Collusion in healthcare refers to the concealment, distortion or selective communication of clinically important information among the patient, family and healthcare team. In palliative care it is frequently motivated by love, protection and fear that diagnostic or prognostic information will produce hopelessness, depression or treatment refusal. Nevertheless, the protective intention can create a 'conspiracy of silence' in which the patient is excluded from conversations about the illness and future care.^1,2

Family caregivers occupy several overlapping roles during life-limiting illness. They provide physical care, coordinate appointments, manage finances, interpret medical information and often serve as surrogate decision-makers. These responsibilities intensify when symptoms worsen or when the patient becomes functionally dependent. Caregivers therefore carry substantial emotional and practical burdens while also mediating communication between clinicians and patients.^3,5 In family-centred cultures, the family may expect to receive serious information first and to determine how much the patient should know. Such practices can reflect relational responsibility, but they can also conflict with the patient's preferences, autonomy and right to participate in decisions.^2,4

The clinical consequences of collusion extend beyond disclosure itself. Withholding information can make symptoms difficult to explain, delay goals-of-care discussions, complicate consent for treatment, create anxiety about accidental discovery and restrict opportunities to complete personal, financial or relational matters. Collusion can also affect quality of life and the therapeutic relationship.^6,7 Conversely, direct and insensitive disclosure can overwhelm a patient or family. Contemporary guidance therefore favours communication that is truthful, paced, empathic and responsive to the amount of information the patient wishes to receive.^8-12

Collusion is especially complex in India, where family involvement in medical decisions is common and where illness trajectories are often shaped by late presentation, financial toxicity, fragmented care and limited access to specialist palliative services. Previous Indian studies have described the prevalence, correlates and consequences of collusion, but fewer have examined how caregivers move between withholding, partial disclosure and open communication as circumstances change.^3,4,7,8 The present study therefore explored the psychological, relational and environmental mechanisms that influence collusion among family caregivers of patients receiving palliative care.

Study design and setting This qualitative interview study was conducted from February to August 2025 in the pain and palliative care department of a tertiary care hospital in Bengaluru, Karnataka, India. The service provides outpatient, inpatient and emergency palliative care to patients referred from several districts. Most patients had advanced cancer; a smaller proportion had progressive pulmonary, cardiac, hepatobiliary or neurological disease. The manuscript was prepared with reference to the Consolidated Criteria for Reporting Qualitative Research (COREQ).14 Participants and recruitment Family caregivers who provided informed consent and could communicate in Kannada or English were eligible. Caregivers were approached during the patient's contact with the palliative care service. Some had received brief counselling regarding disclosure from the palliative care or primary oncology team before the research interview. Recruitment continued until the investigators considered that no substantively new themes were emerging. The final analytic sample comprised nine caregivers of nine patients with life-limiting illness. Interview guide and data collection A semi-structured interview guide was refined after two pilot interviews, review of the literature and discussion within the research team. The guide explored the caregiver's understanding of the illness, immediate emotional reactions, beliefs about disclosure, perceived patient preferences, family influences, previous experiences, treatment decisions and the anticipated consequences of telling or withholding information. After rapport had been established, an investigator conducted an individual interview with the caregiver in the absence of the patient. A brief clinical conversation with the patient preceded the caregiver interview to clarify the illness context. Interviews used open questions, prompts and probes, lasted approximately 30-60 minutes and were audio-recorded. Clinical care continued independently of research participation; treating clinicians remained responsible for communication, goals-of-care discussions and management of collusion. Data preparation and analysis Audio-recordings were transcribed and, when required, translated into English and back-translated. Minor grammatical editing was applied to translated excerpts for readability without altering the intended meaning. The investigators used directed qualitative content analysis, in which initial areas of inquiry were informed by prior literature while categories and subthemes were refined iteratively from the data.13 Two investigators independently coded the transcripts, compared coding decisions and resolved differences through discussion. Themes and subthemes were developed through repeated review and synthesis. Investigator triangulation was used to reduce the influence of a single researcher's interpretation. Numbers shown in parentheses beside selected subthemes represent coding frequencies reported by the study team and should not be interpreted as participant prevalence. Ethical considerations Approval was obtained from the Institutional Ethics Committee before recruitment. Informed consent was obtained from each caregiver. Interviews were conducted privately, and quotations were edited to remove potentially identifying details. The ethics committee name and approval number should be inserted before journal submission.

