Licensed Nurse and Nursing Assistant Recognition of Delirium in Nursing Home Residents With Dementia

Abstract

Many nursing home residents experience delirium. Nursing home personnel, especially nursing assistants, have the opportunity to become familiar with residents’ normal cognitive function and to recognize changes in a resident’s cognitive function over time. The purpose of this study was to determine the accuracy of delirium recognition by licensed nurses and nursing assistants from eight nursing homes over a 12-month period. Participants were asked to complete five case vignette assessments at three different time points (in 6-month intervals) to test their ability to identify different subtypes of delirium and delirium superimposed on dementia (DSD). A total of 760 case vignettes were completed across the different time points. Findings reveal that staff recognition of delirium was poor. The case vignette describing hyperactive DSD was correctly identified by the greatest number participants, and the case vignette describing hypoactive DSD was correctly identified by the least number of participants. Recognition of the case vignette describing hypoactive delirium improved over time. Nursing assistants performed similarly to the licensed nurses, indicating that all licensed nursing home staff require further education to correctly recognize delirium in older adults.

Many nursing home residents experience delirium, a state of confusion characterized by an acute and fluctuating decline in cognitive functioning.¹ Delirium is considered a medical emergency with serious consequences, such as hospitalization, further functional decline, and mortality.^2–4 Nursing home staff—especially nursing assistants—have regular contact with residents over long periods of time, presenting the opportunity for them to become familiar with residents’ normal, everyday cognitive function and to recognize changes in their cognitive function over time. Correctly recognizing delirium-induced changes in cognitive function is challenging, however. Despite the frequent contact between nursing home staff and residents with delirium, little is known about the ability of nursing home staff to identify and manage this condition, especially in residents with dementia.

Case vignette assessments have previously been used to assess clinician understanding of delirium in studies involving acute care nurses,^5–9 home health nurses,¹⁰ medical students,¹¹ and physicians¹² in a variety of countries, including Japan,¹² Australia,^5,7,8 Canada,¹⁰ Taiwan,¹³ and the United States.^6,9,11 In general, physicians and other clinicians have exhibited poor recognition of delirium. Two years ago, our team conducted a descriptive study to determine the ability of nursing home staff to recognize dementia, delirium, and delirium superimposed on dementia (DSD; when delirium presents concurrently with preexisting dementia).¹⁴ To our knowledge, this was the first study of this nature conducted in the nursing home setting. Nursing home licensed nurses and nursing assistants were asked to complete five standardized case vignette assessments at the beginning of the study, 6 months after the start of the study, and 12 months after the start of the study. Overall, we reported poor rates of delirium recognition (18–37%). In that study, we did not collect data on the staff position (ie, licensed nurse or nursing assistant) who responded to the case vignettes and the variations of their responses, making it difficult to compare our results with those reported by other investigators who sampled participants from the acute care or community setting.

The current study is an attempt to clarify that issue by examining the accuracy of delirium recognition by licensed nurses (ie, registered and licensed practical nurses) and nursing assistants separately, using the same five case vignette assessments. First, we compared the rates of delirium recognition among licensed nurses and nursing assistants Second, we compiled participants’ narrative responses to the case vignettes in order to describe licensed nurse and nursing assistant knowledge of the different causes of delirium.

Methods

This study reports on data from the randomized RESERVE for DSD (Recreational Stimulation for Elders as a Vehicle to resolve DSD) clinical trial,¹⁵ a 5-year National Institute of Nursing Research–funded study (ClinicalTrial.gov Identifier: NCT01267682). The purpose of RESERVE for DSD was to test the efficacy of cognitive stimulation for resolving delirium in persons with dementia who were admitted to skilled nursing facilities after hospitalizations. The study was approved by the Pennsylvanie State University Institutional Review Board. The study protocol was published previously elsewhere.¹⁵

RESERVE for DSD was conducted in eight nursing homes located in central and northeast Pennsylvania. Number of beds for each facility ranged from 90 beds to 270 beds; three sites were located in large urban areas, and five sites were located in smaller rural settings. As part of the study protocol, we monitored staff knowledge of delirium because of its potential effect on delirium outcomes. No educational intervention was provided.

