Development and validation of the Neuro-Oncology Grief Index

Abstract

Background

Grief among neuro-oncology healthcare providers (NOPs) affects emotional resilience, professional performance, and patient care. This study adapted and validated the 30-item Neuro-Oncology Grief Index (NOGI) to assess grief experiences and coping strategies in NOPs after patient loss.

Methods

A cross-sectional design was used to adapt the NOGI through item generation, expert review, and field testing. Psychometric properties were evaluated via exploratory factor analysis, internal consistency using Cronbach’s alpha, and stepwise regression was performed using variables that were significantly associated with provider grief to determine their predictive value and the amount of variance.

Results

The sample comprised 80 NOPs (68% female, 75% white, 62% with >5 years’ experience). An exploratory factor analysis identified a single factor for the grief scale. Internal consistency was acceptable with Cronbach’s alpha = .92. High grief was linked to frequent patient deaths, burnout, and compassion fatigue. Adverse coping behaviors, like staying overly busy or isolating, were common; 58% of NOPs avoided emotional support due to workplace norms. Regression analysis identified past-year risk factors, provider protective factors, and adverse behavioral risks as significant predictors of grief.

Conclusions

The NOGI is a reliable, valid tool for assessing NOP grief, highlighting emotional grief, burnout, compassion fatigue, and coping strategies. Interventions addressing emotional support, burnout, and workplace norms are needed to enhance provider well-being and patient care. Future research should further validate the NOGI and address compassion fatigue and emotional support access.

Key Points.

• The NOGI is a valid tool to measure NOP grief and associated variables.

• Up to 60% of NOPs exhibit elevated NOGI scores and show the need for grief support.

Importance of the study

The Neuro-Oncology Grief Index (NOGI) study is important because it offers a validated tool for assessing grief among neuro-oncology healthcare providers. It shows that 58% of providers display negative coping behaviors, such as isolation, often due to frequent patient deaths, burnout, and compassion fatigue, all of which impair professional performance and patient care. By examining 3 key variables, the NOGI accounts for 64% of the variation in grief and identifies factors that predict NOP grief. The study’s results support targeted actions like grief support programs and burnout prevention efforts to strengthen provider resilience and enhance care quality. It also outlines a research agenda for further validation and policy changes aimed at prioritizing grief support in neuro-oncology settings.

Grief is a complex, multidimensional response to loss, characterized by conflicting emotions and behaviors that reflect the interplay of psychological, physiological, and social processes.¹ In the context of healthcare, particularly neuro-oncology, grief extends beyond personal loss to encompass the profound emotional impact of patient death on providers. This grief is often intensified by the deep emotional connections formed with patients, the frequent exposure to adverse end-of-life experiences, and the professional responsibility to manage complex, debilitating conditions such as brain tumors and neurological complications.^2,3 We describe grief as involving “conflicting emotions and behaviors,” suggesting a dynamic experience where providers may simultaneously experience sadness, guilt, powerlessness, and professional duty.¹ The theoretical foundation for this study draws from the Dual Process Model,⁴ which suggests that grievers oscillate between loss-oriented (eg processing emotions) and restoration-oriented (eg resuming professional duties) coping, potentially explaining conflicting behaviors. Additionally, it relies on the emotional and cognitive dimensions of grief proposed by Worden’s task-based model,⁵ which emphasizes processing pain and adjusting to loss.

Furthermore, grief experienced by healthcare providers is recognized as a unique form of bereavement that intertwines personal loss with professional responsibilities and expectations.¹ Over time, consistent exposure to patient death can erode a provider’s emotional resilience, leading to feelings of helplessness, self-doubt, and professional inadequacy.² Since the expressions of grief in the healthcare profession are seen as a weakness, it often prevents providers from seeking necessary emotional support, further exacerbating the risk of burnout and affecting overall well-being and the quality of care offered to patients.³

The literature suggests that many oncologists are drawn to the field for its intellectual rigor, scientific advancements, or personal motivations (eg family experiences with cancer). Still, the emotional burden of frequent loss is often underestimated.^6,7 For neuro-oncologists, this gap is more pronounced. Other researchers² indicate that oncologists often feel unprepared for the grief and ethical dilemmas they encounter, such as managing patient and family expectations in the face of limited treatment options.

