I NTRODUCTION
The diagnosis of cancer irrevocably alters a patient’s life and, profoundly, that of their family. Collusion, defined in healthcare as the withholding or selective sharing of information regarding a diagnosis, prognosis, or medical details from the patient by family members, usually with the implicit or explicit cooperation of healthcare professionals, is often due to a desire to protect the patient from emotional distress.[ 1 , 2 ] This practice is prevalent in cancer care but frequently overlooked.[ 3 ] Collusion is more common in patients with rapidly progressing illnesses (<6 months).[ 4 ] However, in surgical patients, especially those undergoing gastrointestinal, gynecological, or thyroid procedures, concealing the diagnosis is easier due to the prevalence of nonmalignant surgeries.
Collusion with roots in cultural norms that value family decision-making and fear of psychological suffering, frequently takes the form of a “conspiracy of silence,” which skews communication and makes patient-centered care more difficult. Collusion hinders effective communication between patients and doctors resulting in suboptimal treatment.[ 3 , 5 ] Moreover, patients unaware of diagnosis have increased anxiety and distress due to uncertainty surrounding their illness. For caregivers, the decision to withhold information can be emotionally taxing and may contribute to feelings of guilt or regret and may lead to a sense of helplessness and psychological distress.[ 4 , 5 , 6 , 7 ]
The primary objective of our study was to estimate the prevalence of collusion in cancer patients by interviewing the caregiver and the patient using a collusion questionnaire in a South Indian setting. The secondary objectives included assessment of factors influencing collusion like educational and socioeconomic status, age, gender, organ system involved, time since diagnosis, and prognosis.
M ETHODS
A cross-sectional study using a questionnaire was carried out among cancer patients receiving treatment at a tertiary care center, which was conducted after obtaining the Institutional Ethics Committee clearance (MOSC/IEC/135/2024 dated April 04, 2024) and informed consent from patients and their caregivers. The sampling method was convenient sampling and the study duration was 6 months. Adult patients aged > 18 years who were diagnosed with malignancy and undergoing medical treatment were included. Those with psychiatric illness, on antidepressants or anxiolytics or unable to communicate were excluded.
Based on the results of proportion of unwillingness to communicate about illness (75.83%) between caregivers and patients, observed in an earlier publication by Sutar et al .[ 4 ] and with 10% relative precision and 95% confidence, the minimum sample size was calculated to be 121. Therefore, we recruited 121 patients.
Qualitative interviews were conducted using a modified collusion questionnaire.[ 4 ] Patients and care givers were interviewed after taking their consent to be included in the survey. Initially we approached the bystander and interviewed them using the questionnaire [ Table 1 ]. After the interview, the bystander was asked if the patient knew about the diagnosis. If bystander said “NO” we did not approach the patient. However, if the bystander said “YES,” and said the patient knew about his/her diagnosis, we confirmed by asking only the 1 st question from the questionnaire to the patient. No further questions were asked to the patients. No additional information regarding diagnosis, disease condition, treatment, or prognosis were given to the patient or care giver. There was no mention of words like cancer, malignancy, or abnormal cells in the questionnaire. Details like educational and socioeconomic status, age, gender, organ system involved, time since diagnosis, prognosis, and surgery planned, if any, were asked to the bystander or noted from the patient case reports. No patient was informed about the diagnosis or prognosis of the disease, even if requested. After data collection, it was statistically analyzed.
