Abstract
Objective
Head and neck cancer (HNC) patients may be subject to decisional conflict (DC). However, research on DC in the palliative phase remains limited. The primary objective of this study was to assess and describe DC among patients with HNC in the palliative phase and explore the association between clinically significant DC and quality of life (QoL).
Study design
Prospective cohort study.
Setting
Tertiary cancer center.
Methods
Patients with HNC entering the palliative phase between September 2022 and March 2024 were eligible. Patients were assessed in 2 cohorts: incurable HNC and refraining from curative treatment. DC was measured using the Decisional Conflict Scale (DCS). Other measures included patient, tumor, and palliative characteristics, and the EORTC QLQ‐C15‐PAL for QoL.
Results
Of 104 included patients, 72 patients had incurable HNC and 32 refrained from curative treatment. Clinically significant DC (score ≥25) was observed in 47 incurable HNC patients (65.3%) and 14 patients who refrained from curative treatment (43.8%), which was a significant difference. Tumor stage I/II was significantly predictive of less DC in patients who refrained from curative treatment. Three QoL domains showed large differences for incurable HNC patients when differentiating for clinically significant DC, and 6 for those who refrained from treatment.
Conclusion
This study provides valuable insights into DC among palliative HNC patients. Clinically significant DC was more common in patients with incurable HNC than those who refrained from curative treatment. These findings highlight key differences between the cohorts, underscoring the need for tailored counselling to support patient‐centered care.
Keywords: decisional conflict, head and neck cancer, palliative phase, patient‐centered care
Head and neck cancer (HNC) causes over 850.000 new cases and 400.000 deaths annually. 1 Prognosis remains poor, with nearly half of patients dying from the disease. 2 , 3 Patients face cancer‐specific issues—such as airway and swallowing problems or disfigurement—that contribute to their emotional and psychological burden, 3 , 4 , 5 , 6 , 7 and involve decision‐making with delicate trade‐offs between quality of life (QoL) and survival. 8 , 9 Most patients prefer to be involved in decision‐making, 10 , 11 , 12 and given the limited survival in the palliative phase, early engagement in this process is essential.
Over the past decades, shared decision making (SDM) and value‐based healthcare have been gaining recognition as essential elements of optimal care. This on‐going process remains particularly demanding in HNC care due to its unique problems. 13 , 14 To enable SDM, patients should be optimally informed and their preferences should be considered before making a final decision. 15 Decision‐making can be challenging for both doctors and patients, as it involves balancing foundational ethical principles—such as patient autonomy and nonmaleficence—with practical and systemic considerations, including treatment availability, clinical guidelines and time constraints. These complexities can introduce uncertainty, which might lead to decisional conflict (DC) in patients. 9 , 16 , 17 DC refers to uncertainty for patients in decision‐making when facing difficult choices. 17 , 18 , 19 High levels of DC can result in delay in the decision‐making process, choices misaligned with patient preferences, and experiencing decisional regret. 16 , 19 , 20 , 21 , 22 Given the complexity of decision‐making in HNC and multiple factors that might contribute to heightened DC in this population, DC is imminent. 6 , 23 , 24
Although some studies have demonstrated a high prevalence of DC among patients with HNC, 16 , 25 , 26 there is limited research focusing on decision‐making and DC in the palliative phase of HNC. 26 , 27 This phase involves delicate trade‐offs and complex choices, such as life‐prolonging treatment, end‐of‐life care, and options like euthanasia, that can influence end‐of‐life experiences for patients and their families. 28 Therefore, this study aims to assess DC and explore the association between clinically significant DC and QoL among patients in the palliative phase. Insight into DC in this phase is crucial, as care in the palliative phase aims to maintain as long as possible the highest possible QoL, while prioritizing patients' autonomy and personal preferences in decision‐making.
Methods
Ethical approval was granted by the Erasmus Medical Center ethics committee (MEC‐2022‐0473). Follow‐up time was specified as the date of the palliative trajectory's onset (ie, the consultation confirming curative treatment was no longer viable, or patients' declaration of treatment refrainment) until death or October 31, 2024, whichever came first.
Settings and Participants
This prospective single‐center cohort study included patients between September 2022 and March 2024. All study participants gave informed consent. Power calculation was done for a two‐cohort clinical study, of which the present study is the first. Details can be found in the published study protocol. 29 A minimum of 79 patients per cohort was required for statistical significance. Eligible patients were ≥18 years old, diagnosed with HNC at any tumor site, and in a palliative trajectory. Patients were recruited at the outpatient clinic following the consultation discussing the palliative trajectory. Exclusion criteria were cognitive incompetence, illiteracy, or insufficient Dutch proficiency. Included patients were divided into 2 cohorts based on the reason for the palliative trajectory: incurable HNC or refraining from curative treatment, as this reason may influence the levels of experienced DC. In our institution, both patients with an incurable tumor and those refraining from treatment may have options for palliative antitumor treatments (ie, palliative chemo‐ or radiotherapy), depending on factors like the patient's condition and tumor type. These end‐of‐life decisions, among others such as euthanasia, palliative sedation, pain medication, and arranging personal matters, are contemplated during the palliative phase. We hypothesized that patients who refrained from curative treatment are more certain about end‐of‐life decisions and would therefore experience lower DC than patients with an incurable tumor.
Patient, Tumor, and Palliative Characteristics
Baseline characteristics included Adult Comorbidity Evaluation‐27 (ACE‐27) score, age, sex, weight loss (yes/no), marital status, WHO performance status, tumor location and TNM stage, and tumor chronology (primary/recurrence). TNM staging followed the American Joint Committee on Cancer Staging Manual, eighth edition classification of malignant tumors. 30 Follow‐up data included place and cause of death (including euthanasia and palliative sedation), and palliative treatments such as radiotherapy, systemic therapy, and best supportive care (BSC, ie, no antitumor treatment, but comprehensive care addressing patients' physical, emotional, and psychosocial needs 31 ).