Research on provider grief in the field of oncology found high levels of grief that stemmed, at least in part, from a sense of responsibility for the patient’s care and grief-related spillover, which was described as experiencing the effects of patient loss at home due to the inability to separate work from personal life.^1,8–10 Sometimes, oncologists reported facing these effects before the death of a patient due to grappling with the foreknowledge of a poor prognosis, causing feelings of powerlessness, self-doubt, guilt, compartmentalization, denial, dissociation, and failure.¹ Now, while grief over patient loss is seen as an intrinsic part of clinical practice, oncology providers face unique pressures due to their legal responsibility for patient care and the potential blame for a patient’s death.^3,11 This hardship appeared more pronounced during instances of emotional transference, when patients died at an early age, when deaths were unexpected, or when a keen sense of closeness developed between patients and their oncology providers.¹ Notably, however, the burden of grief experienced by providers in the field of neuro-oncology has not yet been explored. As healthcare professionals, neuro-oncology healthcare providers (NOPs) treat highly symptomatic patients suffering from debilitating diseases such as brain tumors, brain metastases, and neurological complications arising from cancer. Consequently, NOPs are regularly exposed to adverse end-of-life experiences, which can lead to maladaptive coping and symptomatology akin to post-traumatic stress disorder.³ Because existing tools have fallen short in effectively evaluating the grief experienced by oncology providers,^12–14 particularly nurse practitioners, we constructed the 30-item Neuro-Oncology Grief Index (NOGI) by adapting items from the validated Texas Revised Inventory of Grief I-II (TRIG)¹² and Brief Cope questionnaires (COPE)¹³ to assess the grief and coping strategies of NOPs after patient loss. Hence, the current study aimed to adapt an instrument and address the following research questions: (1) Do NOPs who experience a greater number of patient deaths in the past year report higher grief? (2) Do NOPs with more years of experience report higher grief? (3) What variables should an intervention focus on to influence NOP grief? and (4) What is the significance of demographic variables on NOP grief?

Methods

Researchers used a descriptive cross-sectional design to develop and evaluate the NOGI. This study consisted of 3 phases: Phase 1, item generation; Phase 2, expert panel review and field test; and Phase 3, evaluation.^15–17

Item Generation

As part of the item generation, 2 researchers carefully reviewed items from the TRIG I-II, a 22-item self-administered questionnaire that evaluates dimensions of grief—emotional response, non-acceptance, thoughts, and past disruption due to loss.^12–14 Additionally, researchers assessed the COPE, a 28-item instrument that evaluates dimensions of coping—problem-focused, emotion-focused, and avoidant coping.¹⁸ Researchers selected both instruments due to their reported validity and reliability in measuring grief and coping strategies.^13,14 Researchers chose 23 items from both instruments (ie 14 items from TRIG I-II and 9 items from COPE) based on their relevance to the study’s purpose and research questions. The first 15 items of the NOGI, mostly adapted from TRIG I-II, measured provider grief on a five-point Likert scale (Strongly Disagree to Strongly Agree). The remaining 15 items, 9 adapted from the COPE, assessed provider coping on a five-point Likert scale (Never to Always). These items were chosen to examine the experience of neuro-oncologists and how they cope with the experience. The NOGI combines these items to provide a holistic measure of how NOPs experience and manage grief, reflecting the reality that coping strategies shape grief outcomes. For example, maladaptive coping (eg grieving alone) may intensify grief-related emotions like guilt or self-doubt, while adaptive coping (eg seeking advice) may buffer these effects. By integrating both, the NOGI can identify patterns of maladaptive versus adaptive responses specific to NOPs, enabling targeted interventions that address both emotional distress and coping behaviors.

