriptions. Coding and theme development were conducted collaboratively and audited, with findings reviewed by oncology and palliative care experts. The second phase (August 2023 – October 2024) involved a cross-sectional study in which the remaining objectives were addressed. We recruited the elderly and their caregivers using the same eligibility criteria as was applied in the qualitative study. Demographic and readiness data were described using descriptive statistics. Comparisons of advance care planning perceptions between patients and caregivers utilized intraclass correlation for consistency, and confirmatory factor analysis was conducted using the lavaan package to validate a pros and cons questionnaire. These models assumed a correlated structure and applied fit indices such as root mean square error of approximation, standardized root-mean-square residual, and comparative fit index, with dynamic fit indices enhancing robustness. Factor scores were computed to standardize results, ensuring alignment with prior studies.Bayesian network learning was applied to explore potential causal relationships between family mechanisms and advance care planning engagement. The qualitative findings informed variable selection, aiming to reduce overfitting and ensure interpretability. Family background variables were grouped using latent class analysis. The Bayesian network structures were built through score-based algorithms. Expert input and mandatory arc specifications strengthened the model's conceptual validity. Sensitivity and intervention analyses were conducted to identify nodes that were sensitive to parameter alterations (due to a change in context) and simulate interventions using a "do-operator". Tools such as GeNIe Modeller and R packages (bnlearn, igraph, and caret) facilitated structure and parameter learning, sensitivity analysis, and intervention analysis.ResultsIn the qualitative study, we interviewed 16 elderly with cancer and 16 caregivers. Three themes regarding perceptions of advance care planning were found: 1) Advantageous opportunity: the patients believed the planning would help them realize their life values, and ensure that their preference would be respected; 2) contemplations and barriers to advance care planning: unfamiliarity and thoughts of it being unnecessary might cause low utility, worry among patients and family members, take away optimism, and would not be a proper activity for the patient at the current health situation; and 3) cues for initiation: perceived conformity with one’s religion, awareness of the current cancer state, having multiple comorbidities or experience with suffering related with medical care, wishing not to burden the family, having close family members, and trust in physicians.In the second phase, we recruited 279 elderly patients with cancer and their caregivers, resulting in 558 participants. Totally, 54.8% and 58.3% of participants were prepared, but never participated in any advance care planning discussions pertaining goals of care and life-sustaining treatment with their physicians. On the other hand, discussions with the family regarding the two types of planning was relatively higher (29.1% and 23.1% had maintained discussions pertaining goals of care and life-sustaining treatment with the families, respectively, for more than 6 months). We also found that less than 5% of the elderly had a durable power of attorney for health care and a living will. The prior awareness of any types of advance care planning were low; approximately 4% of the participants were knowledgeable of legalized advance care planning–durable power of attorney for health care, and living will. Perceptions towards advance care planning and general health values between the elderly and their caregivers were commonly incongruent, although small agreements were found on the ideas that advance care planning can help one control the future care and the present could be a good timing. From the Bayesian network analysis and the directed acyclic graph, the marginal probability suggest a trend of perceiving advance care planning as advantageous among the elderly patients. However, the caregivers were equivocal about the advantages but generally perceived that the disadvantages were low. The Bayesian network learning produced a directed acyclic (causal) graph in which the older persons' pros and cons are directly linked to readiness for certain types of ACP engagement (discussing life-sustaining treatment use with family or clinician, discussing goals of care with clinician, and appointing a durable powever of attorney for healthcare). The effects of the elderly's perceptions demonstrated dominant effects on ACP engagement than the caregiver’s perceptions. Hypothetical interventions on the patients' and caregivers' perceived pros and cons of advance care planning to promote engagement were found to be possible (Bayes factor >10). The patient's cons showed large causal effects on the motivations for advance care planning (14.9% and 43.8% had an increased probability of documenting advance directives and discussing cardiopulmonary resuscitation with physicians, respectively). Finally, the causal graph also shows a lack of causal relationship between family functioning and advance care planning engagement in older patients with cancer. Conclusions Families of older Thai cancer patients often viewed advance care planning discussions as unfamiliar, unnecessary, and potentially distressing, fearing they might cause worry, diminish optimism, or seem inappropriate. However, they recognized its value in ensuring that patient preferences are respected. While many older patients were prepared to engage in discussions about advance care planning with family and care providers, a significant number remained in the precontemplation or contemplation stages regarding formal documents like advance directives or living wills. Favorable attitudes were observed, but disagreements between patients and caregivers were notable. Advance care planning was seen as a way to maintain mindfulness and control over future medical care, suggesting that now might be an opportune time to encourage such discussions.Patients who were motivated for certain advance care planning behaviors were more likely to be ready for others. However, formal documentation was less likely to be adopted. Interventions targeting pros and cons had varying effects, with promoting pros being relatively effective for durable power of attorney for health care, while reducing cons had a stronger impact on other advance care planning behaviors. In time-limited oncology settings, prioritizing cons may be more efficient. Designing personalized interventions based on pros and cons could support patient engagement, but care professionals should remain mindful of differing perspectives between patients and caregivers to ensure patient autonomy is respected.