Training and Practice Gaps in Nursing Home Palliative Care: A Cross-Sectional Study

Abstract

Background:

Palliative care needs are prevalent among nursing home (NH) residents. However, access to and integration of palliative care services remain limited. NHs often rely on a workforce with varying levels of training and exposure to palliative care, which may influence care quality and consistency. Understanding the perspectives of NH personnel across disciplines about palliative care is essential to developing sustainable models of care.

Methods:

We conducted a cross-sectional survey of clinical employees (N=398) at seven NHs within a single integrated health system in New York State to (1) gauge attitudes toward and knowledge about palliative care, (2) estimate the extent of engagement with advance care planning, and (3) elicit perceptions of the most valuable components of palliative care for NH residents. Median regression and ordinal logistic regression were used to assess predictors of palliative care attitudes and of checking for advance care planning documents. Disciplines were grouped into six categories for analysis: MD/NP, RN, social work/case management, rehabilitation services (PT/OT/SLP), CNAs, and LPNs.

Results:

While most respondents believed that palliative care improves quality of life (Median=5, IQR 4–5), only 68% correctly identified it as distinct from hospice. Prior exposure to palliative care and favorable attitudes were associated with more frequent checking for advance care planning documents (OR=1.75 and OR=1.50, respectively; both p<0.05). Attitudes varied by discipline, with physical/occupational/speech therapists reporting less favorable views than other groups (B=−0.70, p=0.005). Respondents prioritized pain management, psychiatric symptom management, and caregiver support as areas where palliative care specialists could offer the greatest help and further training was most desired.

Conclusions:

Despite support for palliative care, gaps in knowledge and clinical engagement persist and vary by discipline. Discipline-tailored training and program design may improve palliative care delivery in NHs and help ensure more consistent, value-aligned care for residents with serious illness.

Approximately 1.3 million U.S. residents live in nursing homes (NHs), a number expected to more than double by 2030.^1,2 Serious illnesses are common among NH residents; approximately half have dementia,³ up to 37% have heart failure,⁴ and 10% have cancer.⁵ Additionally, about half of NH residents die within six months of admission, and a quarter of deaths in the United States occur in NHs.^6,7

Palliative care (PC) is specialized medical care aimed at improving quality of life for individuals with serious illnesses and their caregivers through symptom management, psychosocial support, and advance care planning (ACP).^8,9 Over half of all NH residents have unmet PC needs, including untreated psychological and physical symptoms.^10–12 Furthermore, compared to individuals living in the community, NH residents with serious illnesses are more likely to receive aggressive end-of-life care, including repeated hospitalizations.⁵

Currently, PC is poorly integrated into NH care. The most frequent model of PC in NHs is hospice care.² However, NH residents may be less likely to enroll in hospice than individuals without NH stays.¹³ Currently, only a third of eligible residents enroll in hospice.¹⁴ Other models of PC delivery in NHs, including specialist PC consultants,² remain uncommon.¹⁵ For instance, in a study of care provided at 54 NHs, only 4% of residents received a PC consultation.¹⁶ As a result, many non-hospice eligible residents suffer from unmet PC needs.

Improved PC integration could enhance NH care, but models from acute or ambulatory settings may not translate well to NHs. NHs have more limited access to specialist medical care and rely more heavily on licensed practical nurses (LPNs) and certified nursing assistants (CNAs),¹⁷ who are typically not trained in specialty PC.^18,19 Further, financial disincentives and lack of referral streams limit the availability of PC in NHs.^20,21 Thus, interventions must align with the unique clinical workflows and staffing patterns in NHs.

Prior studies suggest that NH personnel often feel unprepared to address PC needs^22–24 but value PC services and desire more training.^19,25 Additionally, existing survey studies have revealed gaps in knowledge among NH personnel regarding PC.¹⁹ Further studies are needed to assess attitudes toward specific components of PC among NH personnel. These data are critical for informing workforce development and sustainable NH PC models.

To address this knowledge gap, we surveyed interdisciplinary personnel across seven NHs within a single healthcare system in New York to: (1) gauge general attitudes toward and knowledge about PC services, (2) estimate the extent of engagement with ACP, and (3) elicit perceptions of the most valuable components of PC for NH residents and to understand personnel-level factors that impact these outcomes.

Methods:

Study Design:

This is an analysis of cross-sectional data generated through a PC needs assessment survey of personnel at seven NHs in a health system in New York State.