Participant profile

Nine caregivers of nine patients participated. Seven caregivers were classified as colluding because they preferred to withhold or substantially limit information about diagnosis or prognosis. Two caregivers supported open discussion and were classified as non-colluding. All participants completed an interview lasting 30-60 minutes. Because the groups were small and markedly unequal, the descriptive values in Tables 1 and 2 are presented only to characterize the sample; no inferential comparisons were undertaken.

Table 1. Characteristics of patients according to caregiver disclosure position

Data are n (%) unless otherwise stated. BADL, basic activities of daily living; IADL, instrumental activities of daily living.

Table 2. Characteristics of participating caregivers

Data are n (%) unless otherwise stated. Percentages are descriptive because of the small group sizes.

Qualitative findings

The analysis generated seven interconnected themes. Caregivers described an evolving process rather than a single disclosure decision. The position adopted at diagnosis could change after counselling, symptom progression, a major treatment decision or a direct request for information from the patient.

1. Immediate reactions to the diagnosis. Caregivers commonly recalled shock, denial, anger, sadness and uncertainty. Sudden recognition of advanced disease, delayed diagnosis and rapid referral to palliative care intensified distress. Financial strain and dissatisfaction with fragmented explanations across healthcare settings compounded the sense of crisis. A smaller group described gradual acceptance or a readiness to focus on comfort.

2. Emotional challenges surrounding disclosure. Caregivers viewed information control as a way to protect the patient from despair, loss of appetite, withdrawal or loss of motivation. They also tried to preserve the patient's premorbid identity, routines and independence. These goals could coexist with guilt, persistent worry and uncertainty about whether withholding remained helpful.

3. Factors facilitating disclosure. Disclosure was easier when the patient was educated, read medical reports, explicitly requested information or had previously shared an open relationship with the caregiver. Prior experience of cancer in relatives or friends also helped some caregivers anticipate the illness trajectory. Direct communication by clinicians and the patient's independent discovery of information sometimes ended collusion abruptly.
4. Barriers to disclosure. The main barriers were fear of psychological harm, past stories of extreme reactions to a cancer diagnosis, rigid beliefs that the caregiver 'must' find a cure, disagreement within the family, lack of courage and the patient's reliance on relatives for decisions. Disclosure was repeatedly postponed when new crises emerged or when a temporary improvement created hope.

5. Consequences of collusion. Maintaining secrecy produced awkwardness when relatives visited, anxiety that the patient would discover the truth from another source and difficulty explaining progressive symptoms. Caregivers experienced constant pressure to pursue further treatment and often felt personally responsible for preventing suffering. Even after counselling, some preferred to delay disclosure or wished another person to initiate the conversation.

6. Perceived benefits of non-collusion. Open communication enabled joint decisions about chemotherapy, hospital admission and intensive care. Caregivers associated disclosure with respect for autonomy, reduced uncertainty and better explanation of symptoms. It also allowed patients to express preferences, meet relatives and attend to unfinished personal or family matters.

7. Gradually unveiling collusion. Caregivers rarely described disclosure as a single event. They favoured paced communication when the patient asked questions, when a treatment decision required informed participation or when symptoms made concealment implausible. Selective or partial disclosure was used when prognosis remained uncertain or when the caregiver believed that the patient was not emotionally ready.

Table 3. Themes, selected subthemes and representative caregiver quotations

Figure 1. Proposed dynamic model of caregiver collusion in a palliative care setting.