Other collected data included employment-sponsored training received, work-related burnout symptoms, mode of patient aftercare services, number of patient deaths experienced in the past year, patient symptomatology, and risk factors in both professional and personal life. Work-related burnout symptoms and risk factors were assessed through single-item, self-report questions where participants indicated (yes/no) if they had experienced specific psychological symptoms or reactions in the past year following a patient’s death, including compassion fatigue, powerlessness or failure, self-doubt, withdrawal, shame or guilt, intrusive thoughts, work-related burnout, the need to cope with or grieve a patient’s death, and severe grief reactions. These items were developed by the research team based on common themes in the oncology grief literature^1–3,8–10 to provide contextual data on providers’ experiences.

Demographic data collected were provider age, sex, race/ethnicity, parental status, relationship status, medical specialty, and years in practice. Other collected data included employment-sponsored training received, work-related burnout symptomatology, mode of patient aftercare services, number of patient deaths experienced within the last year, patient symptomatology, and risk factors in both professional and personal life.

Expert-Panel

To determine content and face validity, researchers conducted a test using the 30-item pilot version of the NOGI (see Figure 1). In this test, a three-member physician panel recruited from a neuro-oncology department of a large general hospital received a briefing on the study’s purpose and instructions on how to critique the instrument. Based on feedback received from the NOPs, researchers revised the NOGI. After an expert panel review, 1 tenured professor and 1 non-tenured associate professor trained in instrument development examined the NOGI to ensure that items focused on the core concepts of grief and coping. From the feedback provided by this peer review, researchers made additional changes to the instrument.

Figure 1.

Original 30-item Neuro-Oncology Grief Index (NOGI).

Field-testing

Researchers further assessed the NOGI by asking a sample of medical students (n = 10) to determine the readability, feasibility, and time of completion. Eligibility criteria for the field test were (1) individuals ≥18 years of age and older; (2) who were current third-year medical students; (3) who could read, write, and speak English; and (4) who self-selected to complete the NOGI. Eligible participants were informed that their decision to participate or not participate would not impact any benefits they were otherwise entitled to receive. No compensation was given, and no identifying information was collected. The questionnaire was administered online via REDCap v.13.7.6, a secure, web-based software platform designed to support data capture for research studies.^19–20 Each participant read the standardized research information sheet and voluntarily provided consent. Based on feedback from the field test, researchers further revised the NOGI.

Evaluation

Through exploratory factor analysis, researchers used SPSS v29 to assess the psychometric properties of the 15-item provider grief scale.²¹ Only the provider grief scale was evaluated, as it aligned with the study’s primary purpose. Eigenvalues ≥1.0 were used for factor retention, and a Kaiser–Meyer–Olkin cutoff of >0.60 was used for acceptable factor loadings.²² Cronbach’s alpha was set at α > .70 to assess the internal consistency of the provider grief scale.²³ Based on a P-value <.05, the relationship strength between independent variables and the provider grief score was assessed. Based on this analysis, a stepwise regression was performed using variables to determine their predictive value and the amount of variance explained in the provider grief score.

Study sample

Convenience sampling methods were used to recruit eligible participants from a repository of members (>3200) from the Society for Neuro-Oncology, the European Association of Neuro-Oncology, and the International Brain Tumor Alliance to address provider grief among NOPs. To support recruitment efforts, researchers obtained permission to include a brief study description with a REDCap link in each association’s monthly newsletter for 3 months. Eligibility criteria for the main study were (1) individuals ≥18 years of age and older; (2) who were current NOPs; (3) who could read, write, and speak English; and (4) who self-selected to complete the NOGI. Eligible participants were informed that their ­participation would not impact any benefits they were otherwise entitled to receive. No compensation was given, and no identifying information was collected. All study data were collected and managed online using REDCap. Each participant read and approved a standardized research information sheet and voluntarily provided informed consent. These individuals were recruited from May to August 2023.