Survey Development:

We elected not to use existing PC knowledge/attitude measurement tools (e.g., the Palliative Knowledge Test-Basic²⁶) because such tools were either prohibitively lengthy, overly specific to disciplinary knowledge (e.g., prescribing medications), or lacking in focus on domains of PC. Instead, we used a survey design framework to develop an instrument to evaluate attitudes, knowledge, and perceived clinical and training needs in PC among personnel.²⁷ The survey was informed by a literature review and discussions with experts in geriatrics, PC, and NH care (co-authors of this manuscript). We iteratively designed our survey instrument through multiple rounds of refinement with both experts and end-users. This included cognitive interviews with content experts in research methods, palliative medicine, PC social work, and PC nursing, as well as with end-users representing various NH team members, including social workers, registered nurses (RNs), nurse practitioners (NPs), CNAs, medical directors, and administrators. We did not perform formal pilot testing, as the survey was intended for internal quality improvement.

Our instrument (Supplement 1) included 10 demographic/professional background questions, 6 questions related to attitudes, knowledge, and exposure to PC, 5 questions about current participation in PC delivery, 6 questions about perceived PC clinical and training needs. Questions about specific PC functions were informed by domains established by the National Consensus Project on Quality Palliative Care Clinical Practice Guidelines (Supplement 2).⁹

Survey Population and Data Collection:

Patient-facing employees of a large healthcare organization in New York providing NH, a Program of All-Inclusive Care for the Elderly and home care (N=1193; 960 in NHs) were emailed a request to complete this 10-minute, web-based REDCap survey in May of 2023 with the support of institutional and site-specific medical directors. This study’s analytic sample is limited to those respondents working in NH settings.

Statistical Analysis:

Descriptive statistics were used to summarize survey responses. Continuous variables were summarized using medians and interquartile ranges (IQR) whereas categorical variables were presented as frequencies and percentages.

We created a 15-point composite from three 5-point Likert items to assess PC attitudes (one item, for which a lower score was favorable, was reversed to create a 15-point composite). We conducted median regression, a form of quantile regression that models the median response rather than the mean. This approach was selected due to the range of response levels and the sparsity of observations at some levels. Independent variables included age, gender, race/ethnicity, professional discipline, PC availability at the respondent’s worksite, and prior PC exposure.

Checking ACP documents for new patients supports value-concordant care in NHs. To evaluate factors associated with respondents checking for ACP documents when caring for a new patient, we conducted ordinal logistic regression using the proportional odds model. The dependent variable was self-reported frequency of checking for ACP documents (characterized using a 5-point scale from never to always). Independent variables included demographic and professional characteristics, prior exposure to PC, and attitudes toward PC.

Statistical analyses were conducted using R version 4.4.2 and Stata version 18.5. A p-value <0.05 was considered statistically significant.

Research Ethics:

Respondents were informed in the invitation that the survey was voluntary. Consent was assumed by completion of the survey, which was deemed not to be human subjects research by the Weill Cornell IRB.

Results:

Demographics:

398 NH personnel responded (41% response rate). Most respondents identified as female (82%, n=327). 39% (N=156) of respondents identified as Black or African American (39%, n=156) and 29% (N=116) identified as white. For full demographics, see Table 1.

Table 1.

Characteristics of survey respondents.

Characteristic	N (%)
Age (years)
21–30	49 (12)
31–40	102 (26)
41–50	83 (21)
51–60	103 (26)
>61	44 (11)
Prefer not to answer	16 (4)
Gender
Female	327 (82)
Male	61 (15)
Non-binary	1 (0)
Prefer not to answer	9 (2)
Race/Ethnicity
Native American	3 (1)
Asian	55 (14)
Black or African American	156 (39)
Latina/o/x or Spanish origin	43 (11)
Native Hawaiian/Pacific Islander	5 (1)
White	116 (29)
Other	5 (1)
Prefer not to answer	25 (6)
Professional Discipline
Case manager	2 (1)
Certified nursing assistant	109 (27)
Licensed practical nurse	39 (10)
Nurse practitioner	10 (3)
Occupational therapy	21 (5)
Physical therapy	22 (6)
Physician	16 (4)
Registered nurse	104 (26)
Social worker	24 (6)
Speech-language pathologist	4 (1)
Other	47 (12)
Experience (years)
<1	20 (5)
1–5	97 (25)
5–15	139 (35)
>15	136 (35)
Prior Exposure to PC
Professional	264 (66)
Personal	176 (44)
None	71 (18)

PC: palliative care

Professional Background and PC Experience:

The most represented disciplines were CNAs (26%, n=109) and RNs (26%, n=104).

Sixty-six percent (N=264) had formal PC training through courses, rotations, or certification. Forty-four percent (N=176) had personal experience with PC. 18% (N=71) of respondents reported no prior exposure to PC (see Table 2).

Table 2.

Summary of respondents’ knowledge and attitudes about palliative care (PC) by professional discipline.