This study shows that caregiver collusion in palliative care is better understood as an evolving relational process than as a binary act of telling or not telling. Seven caregivers initially favoured withholding important information, usually to protect the patient from psychological harm. Their decisions were shaped by the suddenness of diagnosis, rapid clinical deterioration, symptom burden, financial strain, previous experiences, family influence and uncertainty about prognosis. At the same time, caregivers experienced guilt, vigilance and fear that the patient would learn the truth unexpectedly. These competing pressures explain why positions shifted across the illness trajectory. The protective intent described by participants is consistent with earlier literature from India and other family-centred settings. Collusion is often rooted in beneficence as understood by relatives: they believe that maintaining hope protects the patient and improves willingness to continue treatment.1-4 Yet patients can recognize deterioration, infer the diagnosis from treatment or read reports even when families attempt concealment. A systematic review of prognostic disclosure similarly concluded that many patients retain partial awareness despite nondisclosure and recommended assessing the information preferences of patients and caregivers separately.8 Functional dependence and established family decision patterns were prominent in the colluding group. When patients had long relied on relatives for major decisions, caregivers felt more entitled - and more responsible - to control information. This observation supports a relational account of autonomy in which choices are embedded in family relationships, while still underscoring that dependence does not eliminate the patient's moral or clinical interest in receiving information. Indian end-of-life guidance emphasizes consensus, respect for preferences and ethically defensible communication when major treatment-limitation decisions arise.9,10 The consequences of collusion were practical as well as emotional. Caregivers struggled to explain worsening pain, vomiting, weakness and hospital visits; they feared accidental disclosure and felt pressure to continue searching for curative treatment. Prior work has linked collusion with poorer communication and adverse quality-of-life consequences.6,7 In contrast, the two non-colluding caregivers described greater clarity about chemotherapy, intensive care and personal priorities. Disclosure created space for joint decisions and unfinished business, but the data do not support a simplistic conclusion that immediate full disclosure is always preferable. Participants repeatedly advocated disclosure that was gradual, timed and responsive to the patient's questions. This is compatible with structured approaches to serious-news communication, including preparation, assessment of the patient's information preference, empathic response and planning of next steps.11,12 Clinicians should therefore avoid joining a family in indefinite secrecy, but they should also avoid confronting caregivers in a manner that frames them as obstructive. A more constructive approach is to acknowledge the wish to protect, ask what harm is feared, clarify what the patient already understands, explore the patient's preferred level of information and revisit the conversation as the illness evolves. The proposed model extends previous descriptions by locating collusion on a continuum that includes colluding, partially colluding and non-colluding positions. Environmental pressure, coping style and the patient-caregiver relationship interact with illness tempo and decision urgency. The model is preliminary and requires testing in larger, culturally diverse samples that include patients, clinicians and caregivers. Longitudinal designs would be particularly valuable because a single interview cannot fully capture movement between secrecy, partial disclosure and open communication. Clinical implications • Assess the patient's desire for diagnostic and prognostic information separately from the caregiver's preference. • Acknowledge the caregiver's protective intent and explore specific fears before discussing disclosure. • Use repeated, paced conversations rather than treating disclosure as a one-time event. • Prepare caregivers for questions that arise as symptoms progress and for the possibility of unintentional disclosure. • Link communication with concrete decisions about treatment, symptom control, intensive care and unfinished personal matters. • Involve palliative care and mental health professionals when severe distress, family conflict or persistent avoidance impedes decision-making. Strengths and limitations A strength of the study was that interviews were conducted by a psychiatrist, allowing emotionally difficult narratives to be explored sensitively. Independent coding by two investigators and investigator triangulation strengthened interpretive credibility. The study also generated a clinically useful model that recognizes partial and changing forms of collusion. Several limitations require consideration. The sample was small, from a single centre and predominantly represented caregivers of patients with advanced cancer who were interviewed during periods of clinical crisis. Only two caregivers were classified as non-colluding, limiting contrast between disclosure positions. Patient and clinician perspectives were not included. The study did not report longitudinal follow-up, member checking or focus groups, and the transferability of the findings to non-cancer illness, community care and other cultural settings is therefore restricted.

Collusion among family caregivers in palliative care is a dynamic, emotionally charged response to serious illness rather than a stable communication preference. Caregivers commonly withheld information to protect patients, but secrecy also produced guilt, anxiety, uncertainty and difficulty with treatment decisions. Disclosure became more acceptable when patients requested information, deterioration made concealment difficult or a major decision required participation. Clinicians should assess patient preferences, validate caregiver fears and use staged, culturally sensitive communication supported by repeated counselling. A flexible, family-centred approach can preserve compassion while protecting the patient's opportunity to understand, decide and prepare.