Results

The sample primarily consisted of attending physicians (78.8%; n = 63), followed by trainees/residents (13.8%; n = 11) and nurses/nurse practitioners (7.5%; n = 6). Additional roles included physician assistants, professors, researchers, fellows, physical therapists, and medical students. Most participants reported their specialty as neuro-oncology (76.2%; n = 61), with smaller groups in neurosurgery (8.8%; n = 7), oncology (6.3%; n = 5), neurology (3.8%; n = 3), advanced practice (3.8%; n = 3), and radiation oncology (1.3%; n = 1). More than half (57.5%; n = 46) had over 10 years of experience, 23.8% (n = 19) had 1-5 years, 17.5% (n = 14) had 6-10 years, and 1.3% (n = 1) had less than 1 year.

Demographically, 72.5% (n = 58) of providers reported being parents, and 86.3% (n = 69) were married/partnered. Smaller proportions identified as single (11.3%; n = 9) or divorced/separated (2.5%; n = 2). Detailed demographic data are summarized in Table 1, and occupational characteristics are presented in Table 3.

Table 1.

Provider characteristics (N = 80)

Characteristics	n (%)
Age 48.3 (SD = 10.91)
Sex
Female	51 (63.7%)
Male	29 (36.3%)
Ethnicity
Asian	12 (15.0%)
Middle Eastern/African	2 (2.5%)
Latino/Hispanic	3 (3.8%)
White	63 (78.8%)
Parent
No	22 (27.5%)
Yes	58 (72.5%)
Relationship status
Married/partnered	69 (86.3%)
Divorced/separated	2 (2.5%)
Single	9 (11.3%)
No. of patient deaths in the last year
Less than 10	15 (18.8%)
11-20	29 (36.3%)
21-40	23 (28.7%)
More than 40	13 (16.3%)

Table 3.

Occupational characteristics of providers (N = 80)

Characteristics	n (%)
Occupation
Attending physician	63 (78.8%)
Trainee/resident	11 (13.8%)
Nurse/nurse practitioner	6 (7.5%)
Specialty
Neuro-Oncology	61 (76.2%)
Neurosurgery	7 (8.8%)
Oncology	5 (6.3%)
Neurology	3 (3.8%)
Advanced practice	3 (3.8%)
Radiation oncology	1 (1.3%)
Years of experience
Over 10 years	46 (57.5%)
1-5 years	19 (23.8%)
6-10 years	14 (17.5%)
Less than 1 year	1 (1.3%)
Aftercare service after death of patient
Telephone	45 (56.3%)
Mail	44 (55.0%)
In-person	28 (35.0%)
None	11 (13.8%)

Provider Maladaptive Coping Behaviors

Over 96% tried to remain busy following a patient’s death, and more than 80% felt compelled to be strong for others. Other common strategies included distracting with electronics (72.6%), grieving alone (66.3%), and restricting emotional expression (61.3%) (see Table 2). In terms of Provider Adaptive Coping Behaviors, most providers avoided substance use (alcohol, nicotine, or prescription drugs), and nearly all reported never or rarely engaging in risky behaviors such as risky sex. Common adaptive strategies included seeking advice from others (78.8%), using exercise (72.5%), finding humor (70.0%), and engaging in mindfulness, prayer, or meditation (55.0%).

Table 2.

Frequency (percentage) of provider coping behaviors after the death of a patient (N = 80)