Survey Prompt	Professional Discipline
	Physicians and NPs	RNs	SWs and Case Managers	PT, OT, and SLPs	CNAs	LPNs	Other
PC is another name for hospice care. (N (% correct))	24 (92)	87 (84)	20 (77)	30 (64)	49 (45)	27 (69)	31 (66)
Patients can receive PC while they receive curative or life-prolonging treatments. (N (% correct))	21 (84)	85 (82)	20 (77)	34 (72)	71 (66)	32 (82)	36 (78)
PC can improve the quality-of-life of my patients with serious illnesses. (median (IQR))	5 (4–5)	5 (4–5)	5 (4–5)	4 (2–5)	4 (3–5)	5 (5–5)	5 (4–5)
I worry that consulting PC will make my patients lose hope. (median (IQR))	1 (0–3)	1 (1–3)	2 (1–3)	2 (1–3)	2 (1–4)	2 (1–3)	2 (1–3)
I would like to increase access to PC services for my patients. (median (IQR))	5 (4–5)	4 (3–5)	4 (4–5)	3 (2–5)	4 (3–5)	5 (3–5)	4 (3–5)

NP: nurse practitioner; RN: registered nurse; SW: social worker; PT: physical therapist; OT: occupational therapist; SLP: speech-language pathologist; CNA: certified nursing assistant; LPN: licensed practical nurse: IQR: interquartile range

PC Knowledge:

Sixty-eight percent (n=268) correctly identified that PC is distinct from hospice, with accuracy highest among MD/NPs (92%, N=24) and lowest among CNAs (45%, N=49) (Table 3). Similarly, 76% (N=299) correctly recognized that PC could be provided alongside curative treatment, though knowledge was lower among CNAs and PT/OT/SLPs (66% and 72%, respectively, versus 84% for MDs/NPs).

Table 3.

Percentage of respondents by discipline who rated each component of palliative care (PC) as one of the top three (“Top 3”) areas in which a PC specialist could assist their patients or as an area in which they wanted more training (“Need”).

Component of PC	Professional Discipline
	All (N=398)	Physicians and NPs (N=26)	RNs (N=104)	SWs and Case Managers (N=26)	PTs, OTs, and SLPs (N=47)	CNAs (N=109)	LPNs (N=39)	Other (N=47)	p-value
Identified as one of the top three palliative care clinical needs among the respondent’s patients
Pain management	75	65	75	65	94	68	95	66	<0.001
Non-pain symptom management	24	15	31	12	9	24	39	23	0.011
Psychiatric symptom management	42	46	45	39	32	42	59	30	0.116
Caregiver and family support	49	31	58	62	49	49	33	45	0.053
Advance care planning and GOC discussions	36	39	44	58	32	30	23	36	0.044 *
Spiritual care	17	12	13	12	13	28	23	15	0.066 *
Hospice referral and EOL care	18	39	11	19	19	16	15	30	0.015 *
Prognostication	6	15	4	15	11	2	15	6	0.012 *
Managing ethical conflicts	6	4	11	8	2	5	8	4	0.473*
Identified as an area in which the respondent desires further training
Pain management	55	62	51	46	68	58	62	43	0.164
Non-pain symptom Management	32	24	40	31	15	27	44	38	0.017
Psychiatric symptom management	48	58	47	54	36	49	56	43	0.461
Caregiver and family support	47	23	45	54	47	55	41	49	0.123
Advanced care planning and GOC discussions	43	23	54	58	38	41	38	36	0.040 *
Spiritual care	30	19	26	23	15	42	38	32	0.010 *
Hospice referral and EOL care	29	50	31	27	23	25	23	38	0.144*
Prognostication	18	27	28	19	15	10	15	15	0.033 *
Managing Ethical Conflicts	24	23	35	35	6	19	23	28	0.004 *

p-values are based on either a chi-square test or, in cases of small sample size, a Fisher’s Exact test (*). Disciplines were grouped by similarity of role and/or training background. NP: nurse practitioner; RN: registered nurse; SW: social worker; PT: physical therapist; OT: occupational therapist; SLP: speech-language pathologist; CNA: certified nursing assistant; LPN: licensed practical nurse: EOL: end-of-life; GOC: goals of care

General Attitudes Toward PC:

Most respondents strongly agreed that PC improves the quality of life for seriously ill patients (Median=5, IQR 4–5) (Table 3). Concerns about loss of hope following PC referral were minimal (Median=2, IQR 1–3). Interest in increasing access to PC was high (Median=4, IQR 3–5).

In regression models, Native American/Alaskan Native (B = 2.00, p = 0.001) and Hispanic/Latino/a/e/x respondents (B = 0.50, p = 0.006) reported more favorable attitudes toward PC compared to non-Hispanic white respondents. Lack of prior exposure to PC was associated with less favorable attitudes (B = −0.38, p = 0.012). Additionally, rehabilitation professionals (PT/OT/SLP) reported significantly less favorable attitudes compared to other professional disciplines (B = −0.70, p = 0.005). Other demographic factors, including age, gender, and site, were not significant predictors in the model.