Beng KS. Collusion in palliative care. Malays Fam Physician. 2006;1(2-3):62-64. 2. Chaturvedi SK, Loiselle CG, Chandra PS. Communication with relatives and collusion in palliative care: a cross-cultural perspective. Indian J Palliat Care. 2009;15(1):2-9. doi:10.4103/0973-1075.53485. 3. Sutar R, Chandra PS, Seshachar P, Gowda L, Chaturvedi SK. A qualitative study to assess collusion and psychological distress in cancer patients. Indian J Palliat Care. 2019;25(2):242-249. doi:10.4103/IJPC.IJPC_146_18. 4. Jeba J, Jacob A, Kandasamy R, George R. The patient who 'must not be told': demographic factors associated with collusion in a retrospective study in South India. Postgrad Med J. 2016;92(1093):659-662. doi:10.1136/postgradmedj-2015-133850. 5. Lung EYL, Wan A, Ankita A, Baxter S, Benedet L, Li Z, et al. Informal caregiving for people with life-limiting illness: exploring the knowledge gaps. J Palliat Care. 2022;37(2):233-241. doi:10.1177/0825859720984564. 6. Stiefel F, Nakamura K, Terui T, Ishitani K. Collusions between patients and clinicians in end-of-life care: why clarity matters. J Pain Symptom Manage. 2017;53(4):776-782. doi:10.1016/j.jpainsymman.2016.11.011. 7. Victor A, George CE, Inbaraj LR, Norman G. Benefit or harm? A study on impact of collusion on the quality of life among palliative care patients. Indian J Palliat Care. 2018;24(1):61-66. doi:10.4103/IJPC.IJPC_111_17. 8. Sutar R, Chaudhary P. Prognostic disclosure in cancer care: a systematic literature review. Palliat Care Soc Pract. 2022;16:26323524221101077. doi:10.1177/26323524221101077. 9. Myatra SN, Salins N, Iyer S, Macaden SC, Divatia JV, Muckaden M, et al. End-of-life care policy: an integrated care plan for the dying: a joint position statement of the Indian Society of Critical Care Medicine and the Indian Association of Palliative Care. Indian J Crit Care Med. 2014;18(9):615-635. doi:10.4103/0972-5229.140155. 10. Salins N, Gursahani R, Mathur R, Iyer S, Macaden S, Simha N, et al. Definition of terms used in limitation of treatment and providing palliative care at the end of life: the Indian Council of Medical Research Commission Report. Indian J Crit Care Med. 2018;22(4):249-262. doi:10.4103/ijccm.IJCCM_165_18. 11. Baile WF, Buckman R, Lenzi R, Glober G, Beale EA, Kudelka AP. SPIKES-a six-step protocol for delivering bad news: application to the patient with cancer. Oncologist. 2000;5(4):302-311. doi:10.1634/theoncologist.5-4-302. 12. Engel M, Kars MC, Teunissen SCCM, van der Heide A. Effective communication in palliative care from the perspectives of patients and relatives: a systematic review. Palliat Support Care. 2023;21(5):890-913. doi:10.1017/S1478951523001165. 13. Elo S, Kyngas H. The qualitative content analysis process. J Adv Nurs. 2008;62(1):107-115. doi:10.1111/j.1365-2648.2007.04569.x. 14. Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups. Int J Qual Health Care. 2007;19(6):349-357. doi:10.1093/intqhc/mzm042. 15. James N. Test construction of Caregiver Collusion Questionnaire. Psychol Stud. 2014;59(4):436-438. doi:10.1007/s12646-014-0273-7. 16. Pacheco Barzallo D, Schnyder A, Zanini C, Gemperli A. Gender differences in family caregiving: do female caregivers do more or undertake different tasks? BMC Health Serv Res. 2024;24:730. doi:10.1186/s12913-024-11191-w.