No.	Item	Never	Rarely	Sometimes	Usually	Always
1.	I try to remain busy	2 (2.5%)	1 (1.3%)	10 (12.5%)	48 (60.0%)	19 (23.8)
2.	I try to be strong for others	3 (3.8%)	11 (13.8%)	23 (28.7%)	36 (45.0%)	7 (8.8%)
3.	I grieve alone	8 (10.0%)	19 (23.8%)	21 (26.3%)	26 (32.5%)	6 (7.5%)
4.	I bury my emotions with food	25 (31.3%)	21 (26.3%)	26 (32.5%)	7 (8.8%)	1 (1.3%)
5.	I engage in risky sexual behaviors	77 (96.3%)	2 (2.5%)	0 (0%)	0 (0%)	1 (1.3%)
6.	I use electronics	9 (11.3%)	13 (16.3%)	19 (23.8%)	31 (38.8%)	8 (10.0%)
7.	I use prescription drugs	64 (80.0%)	10 (8.0%)	4 (5.0%)	2 (2.5%)	0 (0%)
8.	I use alcohol	36 (45.0%)	19 (23.8%)	17 (21.3%)	6 (7.5%)	2 (2.5%)
9.	I use nicotine products	75 (93.8%)	1 (1.3%)	2 (2.5%)	1 (1.3%)	1 (1.3%)
10.	I freely express my emotions	4 (5.0%)	11 (13.8%)	34 (42.5%)	24 (30.0%)	7 (8.8%)
11.	I seek advice from others	3 (3.8%)	14 (17.5%)	30 (37.5%)	27 (33.8%)	6 (7.5%)
12.	I seek inner peace	13 (16.3%)	23 (28.7%)	22 (27.5%)	14 (17.5%)	8 (10.0%)
13.	I receive emotional support	32 (40.0%)	26 (32.5%)	9(11.3%)	10 (12.5%)	3 (3.8%)
14.	I exercise	7 (8.8%)	15 (18.8%)	23 (28.7%)	19 (23.8%)	16 (20.0%)
15.	I use laughter as medicine	9 (11.3%)	15 (18.8%)	24 (30.0%)	20 (25.0%)	12 (15.0%)

Occupational Characteristics of Providers

Upon the death of a patient, more than half of the providers (56.3%; n = 45) used the telephone to deliver aftercare services and express condolences to family members, followed by those (55%; n = 44) who used mail, those (35%; n = 28) who provided aftercare and condolences to the family in person, those (38.8%; n = 27) who used a message service, and those (13.8%; n = 11) who provided no patient aftercare services (see Table 3).

Provider Occupational Supports

Most providers reported receiving support from coworkers or colleagues (77.5%; n = 62), followed by family and friends (51.3%; n = 41) and nurses/nurse practitioners (38.8%; n = 31) (see Table 4). Few reported support from counselors or wellness coaches (16.3%; n = 13), and none from non-consultant hospital doctors. Most providers (63.7%; n = 51) had never received employment-sponsored training on work-related stress or communication about grief. Of those who did, only 21.3% (n = 17) received training on grief-related communication, and 15.0% (n = 12) on work-related stress. More than half (57.5%; n = 46) reported rarely or never receiving emotional support after a patient death, although a similar proportion (57.6%) indicated they rarely or never required such support (Table 4).

Table 4.

Provider occupational supports (N = 80)

Source	Frequency (%)
Coworker/colleague	62 (77.5%)
Family/friends	41 (51.3%)
Nurse/nurse practitioner	31 (38.8%)
Counselor/wellness coach	13 (16.3%)

Provider-Reported Characteristics of Patients

Less than one-third of providers (31.3%; n = 25) reported that more than 75% of their patients died from their condition, followed by 30.0% (n = 24) who reported 25%-50% of patients, 23.8% (n = 19) who reported 51%-75% of patients, and 15.0% (n = 12) who reported less than 25% of patients died from their condition (see Table 5). Additional data on patient symptomatology, including physical pain, emotional difficulty, and cognitive decline, are presented in Table 5.

Table 5.

Provider (N = 80)-reported characteristics of patients

	Frequency (%)
Itema	<25% of PT	25%-50% of PT	51%-75% of PT	>75% of PT
Died from condition	12 (15.0)	24 (30.0)	19 (23.8)	25 (31.3)
Physical pain	43 (53.8)	22 (27.5)	8 (10.0)	7 (8.8)
Emotional difficulty	9 (11.3)	12 (15.0)	26 (32.5)	33 (41.3)
Cognitive decline	9 (11.3)	13 (16.3)	20 (25.0)	38 (47.5)

^aBased on the number of patients (PT) treated, the percentage (%) of patients for each item.