Table 1
Questionnaire for caregiver interview
| Care giver interview |
| If knows answer tick Yes (Y), If doesn’t know answer tick No (N), If knows little about it tick Little (L), If not sure about answer tick Not sure (NS) |
| 1. Do you know about this disease/problem/condition of the patient? Y/N/L/NS |
| 2. If yes then what do you know about the condition? |
| a. Is it a serious problem? |
| b. Is it not a serious problem? |
| 3. What is the nature of disease? |
| a. Is it curable? |
| b. Is it not curable? |
| c. Is it partly curable? |
| 4. Do you want the patient to know about it? Y/N/NS If YES then what do you want to know? |
| a. Complete information |
| b. Little information |
| 5. Would you be able to discuss diagnosis with patient? Y/N/NS |
| 6. Would you like to inform the patient? Y/N/NS |
| 7. Would you like him/her to be informed about it by the doctor? Y/N/NS |
| 8. Do you wish patient should know course and outcome of disease? Y/N/NS |
| 9. Have you witnessed anyone suffering from this illness before? Y/N/NS |
All categorical variables were summarized as frequency and percentage and all quantitative variables were summarized using as mean and standard deviation if the data followed normality assumption, else using median and interquartile range (Q3-Q1).
KS test was used to check the normality of the data. The Chi-square test was used to determine whether the difference in the percentage of all categorical variables with collusion was statistically significant. To test the statistical significance of the difference in the mean values of all continuous variables with collusion, independent samples t -test was used. Statistical significance was defined as P < 0.05. Statistical analysis was performed using EZR software (version 1.68, Saitama Medical Center, Jichi Medical University, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing).
R ESULTS
The study was done with a study population of 121 individuals comprising of 71 females (58.7%) and 50 males (41.3%). Majority of patients belonged to the age group >60 years (59.55%, n = 72), followed by 40–60 years ( n = 40, 33%) and 7.4% patients ( n = 9) were aged <40 years. Twenty-seven (22.3%) individuals had only primary school education, while 53 (43.8%) individuals had secondary education and 41 (33.8%) graduates and postgraduates. Majority of the participants (75.2%) belonged to above poverty line category, while 24.8% of the participants were below poverty line category [ Table 2 ].
Table 2
Distribution of gender, age, educational and socioeconomic status and organ system affected with cancer
GIT=Gastrointestinal tract
| Variables | Frequency (%) |
| Gender | |
| Male | 71 (58.7) |
| Female | 50 (41.3) |
| Age (years) | |
| <40 | 9 (7.4) |
| 40–60 | 40 (33) |
| >60 | 72 (59.5) |
| Educational status | |
| Primary | 27 (22.3) |
| Secondary | 53 (43.8) |
| Graduation | 41 (33.8) |
| Socioeconomic status | |
| Below poverty line | 91 (75.2) |
| Above poverty line | 30 (24.8) |
| Organ system diagnosed with cancer | |
| Bone | 8 (6.6) |
| Brain | 1 (0.8) |
| Breast | 39 (32.2) |
| GIT | 17 (14) |
| Leukemia | 12 (9.9) |
| Lungs | 7 (5.8) |
| Multiple myeloma | 8 (6.6) |
| Ovary | 3 (2.5) |
| Prostate | 12 (9.9) |
| Thyroid | 11 (9.1) |
| Uterus | 3 (2.5) |
| Total | 121 (100) |
The distribution of diseases the patients were diagnosed with had shown that breast cancer was the commonest malignancy ( n = 39, 32.2%), followed by gastrointestinal tract cancer ( n = 17, 14%), leukemia ( n = 12, 9.9%), and prostate cancer [ n = 12, 9.9%, Table 2 ].
The prevalence of collusion was 26.4% ( n = 32), while 73.6% of participants were aware of the diagnosis ( n = 89) [ Table 3 ]. Out of the 89 patients who knew their diagnosis, 15.7% were informed by their family, 30.3% by their doctor, 47% were informed by the doctor and family together, and 1.1% of patients found out about the diagnosis by themselves [ Table 3 ].