DC
DC reflects patients' uncertainty regarding their decision‐making and was assessed using the Decisional Conflict Scale (DCS) questionnaire, 32 2 weeks after the onset of the palliative trajectory. The DCS is a validated 16‐item 5‐point Likert scale measurement, encompassing 5 subscales: (1) feeling informed, (2) decisional uncertainty, (3) clear values, (4) support, and (5) quality of decisions. Calculated scores range from 0 (no DC) to 100 (extremely high DC); a score <25 is associated with implementing decisions, while a score >37.5 indicates decision delay and uncertainty. 33 A score of ≥25 is commonly used as a cutoff for clinically significant DC. 20 , 34 , 35 The DCS questionnaire can be found in Supplement 1, available online.
QoL
QoL was assessed using the European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire Core 15 Palliative Care (QLQ‐C15‐Pal), a 15‐item version of the EORTC QLQ‐C30 tailored to palliative cancer patients. 36 It addresses physical and emotional functioning, fatigue, pain, dyspnea, insomnia, appetite loss, constipation, nausea, and global health. At our institution the QLQ‐C15‐Pal is routinely collected during the palliative trajectory.
Statistical Analyses
Statistical analyses were performed using IBM SPSS, version 28. Descriptive statistics were used to analyze demographic variables and tumor characteristics. Univariable binary logistic regression analyses for DC scores were conducted (DCS <25 or ≥25). Differences in the proportion of clinically significant DC between groups were calculated using Pearson's Chi‐square test. Means and medians were compared via t‐test or Mann‐Whitney U, respectively. A P < .05 was considered statistically significant. Absolute differences between medians for QoL domains were reported as effect size metric with 95% confidence interval (CI). Absolute differences were considered “large difference” when there was an effect size of ≥8.5. Bootstrapping was performed in RStudio (version 4.2.3; R Foundation) to estimate a more robust 95% CI through data resampling.
Results
Figure 1 depicts the patient inclusion flowchart. Of 197 eligible HNC patients in the palliative phase, 190 patients were approached to participate and 104 patients participated. Forty‐three patients declined participation, 39 patients did not complete the questionnaire and 4 patients were excluded after completing the questionnaire. Of these 4, 1 ultimately did not enter a palliative trajectory, and the other 3 patients had initially given verbal consent, but did not confirm consent when completing the electronic questionnaire.
Figure 1.
Flowchart of inclusion of patients. Figure 1 illustrates the patient inclusion process for the study. Of 197 eligible patients, 104 patients were included in the study.
DCS
Table 1 shows DCS total and subscale scores. Overall, 61 patients (58.7%) experienced clinically significant DC (≥25). Significantly more patients experienced DC in the incurable HNC group (65.3%) than those who refrained from treatment (43.8%, P = .040). The median DCS in the incurable HNC group was 28.1, compared to 18.8 in the treatment refrainment group (P = .026). The highest scoring subscales in the incurable HNC group were “Uncertainty” and “Clarity” (median 41.7, IQR 25.0–58.3, i.e., “I am clear about the best choice for me” and i.e., “I am clear about which risks and side effects matter most,” respectively). Among patients who refrained from treatment, the highest scoring subscale was “Uncertainty” (median 33.3, IQR 10.4–58.3). Both groups scored lowest for the subscale “Effective” (i.e., “I am satisfied with my decision”), with a median score of 18.8 in patients with incurable HNC (IQR 6.3–37.5) and a score of 0 (IQR 0.0–17.2) in patients who refrained from treatment. Significant differences between the two groups were found for the medians of the subscales “Clarity” (P = .033) and “Effective” (P = <.001).
Table 1.
Decisional Conflict Scale Including the Five Subscales, Score of Incurable Patients and Patients Who Refrained From Treatment
| Decisional Conflict Scale | Total score | Informed | Uncertainty | Clarity | Support | Effective |
|---|---|---|---|---|---|---|
| Incurable (N = 72) | ||||||
| Median (IQR) | 28.1 (18.8‐44.9) | 25.0 (16.7‐50.0) | 41.7 (25.0‐58.3) | 41.7 (25.0‐58.3) | 25.0 (10.4‐33.3) | 18.8 (6.3‐37.5) |
| Categories, No. (%) | ||||||
| Low <25 | 25 (34.7) | 24 (33.3) | 13 (18.1) | 14 (19.4) | 33 (45.8) | 38 (52.8) |
| Medium 25‐37.5 | 23 (31.9) | 19 (26.4) | 18 (25.0) | 20 (27.8) | 22 (30.6) | 17 (23.6) |
| High >37.5 | 24 (33.3) | 29 (40.3) | 41 (56.9) | 38 (52.8) | 17 (23.6) | 17 (23.6) |
| Treatment refrainment (N = 32) | ||||||
| Median (IQR) | 18.8 (12.9‐31.3) | 20.8 (10.4‐39.6) | 33.3 (10.4‐58.3) | 29.2 (16.7‐41.7) | 16.7 (2.1‐33.3) | 0.0 (0.0‐17.2) |
| Categories, No. (%) | ||||||
| Low <25 | 18 (56.3) | 16 (50.0) | 10 (31.3) | 9 (28.1) | 17 (53.1) | 25 (78.1) |
| Medium 25‐37.5 | 9 (28.1) | 8 (25.0) | 7 (21.9) | 14 (43.8) | 9 (28.1) | 5 (15.6) |
| High >37.5 | 5 (15.6) | 8 (25.0) | 15 (46.9) | 9 (28.1) | 6 (18.8) | 2 (6.3) |
Baseline Characteristics and Univariable Analysis
Patient and tumor characteristics, and odds ratios for clinically significant DC are shown per group in Table 2. Of the total cohort, 74 patients were male (71.2%) and the mean age was 71.7 years (SD 13.0). There were no significant differences in sex or age between patients who were incurable (n = 72, 76.4% male, mean age 70.2) and patients who refrained from treatment (n = 32, 59.4% male, mean age 75.1). Multiple tumor sites were identified, of which the oral cavity was the most common in both groups (26.4% and 46.9%). Odds ratios were not calculated for the variable “tumor location” due to small subgroup sizes. No baseline predictors for DC were found in the incurable group, but stage I/II tumors predicted less DC in the group that refrained from treatment (0.127, 95% CI 0.018–0.905).