Provider Grief Scale

A maximum likelihood with Promax rotation exploratory factor analysis was conducted with the original 15-item provider grief scale (see Table 6). The scree plot and eigenvalues (>1.0) were visually inspected to determine the number of factors. One main factor was identified with an eigenvalue of 5.48, accounting for 60.86% of the variance, for which 9 items were acceptable for factor loadings (>0.60). The second factor, with an eigenvalue of 0.69, for which 2 items were acceptable for factor loadings (>0.60), only accounted for 8.5% of the variance. Consequently, this factor was deemed insufficient. Using these 9 retained items in the provider grief scale, each of which was measured on a five-point Likert scale (Strongly Disagree to Strongly Agree), the mean provider grief score was 21.81 (SD = 8.2; min/max score = 9-45), for which the internal consistency of the scale was acceptable with Cronbach’s alpha = .92.

Table 6.

Exploratory factor analysis of the provider grief scale (N = 80)

No.	Item	Factor 1	Factor 2
1.	After the patient died, I found it difficult to get along with others	0.738	0.174
2.	After the patient died, I found it hard to focus at work	0.811	0.087
3.	After the patient died, I lost interest in my family, friends, and outside activities	0.648	0.269
4.	After the patient died, I blamed myself for things I could have done better	0.671	0.152
5.	After the patient died, I was irritable	0.849	0.196
6.	After the patient died, I became angry	0.728	0.291
7.	After the patient died, I experienced a variety of feelings (eg fear or relief)	0.682	0.284
8.	It is still difficult to accept the death of this patient	0.680	0.463
9.	I still cry when I think about the death of this patient	0.608	0.454
10.	I still get upset when I think about the death of this patient	0.478	0.532
11.	I still experience painful memories of the death of this patient	0.567	0.508
12.	Sometimes, I miss the patient who died	0.066	0.748
13.	I often think about the patient who died	0.216	0.817
14.	No one will ever compare to patient who died	0.372	0.446
15.	It is unfair that this patient died	0.117	0.480

Shaded cells indicate the factor onto which each item loaded. If neither cell is shaded, that item did not load on either factor.

Regression Analysis

A stepwise regression analysis was conducted to evaluate the association, predictive value, and variance explained by 7 independent variables on the dependent variable of provider grief, as measured by the nine-item provider grief scale. The variables assessed were provider support (PS), protective training (PT), provider protective factors (PPF), past-year risk (PYR), patient characteristics (PC), work-related risk factors (WRF), and adverse behavioral risk factors (ABF). Only 3 variables—PYR, PPF, and ABF—demonstrated statistical significance and were retained in the final model, accounting for 55% of the variance in provider grief scores (F_3,76 = 31.61, P < .001). Past-year risk factors, including powerlessness/failure, self-doubt, compassion fatigue, and burnout, were the strongest predictor, explaining 34% of the variance (F_1,78 = 41.78, P < .001, β = 1.82, 95% CI, 1.11-2.53). Adverse behavioral risk factors, such as remaining busy or grieving alone, accounted for 15% of the variance (F_1,77 = 24.75, P < .001, β = .56, 95% CI, 0.25-0.87). Provider protective factors, including adaptive coping strategies like seeking advice or using exercise, contributed 6% of the variance (F_1,76 = 11.30, P = .001, β = .75, 95% CI, 0.30, 1.19). Multicollinearity was assessed using the Variance Inflation Factor (VIF) and tolerance statistics. The VIF values ranged from 2.04 to 2.31, all well below the conservative threshold of 5, indicating no evidence of significant multicollinearity among the independent variables. Similarly, tolerance values ranged from 0.43 to 0.49, all above the critical threshold of 0.1, further confirming the absence of multicollinearity.^24,25

Discussion

This study introduced a novel instrument designed to measure grief among healthcare providers, specifically those in neuro-oncology. The original 15-item provider grief scale was reduced to 9 items with strong internal consistency. Applying this tool to identify elevated levels of provider grief could serve as a precautionary measure to better support providers and help prevent work-related burnout, compassion fatigue, and other risk factors that could lead to poor work performance and negatively impact patient health. In the current study, almost half of all providers treated more than 21 patients who died within the last year. Surprisingly, there was no statistically significant difference in provider grief scores based on the number of patient deaths experienced per year or years of work-related experience.