Table 3
Prevalence of collusion and distribution of person who informed diagnosis
| Variables | Frequency (%) |
| Prevalence of collusion | |
| Yes | 32 (26.4) |
| No | 89 (73.6) |
| Total | 121 (100) |
| Person who informed diagnosis to patient in collusion positive cases | |
| Family | 14 (15.7) |
| Doctor | 27 (30.3) |
| Family and doctor | 47 (52.8) |
| Self | 1 (1.1) |
| Total | 89 (100) |
Majority of the patients (44.9%) knew the diagnosis ( n = 40) within 1 month of diagnosis, only 32.6% ( n = 29) were informed about their disease at the time of diagnosis, 17 patients (19.1%) were informed between 1 and 6 months of diagnosis, and 3 patients (3.4%) were informed after 6 months [ Figure 1 ]. A significant association was noted between collusion and time since diagnosis. The incidence of collusion decreased with increase in time from initial diagnosis. Collusion was seen in 88.9% ( n = 16) of patients with time of diagnosis < 1 month and this decreased to 54.2% ( n = 13) in 1–6 month time interval since diagnosis. In 6–12 month time frame since diagnosis only 3.4% ( n = 1) had collusion and 4% ( n = 2) in those patients with initial diagnosis made more than 1 year ago [ Figure 1 and Table 4 ].
Table 4
Association of collusion with education status of patient, prognosis according to patient, level of information caregivers want patients to know and caregiver’s perspective of own ability to discuss diagnosis with patient
NS=Not sure
| Variables | Collusion, n (%) | No collusion, n (%) | P |
| Education status of patient | |||
| Primary ( n =27) | 11(40.7) | 16 (59.3) | 0.024 |
| Secondary ( n =53) | 16 (30.2) | 37 (69.8) | |
| Graduate ( n =41) | 5 (12.2) | 36 (87.8) | |
| Prognosis according to patient | |||
| Curable | 0 | 29 (32.6) | <0.001 |
| Partly curable | 17 (53.1) | 48 (53.9) | |
| Incurable | 15 (46.9) | 12 (13.5) | |
| Level of information caregivers want patients to know | |||
| No | 25 (78.1) | 2 (2.2) | <0.001 |
| Yes (little) | 7 (21.9) | 26 (29.2) | |
| Yes (complete) | 0 | 61 (68.5) | |
| Caregiver’s perspective of own ability to discuss diagnosis with patient | |||
| No | 31 (96.9) | 16 (18) | <0.001 |
| NS | 1 (3.1) | 17 (19.1) | |
| Yes | 0 | 56 (62.9) |
There was a significant association between education and collusion ( P = 0.024). In cases of patients with primary school education, the prevalence of collusion was high at 40.7% ( n = 11), decreased to 30.2% ( n = 16) in secondary school educated patients, and in graduates and postgraduates prevalence of collusion was only 12.2% [ n = 5, Figure 2 and Table 4 ]. There was a significant association between nature of disease and collusion ( P < 0.001). When asked about the nature of the disease, in collusion positive cases 46.9% ( n = 15) of bystanders thought the disease to be incurable while 53.1% ( n = 17) thought disease to be partly curable. However, in case of absence of collusion, 32.6% ( n = 29) thought the disease to be curable, while only 13.5% ( n = 12) said disease was incurable and 53.9% ( n = 48) as partly curable [ Table 4 ]. When asked if the bystander would be able to talk about diagnosis to patient, in collusion-positive cases 96.9% ( n = 31) bystanders responded no, while in absence of collusion 62.9% ( n = 56) felt they could talk about the diagnosis. This showed a positive association [ P < 0.001, Table 4 ].
When asked if bystander wished the patient to be informed about the course and outcome of the disease, in collusion positive cases only 9.4%( n = 3) wanted patient to be informed while 3.1%( n = 1) was not sure and majority, 87.5% ( n = 28) did not want the patient to be informed. However, in case of collusion absent cases, a majority of 64% ( n = 57) wanted the patient to be informed of the course and outcome of the disease, with 16.9% ( n = 15) not wanting to inform the patient. This showed a significant association between collusion and wanting patient to know course and outcome of the disease [ P < 0.001, Table 4 ].