Table 2.
Baseline Characteristics and Binary Logistic Regression of Clinically Significant DCS Scores of Incurable Patients and Patients Who Refrained From Treatment
| Incurable | Treatment refrainment | |||
|---|---|---|---|---|
| Characteristics | Frequency (%) | Univariable OR on DCS ≥ 25 (95 CI) | Frequency (%) | Univariable OR on DCS ≥ 25 (95 CI) |
| No. of patients | 72 | NA | 32 | NA |
| Age, mean in years (SD) | 70.2 (13.5) | 1.009 (0.973‐1.045) | 75.1 (11.1) | 0.966 (0.903‐1.033) |
| Sex | ||||
| Male | 55 (76.4) | 1 [Reference] | 19 (59.4) | 1 [Reference] |
| Female | 17 (23.6) | 0.968 (0.310‐3.024) | 13 (40.6) | 1.179 (0.285‐4.879) |
| Tumor location | ||||
| Glottic | 6 (8.3) | NA | 1 (3.1) | NA |
| Supraglottic | 4 (5.6) | NA | 2 (6.3) | NA |
| Oropharynx | 16 (22.2) | NA | 5 (15.6) | NA |
| Oral cavity | 19 (26.4) | NA | 15 (46.9) | NA |
| Hypopharynx | 7 (9.7) | NA | 4 (12.5) | NA |
| Nasopharynx | 2 (2.8) | NA | 1 (3.1) | NA |
| Salivary gland | 3 (4.2) | NA | 1 (3.1) | NA |
| Sinonasal | 6 (8.3) | NA | 2 (6.3) | NA |
| Skin, head & neck region | 4 (5.6) | NA | 1 (3.1) | NA |
| Lip | 1 (1.4) | NA | NA | NA |
| Unknown primary | 3 (4.2) | NA | NA | NA |
| Other | 1 (1.4) | NA | NA | NA |
| Tumor chronology | ||||
| Frist primary | 22 (30.6) | 1 [Reference] | 18 (56.3) | 1 [Reference] |
| Second or later primary | 3 (4.2) | 0.233 (0.018‐3.026) | 7 (21.9) | 0.167 (0.017‐1.679) |
| Recurrent (first or second recurrence) | 41 (56.9) | 0.809 (0.269‐2.429) | 6 (18.8) | 2.000 (0.290‐13.814) |
| Residual | 6 (8.3) | 2.333 (0.228‐23.908) | 1 (3.1) | 0.000 (0.000) |
| Tumor stage | ||||
| I + II | 2 (2.8) | 0.564 (0.034‐9.470) | 9 (28.1) | 0.127 (0.018‐0.905) |
| III | 7 (9.7) | 1.410 (0.252‐7.884) | 9 (28.1) | 0.0222 (0.036‐1.370) |
| IV | 61 (84.7) | 1 [Reference] | 13 (40.6) | 1 [Reference] |
| Missing | 2 (2.8) | NA | 1 (3.1) | NA |
| ACE‐27 | ||||
| 0 (none) | 5 (6.9) | 1 [Reference] | 2 (6.3) | 1 [Reference] |
| 1 (mild) | 10 (13.9) | 11 (34.4) | ||
| 2 (moderate) | 36 (50.0) | 0.925 (0,278‐3.082) | 16 (50.0) | 1.440 (0.343‐6.048) |
| 3 (severe) | 21 (29.2) | 3 (9.4) | ||
| WHO‐status | ||||
| 0 | 13 (18.1) | 1 [Reference] | 8 (25.0) | 1 [Reference] |
| I | 33 (45.8) | 18 (56.3) | ||
| II | 20 (27.8) | 0.597 (0.220‐1.622) | 3 (9.4) | 1.364 (0.230‐8.081) |
| III | 6 (8.3) | 3 (9.4) | ||
| Smoking | ||||
| Current/former | 56 (77.8) | 1 [Reference] | 22 (68.7) | 1 [Reference] |
| No | 15 (20.8) | 1.528 (0.430‐5.430) | 10 (31.3) | 0.800 (0.175‐3.651) |
| Missing | 1 (1.4) | NA | NA | NA |
| Alcohol | ||||
| Yes/former | 60 (83.3) | 1 [Reference] | 24 (75.0) | 1 [Reference] |
| No | 11 (15.3) | 0.556 (0.151‐2.052) | 7 (21.9) | 0.473 (0.076‐2.935) |
| Missing | 1 (1.4) | NA | 1 (3.1) | NA |
| Prior HNC tumor | ||||
| Yes | 48 (66.7) | 0.912 (0.324‐2.565) | 13 (40.6) | 0.694 (0.165‐2.917) |
| No | 24 (33.3) | 1 [Reference] | 19 (59.4) | 1 [Reference] |
| Education level | ||||
| Lower | 28 (38.9) | 1 [Reference] | 13 (40.6) | 1 [Reference] |
| Intermediate | 15 (20.8) | 2.222 (0.505‐9.788) | 3 (9.4) | 0.222 (0.015‐3.221) |
| Tertiary | 18 (25.0) | 0.873 (0.257‐2.966) | 6 (18.8) | 0.089 (0.008‐1.029) |
| Missing | 11 (15.3) | NA | 10 (31.3) | NA |
| Employment | ||||
| Retired | 50 (69.4) | 1 [Reference] | 27 (84.4) | 1 [Reference] |
| Yes | 16 (22.2) | 0.859 (0.267‐2.764) | 3 (9.4) | 0.625 (0.050‐7.749) |
| No | 6 (8.3) | 1.030 (0.171‐6.204) | 2 (6.3) | 1.250 (0.071‐22.132) |
| Marital status | ||||
| Partner or married | 42 (58.3) | 1 [Reference] | 17 (53.1) | 1 [Reference] |
| Widow | 12 (16.7) | 1.846 (0.434‐7.850) | 8 (25.0) | 0.675 (0.121‐3.767) |
| Single | 18 (25.0) | 1.231 (0.385‐3.930) | 7 (21.9) | 0.844 (0.143‐4.974) |
Bold values represent significant effect sizes.