As expected, providers in the sample reported that most patients suffered from cognitive decline and emotional difficulties due to their prognosis, while only a minority suffered from physical pain. These findings aligned with previous reports that showed the likelihood of systematic metastasis of a brain tumor was rare,^26,27 meaning that neuro-oncology patients typically did not suffer from physical pain, although they might experience headaches.^28,29 However, providers in the sample did report treating a high volume of patients with emotional and cognitive difficulties, which could become challenging, mainly because tools to manage these symptoms are historically limited and insufficiently studied.^30–32

Remarkably, despite the numerous challenges in treating neuro-oncology patients, PS was not significant in the model because most providers reported that they never or rarely received emotional support when a patient died and did not feel they required it. Unfortunately, among the minority who sought emotional support, not all received it, raising concerns about the perceived availability of both professional and personal emotional resources for NOPs. Previous literature has shown that social support at work is critical for coping with provider grief.³² The current study echoed this finding, as most providers sought support from coworkers and colleagues, suggesting that providers may have found it easier to seek support from individuals directly involved with patient care who understood the unique challenges of treating this patient population. Additionally, with nearly half of the sample reporting that showing grief at work was unprofessional, it could be possible that the lack of readily available professional support in the workplace contributed to the perception that expressing emotions is discouraged or undervalued.

In the model, provider grief was significantly associated with variables representing PYR factors, PPF, and adverse behavioral risks. The most common PYR factors included coping with a patient’s death, experiencing symptoms of work-related burnout, compassion fatigue, powerlessness/failure, and self-doubt. The most common adverse behavioral risks were remaining busy, trying to be strong for others, and grieving alone. By keeping their grief bottled up or trying to be strong for others, providers might unknowingly worsen their mental health, leading to severe grief reactions, as reported by some of the sample. These findings highlight the critical role of emotional experiences and coping strategies in shaping provider grief, with maladaptive coping exacerbating grief and adaptive coping potentially mitigating its impact. The non-significant variables (PS, PT, PC, WRF) suggest that factors such as patient mortality rates or workplace support may have less direct influence on grief compared to individual coping behaviors and emotional responses.

The finding that identifying elevated levels of grief via the NOGI could support NOPs by mitigating risks such as burnout, compassion fatigue, and reduced work performance warrants further consideration of the directionality of these relationships. The regression analysis showed that PYR factors, including burnout and compassion fatigue, were the strongest predictors of provider grief, explaining 34% of the variance (Table 6), suggesting that these states may amplify grief responses. However, the relationship is likely bidirectional: elevated levels of grief may also exacerbate burnout or compassion fatigue by depleting providers’ emotional resources, creating a feedback loop that impacts well-being and work performance.^1–3 Notably, the lack of association between grief scores and the number of patient deaths or years of experience indicates that grief is not solely driven by cumulative exposure but may reflect underlying vulnerabilities, such as pre-existing burnout or ineffective coping strategies. By identifying providers with high grief scores, the NOGI can serve as a screening tool to flag those at risk for or already experiencing burnout and compassion fatigue, enabling early interventions (eg peer support, counseling) to disrupt this cycle. The cross-sectional nature of this study limits conclusions about causality. Still, the strong association between grief, burnout, and compassion fatigue underscores the need for longitudinal research to clarify directionality and inform targeted support strategies for NOPs.