D ISCUSSION
The nature of collusion-prone situations varies differently based on geographic location, culture and social values.[ 6 , 8 , 9 , 10 ] The roots of collusion in South India are deeply embedded in its socio-cultural fabric. First, family-centeredness is a cornerstone of Indian society. Healthcare decisions are rarely individual endeavors; instead, they are often collective undertakings, with the family playing a dominant role in protecting and caring for their members.[ 11 , 12 ] This communal responsibility can lead to families feeling obligated to decide what is “best” for the patient, which, in the context of a severe illness like cancer, often translates to shielding them from distressing truths.
Studies indicate that collusion occurs in 71% of cases between doctors and patients, 61.3% between doctors and caregivers, and 75.83% between patients and caregivers in medically managed cancer patients in an Indian palliative care setting. However, global meta-analyses reported a pooled prevalence of 24.15% only.[ 3 ] This could be because collusion in Indian cancer care reflects a complex interplay of sociocultural, psychological, and systemic factors. In our study, the prevalence of collusion was 26.4%, and the common reasons for collusion cited by caregivers were old age of patient, mental trauma, and not knowing how to address the topic.
Several factors were identified as influencing the likelihood of collusion. Education level emerged as a key determinant in our study, with patients with lower education being more likely to experience collusion. We found that among patients with primary school education the prevalence of collusion was as high as 40.7%. This suggests that a lack of understanding or awareness of the disease may contribute to caregivers’ decisions to withhold information. Collusion was less prevalent among professionals and those with higher educational status.[ 13 ] Individuals with higher education more actively seek information than an illiterate person. However, a study in North India revealed that the majority of the illiterate and patients belonging to low socioeconomic class also felt that they needed to know about their disease.[ 14 ]
In addition, the time elapsed since diagnosis was found to be inversely related to collusion. Patients diagnosed more recently were more likely to be unaware of their condition, potentially reflecting a desire to protect them from immediate emotional distress. In our study we also made similar observations. We found that collusion was seen in 88.9% patients with time of diagnosis <1 month and this decreased to 54.2% in 1–6 months time interval since diagnosis. It has been shown that collusion is more prevalent among female patients, manual workers, patients who had not received oncological treatment, and not accompanied by a spouse.[ 11 ]
Another study had concluded that according to the patients’ preferences about being fully informed about the diagnosis, it was suggested that the disclosure of cancer diagnosis should be done by a physician and in the presence of a family member.[ 15 ] In the Schofield study, it was shown that 69% of the patients wanted to learn about the bad news and 26% sought only moderate amount of information about it.[ 16 ]
The impact of collusion extends beyond ethical considerations. A cross-sectional study in South India revealed that collusion was independently associated with poor quality of life, secondary to unresolved physical pain, anxiety, and depressive symptoms. It leads to delayed or inadequate treatment, poor adherence to treatment plans, and unrealistic expectations.[ 13 ] Collusion significantly delays or prevents timely integration of palliative care and these patients often miss out on the crucial benefits of early palliative interventions, such as symptom management, psychosocial support, and spiritual care.[ 17 ] Delivering bad news according to the patients’ preferences helps to fulfill their wishes in palliative care as well.[ 18 ] Another major impact of collusion is that the deception inherent in collusion can create internal family conflicts, tension, and communication breakdowns, especially as the patient’s condition deteriorates and the truth becomes increasingly difficult to conceal.
Addressing collusion demands multifaceted strategies. Validated assessment tools should be used to standardize identification,[ 4 , 5 ] protocols should be implemented to navigate collusion without alienating families,[ 3 ] and early psycho-oncology interventions should be initiated to reduce psychological distress to reduce anxiety and improve postdisclosure quality of life.[ 13 ] Future research must prioritize scalable interventions, such as community-based awareness programs to destigmatize cancer and promote open dialogue. Longitudinal studies tracking collusion’s impact on survival outcomes and caregiver burden in South India are also warranted.
C ONCLUSION
The prevalence of collusion in cancer patients was found to be 26.4% with a significant association with educational status, time since diagnosis, and prognosis of the disease. It was a highly prevalent among patients with primary school education and during the 1 st month after diagnosis. There was a significant association between nature of disease and collusion as the caregivers colluded even when the disease was thought to be partly curable and did not want to disclose the complete information regarding the diagnosis, course and outcome to the patient.