Abbreviations: CI, confidence interval; DCS, Decisional Conflict Scale; OR, odds ratio.
DC and QoL
In total, 99 patients completed both the DCS and the QLQ‐C15‐Pal (70 patients with incurable HNC, 29 patients who refrained from curative treatment). Median time to DCS completion was 21 days (IQR 15‐36). Median scores per domain for patients who experienced DC (≥25) and those who did not (<25) and effect size with 95% CI are shown in Table 3. In the incurable HNC group, large differences between patients who experienced DC and those who did not were found for the following domains: physical functioning, pain, and constipation. For the patients who refrained from treatment, domains with large differences in median score when differentiating for DC were global health status, physical functioning, emotional functioning, fatigue, insomnia, and appetite loss.
Table 3.
Clinically Significant Decisional Conflict Versus Quality of Life Domain of EORTC‐QLQ‐C15‐PAL of Incurable Patients and Patients Who Refrained From Treatment
| EORTC‐PAL15 score, median (Q1‐Q3) | ||||
|---|---|---|---|---|
| Variable | Clinical cutoff | Total DCS < 25 | Total DCS ≥ 25 | Effect size (95% CI) |
| Incurable (N = 70) | ||||
| No. of patients (%) | NA | 24 (34.3) | 46 (65.7) | NA |
| Global health status | ≥64.5 | 66.7 (50.0‐83.3) | 66.7 (66.7‐83.3) | 0 (−25.0 to 16.7) |
| Physical functioning | ≥81.2 | 73.3 (50.0‐93.3) | 93.3 (60.0‐93.3) | −20 (−26.7 to 20.0) |
| Emotional functioning | ≥72.5 | 75.0 (43.8‐100.0) | 66.7 (50.0‐87.5) | 8.3 (−16.7 to 16.6) |
| Fatigue | ≤26.9 | 33.3 (22.2‐44.6) | 33.3 (0.0‐33.3) | 0 (0.0‐11.1) |
| Nausea and vomiting | ≤5.3 | 0.0 (0.0‐0.0) | 0.0 (0.0‐0.0) | 0 (0.0‐0.0) |
| Pain | ≤23.2 | 33.3 (16.7‐50.0) | 16.7 (0.0‐33.3) | 16.7 (0.0‐33.3) |
| Dyspnea | ≤18.2 | 0.0 (0.0‐33.3) | 0.0 (0.0‐8.3) | 0 (0.0‐33.3) |
| Insomnia | ≤27.3 | 33.3 (0.0‐33.3) | 33.3 (0.0‐33.3) | 0 (−33.3 to 33.3) |
| Appetite loss | ≤17.7 | 0.0 (0.0‐33.3) | 0.0 (0.0‐33.3) | 0 (0.0‐33.3) |
| Constipation | ≤11.1 | 33.3 (0‐33.3) | 0.0 (0.0‐33.3) | 33.3 (0.0‐33.3) |
| Treatment refrainment (N = 29) | ||||
| No. of patients (%) | NA | 17 (58.6) | 12 (41.4) | NA |
| Global health status | ≥64.5 | 83.3 (66.7‐83.3) | 66.7 (50.0‐66.7) | 16.7 (0.0‐33.3) |
| Physical functioning | ≥81.2 | 93.3 (53.4‐93.3) | 73.3 (60.0‐93.3) | 20 (−20.0 to 33.3) |
| Emotional functioning | ≥72.5 | 100.0 (66.7‐100.0) | 66.7 (41.7‐83.3) | 33.3 (−8.3 to 58.3) |
| Fatigue | ≤26.9 | 33.3 (0.0‐50.0) | 44.5 (22.2‐66.7) | −11.2 (−44.5 to 11.1) |
| Nausea and vomiting | ≤5.3 | 0.0 (0.0‐0.0) | 0.0 (0.0‐0.0) | 0 (0.0‐0.0) |
| Pain | ≤23.2 | 16.7 (0.0‐66.7) | 25.0 (0.0‐62.5) | −8.3 (−41.7 to 50.0) |
| Dyspnea | ≤18.2 | 0.0 (0.0‐33.3) | 0.0 (0.0‐33.3) | 0 (−33.3 to 0.0) |
| Insomnia | ≤27.3 | 33.3 (0.0‐50.0) | 0.0 (0.0‐58.3) | 33.3 (−33.3 to 33.3) |
| Appetite loss | ≤17.7 | 0.0 (0.0‐33.3) | 16.7 (0.0‐58.3) | −16.7 (−50.0 to 33.3) |
| Constipation | ≤11.1 | 0.0 (0.0‐33.3) | 0.0 (0.0‐0.0) | 0 (0.0‐33.3) |
Bold values represent significant effect sizes.
Palliative Characteristics
Table 4 shows the palliative characteristics per group. Acceptance of the proposed treatment plan from multidisciplinary meetings was similar (~80%) in both groups. In the incurable HNC group, most received BSC (44.4%), followed by systemic treatment (29.2%) and radiotherapy (19.4%). Among patients who refrained from curative treatment, the majority received BSC (81.3%), while the remaining 18.8% underwent palliative radiotherapy. Patients who refrained from curative treatment significantly more often received BSC than patients with incurable HNC (P = .003). Furthermore, patients who refrained from curative treatment died significantly more often due to euthanasia, whereas patients with incurable HNC died significantly more often due to palliative sedation (P = .008). In both groups, most patients died at home (incurable HNC 45.8%, treatment refrainment 59.4%).
Table 4.