Surprisingly, only 31.3% of providers reported that more than 75% of their patients died from their condition, with 45% reporting 50% or fewer, which is lower than expected given the high mortality rates associated with neuro-oncologic conditions such as brain tumors.²⁶     This may reflect ambiguity in the questionnaire item, as providers may have interpreted “died from their condition” as excluding deaths from complications (eg infections, seizures). Alternatively, the sample’s diversity, including providers from subspecialties like pediatric neuro-oncology or hospice, may have included those managing less fatal conditions. The convenience sampling method and small sample size may also have overrepresented providers with atypical patient populations or lower mortality rates, potentially skewing results. While the regression analysis found no significant association between patient deaths and provider grief, the possibility of a sample skewed toward providers experiencing more unexpected deaths (eg from complications) could not be ruled out, as unexpected deaths are known to intensify grief.¹ Future research should clarify the cause of death (disease vs complications) and assess the proportion of unexpected deaths to better understand their impact on provider grief.

To address these issues sufficiently, educational and supportive interventions for managing grief related to patient death are needed to assist providers in their emotionally and mentally taxing work. These interventions should address challenges in accessing social-emotional support, expressions of grief, work-related burnout, difficulties in maintaining emotional boundaries when patients die, normal reactions to patient death, and signs of emotional distress or when providers should seek professional help. Addressing these challenges is essential to ensure the best quality of life for providers and the best quality of care for their patients.

Limitations

This cross-sectional study collected data at a single point, and researchers did not control for any confounders. Consequently, the findings indicated associations rather than causations and may have introduced a degree of bias. The convenience sampling method meant that providers in the study might not represent the broader population, limiting the generalizability of results. Furthermore, the small sample size and lack of participant diversity prevented meaningful analysis of the effect of demographic variables on provider grief.

Another limitation is the sample size of 80 NOPs for the exploratory factor analysis. While within the COSMIN ¹⁶ guideline’s “adequate” range (50-99 participants), it is at the lower threshold for psychometric testing and may introduce risks of unstable factor solutions or reduced statistical power. According to COSMIN, a sample size of ≥100 would be considered “very good.” The modest sample size, combined with the convenience sampling method, may limit the generalizability of the factor structure. Future studies should aim for larger, more diverse samples to confirm the psychometric properties of the NOGI and establish test-retest reliability. We recommend that future research establish test–retest reliability to strengthen the scale further and establish values that are associated with adverse health outcomes. Acknowledging these limitations is important for interpreting the findings so that future research can address these issues.

Author Contributions

Conceptualization: Rachael D. Nolan, Abigail Beausir, Barbara O’Brien, Katy Peters, Shlomit Yust-Katz; methodology: Abigail Beausir, Rachael D. Nolan, Shlomit Yust-Katz; formal analysis: Abigail Beausir, Rachael D. Nolan, Shlomit Yust-Katz, Jeffrey S. Hallam, Raj Mukherjee; data curation: Abigail Beausir, Raj Mukherjee, Rachael D. Nolan, Shlomit Yust-Katz; writing—original draft preparation: Rachael D. Nolan, Shlomit Yust-Katz, Jeffrey S. Hallam, Abigail Beausir; writing—review and editing: Rachael D. Nolan, Abigail Beausir, Barbara O’Brien, Katy Peters, Shlomit Yust-Katz, Jeffrey S. Hallam, Raj Mukherjee; supervision: Rachael D. Nolan, Shlomit Yust-Katz; project administration: Rachael D. Nolan, Shlomit Yust-Katz.

Conflict of Interest Statement

Dr. Yust-Katz reports: Consulting, honoraria or travel support: Medison, AbbVie and Pfizer. Board participation: Society for Neuro Oncology, Pfizer, and Organon. All remaining authors have declared no conflicts of interest.

Funding

The authors received no funding for this research project.

Ethics Statement

An external review board deemed this study exempt from ethical review and approval because the proposed research involved no more than minimal risk to participants, utilized only survey procedures, and did not collect any identifying information.

Data Availability

Data will be made available upon reasonable request.