Conflicts of interest
There are no conflicts of interest.
Funding Statement
Nil.
R EFERENCES
1. Chaturvedi SK, Loiselle CG, Chandra PS. Communication with relatives and collusion in palliative care: A cross-cultural perspective. Indian J Palliat Care. 2009;15:2–9. doi: 10.4103/0973-1075.53485.
2. Low JA, Kiow SL, Main N, Luan KK, Sun PW, Lim M. Reducing collusion between family members and clinicians of patients referred to the palliative care team. Perm J. 2009;13:11–5. doi: 10.7812/tpp/09-058.
3. Sutar R, Chaudhary P, Yadav V. Prevalence of collusion in cancer communications: A meta-analysis. Psychooncology. 2022;31:372–87. doi: 10.1002/pon.5824.
4. 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;25:242–9. doi: 10.4103/IJPC.IJPC_146_18.
5. Mathew B, Mohanti BK, Tewari S, Kabra V, Gulia P, Bajpai P, et al. Collusion: The facade and its implications on total pain management in palliative care. Indian J Palliat Care. 2021;27:176–9. doi: 10.4103/IJPC.IJPC_81_20.
6. Biswas J, Khanam SA, Tauhid MS, Sarker SR, Afsar N, Islam N. Situations in which caregivers and patients are likely to collude: Perspectives from caregivers of advanced cancer patients in Bangladesh. Palliat Support Care. 2025;23:e68. doi: 10.1017/S1478951524002074.
7. Mohamed Hussin NA, Mohd Sabri NS. A qualitative exploration of the dynamics of guilt experience in family cancer caregivers. Support Care Cancer. 2023;31:659. doi: 10.1007/s00520-023-08060-3.
8. Siraj S, Hens K, Ali Y. Disclosure of true medical information: The case of Bangladesh. BMC Med Ethics. 2024;25:112. doi: 10.1186/s12910-024-01115-y.
9. Ma Q, Wu Y, Fang R. Truth-telling, and ethical considerations in terminal care: An Eastern perspective. Nurs Ethics. 2025;32:971–9. doi: 10.1177/09697330241312376.
10. Borgan SM, Amarin JZ, Othman AK, Suradi HH, Qwaider YZ. Truth disclosure practices of physicians in Jordan. J Bioeth Inq. 2018;15:81–7. doi: 10.1007/s11673-018-9837-x.
11. 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;92:659–62. doi: 10.1136/postgradmedj-2015-133850.
12. Ramya N, Sunderraj RD, Jeba J, Jacob A. Family involvement in healthcare decision-making for cancer patients in India. J Palliative Care. 2017;33:211–7.
13. 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;24:61–6. doi: 10.4103/IJPC.IJPC_111_17.
14. Emanuel EJ, Fairclough DL, Wolfe P, Emanuel LL. Talking with terminally ill patients and their caregivers about death, dying, and bereavement: Is it stressful?Is it helpful?Arch Intern Med. 2004;164:1999–2004. doi: 10.1001/archinte.164.18.1999.
15. Arbabi M, Rozdar A, Taher M, Shirzad M, Arjmand M, Ansari S, et al. Patients'preference to hear cancer diagnosis. Iran J Psychiatry. 2014;9:8–13.
16. Schofield PE, Beeney LJ, Thompson JF, Butow PN, Tattersall MH, Dunn SM. Hearing the bad news of a cancer diagnosis: The Australian melanoma patient's perspective. Ann Oncol. 2001;12:365–71. doi: 10.1023/a:1011100524076.
17. Beng KS. Collusion in palliative care. Malays Fam Physician. 2006;1:62–4.
18. Ayalew EA, Mphuthi DD, Matlhaba KL. Patients'preferences for delivering bad news in palliative care in Ethiopia: A qualitative study. BMC Palliat Care. 2023;22:170. doi: 10.1186/s12904-023-01275-5.