Palliative Characteristics of Incurable Patients and Patients Who Refrained From Treatment
| Incurable (N = 72) | Treatment refrainment (N = 32) | ||||
|---|---|---|---|---|---|
| Characteristics | Frequency (%) | Missing (%) | Frequency (%) | Missing (%) | P‐value |
| Accepted palliative treatment proposal? | 3 (4.2) | NA | .592 | ||
| Yes | 57 (79.2) | 25 (78.1) | |||
| No | 12 (16.7) | 7 (21.9) | |||
| Received palliative treatments | 1 (1.4) | NA | .003 | ||
| Radiotherapy | 14 (19.4) | 6 (18.8) | |||
| Systematic | 21 (29.2) | NA | |||
| Systematic + radiotherapy | 3 (4.2) | NA | |||
| Systematic + surgery | 1 (1.4) | NA | |||
| Best supportive care | 32 (44.4) | 26 (81.3) | |||
| Deceased | 1 (1.4) | NA | .558 | ||
| Yes, due to HNC | 47 (65.3) | 25 (78.1) | |||
| Yes, other cause | 2 (2.8) | 1 (3.1) | |||
| Yes, unknown cause | 5 (6.9) | NA | |||
| No | 17 (23.6) | 6 (18.8) | |||
| Way of death | 9 (12.5) | 3 (9.4) | .008 | ||
| Euthanasia | 6 (8.3) | 11 (34.4) | |||
| Palliative sedation | 27 (37.5) | 6 (18.8) | |||
| Neither/other way | 13 (18.1) | 6 (18.8) | |||
| Not applicable | 17 (23.6) | 6 (18.8) | |||
| Place of death | 5 (6.9) | 1 (3.1) | .628 | ||
| At home | 33 (45.8) | 19 (59.4) | |||
| Hospice | 10 (13.9) | 3 (9.4) | |||
| Hospital | 5 (6.9) | 1 (3.1) | |||
| Nursing home | 2 (2.8) | 2 (6.3) | |||
| Not applicable | 17 (23.6) | 6 (18.8) | |||
Bold values represent significant effect sizes.
Discussion
This study assessed DC in HNC patients in the palliative phase, describing initial insights and highlighting differences between 2 distinct patient cohorts. Patients who refrained from curative treatment experienced significantly less DC (43.8%) compared to those who had no curative options (65.3%). Patients with incurable cancer experienced clinically significant DC in all domains, except in the “effective” subscale of the DCS, which scored lowest in both groups.
DC
Patients who refrained from curative treatment had a median DCS < 25, indicating no DC, while those with incurable HNC scored ≥25, reflecting clinically significant DC. This finding highlights a key difference in characteristics between the two groups, namely that individuals who autonomously decline treatment appear to exhibit greater decisional certainty. Contrarily, patients who report more DC may experience more anxiety, stress, and decisional regret, 26 , 37 potentially leading to less optimal outcomes. Therefore, these patients could benefit from more personalized counselling, and psychosocial and decisional support. 37 These should be important considerations for clinicians when engaging in end‐of‐life discussions and shared decision‐making with HNC patients.
This study reports higher levels of DC for the group incurable HNC patients than previous literature on DC in HNC patients. 25 , 26 , 34 In these previous studies, patients were treated with curative intent, which could likely involve decisions associated with lower levels of DC. Moreover, patients who have not yet accepted their palliative trajectory may struggle with end‐of‐life choices, resulting in more DC during the decision‐making process. Patients with incurable HNC experienced clinically significant DC in all domains expect for the “effective” subscale, which scored lowest for both groups. This suggests that both groups felt they had made an informed decision with which they were satisfied, that reflected what was important to them and that they expect to adhere to. 38 Only tumor stage I/II was predictive of lower DC in the treatment‐refraining group, indicating that even with favorable prognosis, these patients had a high degree of certainty and clear preferences. No other predictors of clinically significant DC were found in this study. To our knowledge, this is the first study to report on predictors of DC specifically in the palliative phase. While several studies have explored predictors of DC in the curative phase, the findings are inconsistent. Two studies found no significant predictive correlation between patient characteristics, such as age, gender, and educational level, and DC. 25 , 39 In contrast, another study reported that older patients (>65 years) and those with lower educational levels experienced higher levels of DC, 40 and the latter finding is also supported by two additional studies. 41 , 42 However, some contradictory evidence suggests that younger age may contribute to higher DC. 41 The same study identified lower income as a predictor of DC.
QoL Domains
In this study, more QoL domains showed large effect size differences when differentiating for clinically significant DC in the treatment‐refraining group than in the incurable HNC group (6 vs 3). Interestingly, patients who refrained from treatment scored better on 5 out of 6 QoL domains with large differences (global health status, physical functioning, emotional functioning, fatigue and appetite loss) when they experienced no DC (<25), whereas incurable patients scored better on all 3 QoL domains with large differences (physical functioning, pain, and constipation) when they did experience DC (≥25). In other words, patients with no curable options were more prone to experience DC when they had higher self‐reported QoL in the beforementioned 3 domains. This suggests that patients who experience higher QoL and are unwillingly in a palliative trajectory may find it challenging to make difficult end‐of‐life decisions, leading to DC. This is in concordance with another study that found a positive relation between DC and QoL in cancer patients, indicating higher QoL leads to more DC. 43 Conversely, patients who refrained from curative treatment experienced more DC when they had lower self‐reported QoL in 5 domains, which is supported by a study of Van Lent. 44 This outcome suggests that patients with a lower QoL may feel that they have fewer options for improving their QoL, leading to feelings of uncertainty about decisions, as the possibility of further deterioration of their QoL may not weigh up to the chance of benefit. Contrarily, when patients experience better QoL this may lead to more certainty about their decisions of declining treatment, as treatments can lead to side‐effects and deteriorated QoL. One example from our study is the “emotional functioning” domain among patients who refrained from curative treatment. Within this group, patients experiencing no DC had a perfect score of 100 on this domain, whereas those with high DC scored significantly lower (66.7/100). This suggests that better emotional functioning is associated with less DC. A similar but smaller difference was observed in patients with incurable tumors (75.0 vs 67.7), likely reflecting greater psychological burden when no curative options exist.