ACP:

Fifty-eight percent (N=230) of respondents reported always checking for ACP documents. In ordinal logistic regression, respondents with exposure to PC (either personally or professionally) were more likely to frequently check for ACP documents (OR = 1.75, 95% CI: 1.25–2.45, p < 0.001). Positive attitudes toward PC were associated with higher frequencies of checking for ACP documents (OR = 1.50, 95% CI: 1.10–2.05, p = 0.015).

PC Clinical and Training Needs:

Respondents identified each PC domain as an unmet need in at least 50% of the residents cared for by the respondents. The domains identified as an unmet need in ≥75% of residents were pain management, psychiatric symptom management, caregiver and family support, ACP and goals of care discussions, spiritual care, and hospice referral and end-of-life care.

Respondents indicated that PC specialists would be most helpful in addressing pain management (75% of respondents, N=298), caregiver and family support (49%, N=194), and psychiatric symptom management (42%, N=167). Over 90% of rehabilitation therapists (PT/OT/SLP) felt that pain management was among the most important roles of PC. In contrast, respondents felt there was the least need for a PC specialist to manage ethical conflicts (6%, N=25) and prognosticate (6%, N=23).

Respondents were asked to select domains of PC in which they were interested in more training. Respondents voiced the greatest interest in receiving further training in the areas of pain management (55% of respondents, N=220), psychiatric symptom management (48%, N=190), and caregiver/family support (47%, N=188) (See Table 3). Across disciplines, most respondents indicated that they preferred to learn more about these topics by discussing cases with a PC expert (54%, N=216). This was most preferred by MD/NPs (73%, N=19) and least preferred by social workers (42%, N=11) and CNAs (47%, N=51). In contrast, many CNA and LPNs preferred learning by computer modules (CNA: 40%, N=644 LPN: 56%, N=22) (see Supplement 3).

DISCUSSION

This study offers a nuanced examination of the perspectives of NH personnel on PC across diverse disciplines within a multi-site health system. Our findings provide insights into the current state of PC workforce preparedness and discipline-specific educational needs within the NH setting. We found that while most respondents expressed positive attitudes towards PC, substantial variation existed in knowledge, practices, and perceived needs across disciplines. We also found that perspectives on PC varied by discipline and were strongly associated with attitudes toward and exposure to PC. These findings highlight an opportunity to increase the uptake of best practices among frontline personnel through education.

In terms of educational content, respondents identified pain management, psychiatric symptom management, and caregiver support as priorities for both clinical care and training. However, variation across disciplines in perceived needs, desired training domains, and preferred training formats reinforces the need for training and clinical program development in the NH setting to reflect and leverage disciplinary diversity. The prioritization by all disciplines of case discussion is striking, given that some disciplines, such as CNAs, may have less direct interface with traditional consultants. This prioritization suggests the possibility of developing more educationally-focused consultative processes in which expert consultants provide feedback and teaching to a broader constituency of personnel as part of the consultation process.

A strength of this study is its large interprofessional sample drawn from seven NHs. However, our study has several limitations. First, this was a cross-sectional, self-report survey conducted within a single system in a single state, limiting its generalizability. Furthermore, our sample is subject to response and social desirability biases; respondents with an a priori interest or more positive attitudes towards PC may have been more likely to respond. Second, differences in the disciplinary scope of practice may have confounded our findings. Third, small sample sizes in some discipline categories limited our conclusions. Finally, some of the items may not correlate well with meaningful engagement with PC; for example, the checking for ACP documents (one of the behaviors asked about in the survey) may entail cursorily checking for such documents without engaging with patients’ goals and wishes meaningfully.

Our findings underscore the need for systematic investment in discipline-specific PC education and clinical programming. The high reported burden of unmet needs, particularly in psychosocial, spiritual, and family domains, demonstrates the potential for increased PC involvement in NH care. Investing in PC programs that incorporate both workforce development and clinical programming may help bridge these gaps and reduce unnecessary suffering in the NH setting.

Key Points.

• Nursing home personnel across disciplines endorsed the importance of palliative care but varied in knowledge, engagement, and perceived clinical and training needs.

• Respondents with prior exposure to palliative care and more favorable attitudes were more likely to engage in advance care planning practices.

• Pain management, psychiatric symptom management, and caregiver support were consistently identified as top areas where palliative care specialists could assist.

Why does this paper matter?

Understanding discipline-specific differences in palliative care knowledge, attitudes, and priorities can inform targeted training and care models in nursing homes. Tailored approaches are essential to improve the quality of life for seriously ill residents and reduce unmet palliative care needs in this setting.