Palliative Characteristics
The differences between the two cohorts are further observed in the palliative characteristics. A significantly smaller proportion of patients who refrained from curative treatment opted for palliative antitumor therapy (18.8%) compared to those with an incurable tumor (55.6%), suggesting a preference for minimal medical intervention in the treatment‐refrainment group. Additionally, this group more frequently chose euthanasia as their preferred way of dying (34.4% vs 8.3%), suggesting that having this option may influence the decision to forgo curative treatment. These findings align with the lower degree of DC observed for this group, indicating that patients who refrained from curative treatment seem to make more pronounced decisions regarding end‐of‐life choices.
Clinical Implications and Future Research
Our findings suggest that HNC patients who refrained from curative treatment—especially those with early‐stage tumors—tend to be more certain in their decisions. In contrast, patients with incurable tumors may benefit from enhanced psychosocial support or individualized prognostic counseling, given their higher levels of DC. To support this need, our team developed and validated OncologIQ Palliative, a prognostic model that provides personalized survival estimates for HNC patients in the palliative phase. 29 , 45 , 46 Our hypothesis is that sharing personalized prognostic information on life expectancy with HNC patients in the palliative phase will enhance patient empowerment throughout the decision‐making process from the palliative diagnosis till death. Specifically, we hypothesize that sharing individualized prognostic information will reduce DC, increase satisfaction with the decision‐making process, lead to fewer treatments during the palliative phase, and improve satisfaction with advanced care planning, including end‐of‐life discussions about preferred place of death and wishes regarding palliative sedation and euthanasia. We also hypothesize that this approach may contribute to longer overall survival. To evaluate these potential effects of the model OncologIQ Palliative during palliative consultations, we are conducting a clinical study, 29 in which we are measuring DC in a second cohort of patients who have received individualized prognostic counseling with OncologIQ Palliative. Additionally, we will conduct interviews with these patients to gain deeper insights into their experiences with this counseling approach. These initiatives are part of our broader efforts to enhance specialized palliative and oncological care and to empower patients in the palliative phase, particularly in our Expert Center of Palliative Care for HNC patients. Ultimately, we hope these findings will benefit other patients in palliative care settings and contribute to broader clinical practice.
Furthermore, in our Expert Center of Palliative Care for HNC patients, patients are offered hybrid remote palliative care, in which electronic patient‐reported outcomes (ePROs) enable patients to remain in their home environment in the palliative phase while being remotely monitored. 7 This unique type of care for HNC patients in the palliative phase is, to our knowledge, the first of its kind that aligns with the latest recommendations of the SONCOS (“Stichting Oncologische Samenwerking,” Dutch Foundation for Oncological Collaboration) standards and guidelines report of 2025, 47 which advises periodic and systematic symptom monitoring for expert centers in palliative care in The Netherlands. As a frontrunner in this field and with the growing body of evidence for its benefits, we hope to standardize ePRO‐guided palliative care for HNC patients in the Netherlands. A recent study showed high patient satisfaction with most aspects of this type of care, 7 though one third considered the provided psychosocial support to be insufficient. To address this, we are enhancing our remote care pathway by exploring compassionate artificial intelligence solutions, guided by an extensive literature review and cocreation sessions with healthcare professionals and patients. This user‐centered approach increases the likelihood of successful implementation. 48 , 49
Strengths and Limitations
To our knowledge, this is the first study to assess DC in palliative HNC patients, offering valuable first insights into how HNC patients experience decision‐making in the critical palliative stage of their illness. Prospective data collection has ensured a high level of data reliability. A median time of 21 days between the palliative consultation and DCS completion significantly reduces the risk of recall bias in this study. A limitation of this study is the relatively small sample size, as only palliative patients were eligible. Given the sensitivity of this phase in the lives of patients and their families, patients may be more inclined to decline participation in a clinical study. Ultimately, 104 patients were included, with 32 patients in the refrainment group and 72 in the group with incurable HNC. Despite the unequal group sizes, both groups were sufficiently large to detect statistically significant differences.
Conclusions
In this cohort study of patients with HNC in the palliative phase, more than half of patients (58.6%) experienced clinically significant DC, with significantly higher rates among those with incurable HNC (65.3%) compared to those who refrained from curative treatment (43.8%). Several QoL domains were linked to clinically significant DC. Additionally, multiple palliative characteristics—such as choosing for palliative anti‐tumor therapy or euthanasia—were significantly different between the two groups. Our findings suggest two distinct patient cohorts based on the reason for entering the palliative phase. This insight may guide the development of personalized interventions and providing more patient‐centered care, aiming to diminish DC in this vulnerable group of patients.
Author Contributions
Boyd N. van den Besselaar and Aimée F. Herkendaal collected and analyzed the data and drafted the main manuscript; Marinella P. J. Offerman and Aniel Sewnaik, coordinated the study and helped interpreting the results and revised the main manuscript and supervised; Robert J. Baatenburg de Jong, revised the manuscript and supervised.
Disclosures
Competing interests
None.
Funding source
We received full funding from the KWF Dutch Cancer Society (project number 13528). The funding body had no role in the design of the study, collection, analysis, interpretation of data, or writing of the manuscript.
Supporting information
Supplement 1. The Decisional Conflict Scale questionnaire. Supplement 1 consists of the English version of the Decisional Conflict Scale questionnaire.
Acknowledgments
The authors used Microsoft Copilot to assist with language refinement and clarity during manuscript preparation.
Data Availability Statement
The datasets generated and/or analyzed during this study are not publicly available. The full dataset could contain information that might compromise research participants’ privacy and/or their conditions of consent. The data that will support the findings of this study may be available on request from the corresponding author (Aimée F. Herkendaal).
References
- 1. Barsouk A, Aluru JS, Rawla P, Saginala K, Barsouk A. Epidemiology, risk factors, and prevention of head and neck squamous cell carcinoma. Med Sci. 2023;11(2):42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. van den Besselaar BN, Sewnaik A, Hoesseini A, Dorr MC, Baatenburg de Jong RJ, Offerman MPJ. Causes and ways of death in patients with head and neck cancer. JAMA Otolaryngol Head Neck Surg. 2024;150(4):303‐310. 10.1001/jamaoto.2023.4694 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Lokker ME, Offerman MPJ, van der Velden LA, de Boer MF, Pruyn JFA, Teunissen SCCM. Symptoms of patients with incurable head and neck cancer: prevalence and impact on daily functioning. Head Neck. 2013;35(6):868‐876. [DOI] [PubMed] [Google Scholar]
- 4. Hung T‐M, Lin C‐R, Chi Y‐C, et al. Body image in head and neck cancer patients treated with radiotherapy: the impact of surgical procedures. Health Qual Life Outcomes. 2017;15(1):165. 10.1186/s12955-017-0740-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Barber B, Dergousoff J, Slater L, et al. Depression and survival in patients with head and neck cancer: a systematic review. JAMA Otolaryngol Head Neck Surg. 2016;142(3):284‐288. [DOI] [PubMed] [Google Scholar]
- 6. Hammermüller C, Hinz A, Dietz A, et al. Depression, anxiety, fatigue, and quality of life in a large sample of patients suffering from head and neck cancer in comparison with the general population. BMC Cancer. 2021;21(1):94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. van den Besselaar BN, van Hof KS, Sewnaik A, Baatenburg de Jong RJ, Offerman MPJ. Electronic health in the palliative care pathway for patients with head and neck cancer. JAMA Otolaryngol Head Neck Surg. 2025;151(1):19‐27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Davies L, Rhodes LA, Grossman DC, Rosenberg M‐C, Stevens DP. Decision making in head and neck cancer care. Laryngoscope. 2010;120(12):2434‐2445. 10.1002/lary.21036 [DOI] [PubMed] [Google Scholar]
- 9. Zafar SY, Alexander SC, Weinfurt KP, Schulman KA, Abernethy AP. Decision making and quality of life in the treatment of cancer: a review. Support Care Cancer. 2009;17(2):117‐127. [DOI] [PubMed] [Google Scholar]
- 10. Ihrig A, Maatouk I, Friederich HC, et al. The treatment decision‐making preferences of patients with prostate cancer should be recorded in research and clinical routine: a pooled analysis of four survey studies with 7169 patients. J Cancer Educ. 2022;37(3):675‐682. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Singh JA, Sloan JA, Atherton PJ, et al. Preferred roles in treatment decision making among patients with cancer: a pooled analysis of studies using the Control Preferences Scale. Am J Manag Care. 2010;16(9):688‐696. [PMC free article] [PubMed] [Google Scholar]
- 12. Josfeld L, Keinki C, Pammer C, Zomorodbakhsch B, Hübner J. Cancer patients' perspective on shared decision‐making and decision aids in oncology. J Cancer Res Clin Oncol. 2021;147(6):1725‐1732. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Takes RP, Halmos GB, Ridge JA, et al. Value and quality of care in head and neck oncology. Curr Oncol Rep. 2020;22(9):92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Forner D, Noel CW, Shuman AG, et al. Shared decision‐making in head and neck surgery: a review. JAMA Otolaryngol Head Neck Surg. 2020;146(9):839‐844. 10.1001/jamaoto.2020.1601 [DOI] [PubMed] [Google Scholar]
- 15. Stiggelbout AM, Pieterse AH, De Haes JCJM. Shared decision making: concepts, evidence, and practice. Patient Educ Couns. 2015;98(10):1172‐1179. [DOI] [PubMed] [Google Scholar]
- 16. Graham ME, Westerberg BD, Lea J, et al. Shared decision making and decisional conflict in the Management of Vestibular Schwannoma: a prospective cohort study. J Otolaryngol Head Neck Surg. 2018;47(1):52. 10.1186/s40463-018-0297-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Pecanac KE, Brown RL, Kremsreiter HB. Decisional conflict during major medical treatment decision‐making: a survey study. J Gen Intern Med. 2021;36(1):55‐61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. LeBlanc A, Kenny DA, O'Connor AM, Légaré F. Decisional conflict in patients and their physicians: a dyadic approach to shared decision making. Med Decis Making. 2009;29(1):61‐68. [DOI] [PubMed] [Google Scholar]
- 19. Zhang Q, Gu X, Liu S, et al. Decisional conflicts in patients with low‐risk papillary thyroid microcarcinomas considering active surveillance. JAMA Otolaryngol Head Neck Surg. 2024;150:952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Taylor BA, Hart RD, Rigby MH, Trites J, Taylor SM, Hong P. Decisional conflict in patients considering diagnostic thyroidectomy with indeterminate fine needle aspirate cytopathology. J Otolaryngol Head Neck Surg. 2016;45:16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Hickman RL Jr., Daly BJ, Lee E. Decisional conflict and regret: consequences of surrogate decision making for the chronically critically ill. Appl Nurs Res. 2012;25(4):271‐275. 10.1016/j.apnr.2011.03.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Knops AM, Goossens A, Ubbink DT, Legemate DA, Stalpers LJ, Bossuyt PM. Interpreting patient decisional conflict scores: behavior and emotions in decisions about treatment. Med Decis Making. 2013;33(1):78‐84. 10.1177/0272989x12453500 [DOI] [PubMed] [Google Scholar]
- 23. Bakas AT, Polinder‐Bos HA, Streng F, et al. Frailty in non‐geriatric patients with head and neck cancer. Otolaryngol Head Neck Surg. 2023;169(5):1215‐1224. 10.1002/ohn.388 [DOI] [PubMed] [Google Scholar]
- 24. Nokovitch L, Kim Y, Zrounba P, et al. Addictions, social deprivation and cessation failure in head and neck squamous cell carcinoma survivors. Cancers. 2023;15(4):1231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Hoesseini A, Dorr MC, Dronkers EAC, Baatenburg de Jong RJ, Sewnaik A, Offerman MPJ. Decisional conflict in patients with head and neck cancer. JAMA Otolaryngol Head Neck Surg. 2023;149(2):160‐167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Heirman AN, Arends CR, de Jel DVC, et al. Decisional conflict and decision regret in head and neck oncology: a systematic review and meta‐analysis. JAMA Otolaryngol Head Neck Surg. 2024;150(5):393‐404. 10.1001/jamaoto.2024.0214 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Stacey D, Légaré F, Lewis K, et al. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev. 2017;4(4):CD001431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Bakitas M, Kryworuchko J, Matlock DD, Volandes AE. Palliative medicine and decision science: the critical need for a shared agenda to foster informed patient choice in serious illness. J Palliat Med. 2011;14(10):1109‐1116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. van den Besselaar BN, Sewnaik A, Dorr MC, et al. A study protocol for individualized prognostic counselling in the palliative phase. BMC Palliat Care. 2025;24(1):9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Zanoni DK, Patel SG, Shah JP. Changes in the 8th Edition of the American Joint Committee on Cancer (AJCC) Staging of Head and Neck Cancer: Rationale and Implications. Curr Oncol Rep. 2019;21(6):52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Hui D, Nooruddin Z, Didwaniya N, et al. Concepts and definitions for “actively dying,” “end of life,” “terminally ill,” “terminal care,” and “transition of care”: a systematic review. J Pain Symptom Manage. 2014;47(1):77‐89. 10.1016/j.jpainsymman.2013.02.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. O'Connor AM. Validation of a decisional conflict scale. Med Decis Making. 1995;15(1):25‐30. [DOI] [PubMed] [Google Scholar]
- 33. O'Connor AM. User manual—decisional conflict scale; 1993. https://decisionaid.ohri.ca/docs/develop/User_Manuals/UM_Decisional_Conflict.pdf
- 34. Heirman AN, de Kort DP, Petersen JF, et al. Decisional conflict in patients with advanced laryngeal carcinoma: a multicenter study. Laryngoscope. 2024;134(8):3604‐3610. [DOI] [PubMed] [Google Scholar]
- 35. Thompson‐Leduc P, Turcotte S, Labrecque M, Légaré F. Prevalence of clinically significant decisional conflict: an analysis of five studies on decision‐making in primary care. BMJ Open. 2016;6(6):e011490. 10.1136/bmjopen-2016-011490 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Groenvold M, Petersen MA, Aaronson NK, et al. The development of the EORTC QLQ‐C15‐PAL: a shortened questionnaire for cancer patients in palliative care. Eur J Cancer. 2006;42(1):55‐64. [DOI] [PubMed] [Google Scholar]
- 37. Zhang Q, Gu X, Liu S, et al. Decisional conflicts in patients with low‐risk papillary thyroid microcarcinomas considering active surveillance. JAMA Otolaryngol Head Neck Surg. 2024;150:952‐959. 10.1001/jamaoto.2024.2718 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. O'connor AM. User Manual‐Decisional Conflict Scale. 1993. Ottawa Hospital Research Institute; 1993. [Google Scholar]
- 39. Ter Stege JA, Oldenburg HSA, Woerdeman LAE, et al. Decisional conflict in breast cancer patients considering immediate breast reconstruction. Breast. 2021;55:91‐97. 10.1016/j.breast.2020.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Lee Y‐H, Chou X‐Y, Lai Y‐H, et al. Decisional conflict and its determinants among patients with cancer undergoing immunotherapy combined with chemotherapy or targeted therapy: a cross‐sectional study. Sci Rep. 2023;13(1):12715. 10.1038/s41598-023-39280-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Manne S, Smith B, Mitarotondo A, Frederick S, Toppmeyer D, Kirstein L. Decisional conflict among breast cancer patients considering contralateral prophylactic mastectomy. Patient Educ Couns. 2019;102(5):902‐908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Chien CH, Chuang CK, Liu KL, Li CL, Liu HE. Changes in decisional conflict and decisional regret in patients with localised prostate cancer. J Clin Nurs. 2014;23(13‐14):1959‐1969. 10.1111/jocn.12470 [DOI] [PubMed] [Google Scholar]
- 43. Kates JM. Treatment‐related decisional conflict, quality of life, and comorbidity in older adults with cancer. Asia Pac J Oncol Nurs. 2018;5(4):421‐429. 10.4103/apjon.apjon_32_18 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. van Lent LGG, de Jonge MJA, van der Ham M, et al. Decisional conflict after deciding on potential participation in early phase clinical cancer trials: dependent on global health status, satisfaction with communication, and timing. Cancers (Basel). 2022;14(6):1500. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Hoesseini A, Sewnaik A, van den Besselaar BN, et al. Prognostic model for overall survival of head and neck cancer patients in the palliative phase. BMC Palliat Care. 2024;23(1):54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. van den Besselaar BN, van Klaveren D, Berzenji D, et al. Validation of the prognostic model “oncologiq palliative” for head and neck cancer patients. Eur Arch Otrhinolaryngol. 2025;282:4735‐4743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. SONCOS Normeringsrapport 13 2025 “Multidisciplinaire normering oncologische zorg in Nederland”. 2025. Stichting Oncologische Samenwerking, Federatie Medisch Specialisten.
- 48. Laurisz N, Ćwiklicki M, Żabiński M, Canestrino R, Magliocca P. Co‐creation in health 4.0 as a new solution for a new era. Healthcare (Basel, Switzerland). 2023;11(3):363. 10.3390/healthcare11030363 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Mimoso I, Figueiredo T, Midão L, et al. Co‐creation in the development of digital therapeutics: a narrative review. Int J Environ Res Public Health. 2024;21(12):1589. 10.3390/ijerph21121589 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplement 1. The Decisional Conflict Scale questionnaire. Supplement 1 consists of the English version of the Decisional Conflict Scale questionnaire.
Data Availability Statement
The datasets generated and/or analyzed during this study are not publicly available. The full dataset could contain information that might compromise research participants’ privacy and/or their conditions of consent. The data that will support the findings of this study may be available on request from the corresponding author (Aimée F. Herkendaal).