The case vignettes used for this study were developed by one of the authors (DF) using a literature review of delirium motoric subtypes.⁶ The case vignettes were intended to assess staff ability to identify different subtypes of delirium and DSD using a standardized format. The five case vignettes described: (1) dementia, (2) hypoactive delirium, (3) hyperactive delirium, (4) hypoactive DSD, and (5) hyperactive DSD. Expert panel members with geriatric and psychiatric expertise independently rated the participants’ diagnosis and delirium motoric subtype (where appropriate) in each of the case vignettes. Their overall agreement on the cases was 84% (κ=.69). Their agreement on the identification of delirium motoric subtype was 100% (κ=1.0).⁶

To recruit participants, research assistants displayed posters centrally and on each nursing unit that described the purpose of the case vignette data collection and invited all staff to participate (participation was voluntary). Research assistants visited each site and distributed case vignettes at times that did not conflict with the typical workflow at that site. Implied consent was obtained and written and verbal instructions for completing the case vignettes were provided. Staff members were encouraged to complete the anonymous case vignettes (see additional information in the Measures section) during their break periods and were asked to return them to a centrally located drop-box or to the research assistant. Staff members who completed the case vignettes were entered into a lottery for a cash prize. Over three time points, each 6 months apart, a total of 316 licensed nurses (ie, registered and licensed practical nurses) and 444 nursing assistants completed the case vignette assessments (Testing Group 1, n=210; Testing Group 2, n=276; Testing Group 3, n=274). The mean response rate across all facilities was 28% (range, 12–40%).

Participants were asked to read the cases presented in each of the five case vignettes, and then they were asked to respond to the question: “What caused the change in cognitive status described in each case vignette?” 6 Each participant’s response was scored for correctness. Descriptive content analysis was used to categorize the responses for qualitative analysis. To control for type 1 error in the test of multiple effects,¹⁶ multivariate multilevel modeling was used to evaluate the effects of testing group, staff position (ie, licensed nurse or nursing assistant), and the facility’s number of nursing hours per resident on rates of correct identification of the case vignettes, as well as the interaction between group and staff position. We calculated the nursing hours of care per resident for each facility throughout the study, as listed on the Pennsylvania Department of Health Website.¹⁷ Nursing hours per resident ranged from 2.97 to 3.88 across the facilities.

Results

The percentages of licensed nurses and nursing assistants who responded correctly on each of the case vignettes in each of the three testing groups are presented in Table 1. Overall, staff recognition of delirium was poor. In all three testing groups, the dementia case was identified correctly by a greater number of participants than the delirium cases. A greater number of licensed nurses recognized dementia than nursing assistants in all three testing groups. Out of the four case vignettes that described delirium-related conditions either alone or in combination with dementia, the case that was correctly identified by the most participants was the one describing hyperactive DSD. Yet, only 53% or fewer of licensed nurses and 39% or fewer of nursing assistants correctly identified this case at each time point. Conversely, the case vignette describing hypoactive DSD had the lowest level of staff recognition among the four delirium-related case vignettes. Only 24% or fewer of licensed nurses and 26% or fewer of nursing assistants correctly identified this case at each time point.

Table 1.

Rates of Correct Responses on Case Vignette Assessments

Case Vignette	Licensed Nurses			Nursing Assistants
	Testing Group 1	Testing Group 2	Testing Group 3	Testing Group 1	Testing Group 2	Testing Group 3
1 (Dementia)	70.6%	68.6%	66.4%	56.0%	55.5%	51.2%
2 (Hypoactive delirium)	30.6%	39.7%	52.7%	26.4%	36.1%	25.0%
3 (Hyperactive delirium)	45.9%	51.2%	40.9%	32%	28.4%	28.0%
4 (Hyperactive DSD)	49.4%	52.9%	51.8%	33.6%	39.4%	39.0%
5 (Hypoactive DSD)	16.5%	17.4%	21.8%	25.6%	17.4%	20.7%

Abbreviations: DSD, delirium superimposed on dementia.

Results of the multivariate analysis for the individual case vignette assessments appear in Table 2. Recognition of the case vignette depicting hypoactive delirium among licensed nurses varied significantly between testing groups (P=.02) but not among nursing assistants. No other case vignettes assessment demonstrated significant differences in recognition between the testing groups for either staff position. Compared with nursing assistants, licensed nurses were more likely to correctly identify the case vignettes depicting hyperactive delirium (P<.0002) and hyperactive DSD (P=.005). There were no other significant differences between staff positions.

Table 2.

Effects of Testing Group, Staff Position, and Number of Nursing Hours per Resident on Case Vignette Assessment Performance

Case Vignette	1 (Dementia)		2 (Hypoactive Delirium)		3 (Hyperactive Delirium)		4 (Hyperactive DSD)		5 (Hypoactive DSD)
	Estimated Effect Size	SE	Estimated Effect Size	SE	Estimated Effect Size	SE	Estimated Effect Size	SE	Estimated Effect Size	SE
Testing Group	−0.0267	0.042	−0.014	0.027	−0.0283	0.024	0.021	0.053	−0.025	0.032
Staff Position	0.1215	0.083	−0.0366	0.064	0.1729	0.046a	0.131	0.047a	−0.0382	0.061
Number of Nursing Hours Per Resident	0.104	0.044a	0.1929	0.072a	−0.1194	0.049a	0.1073	0.074	−0.0408	0.036
Staff Position × Testing Group Interaction	0.0081	0.063	0.1356	0.058a	−0.0206	0.050	0.0058	0.029	0.0342	0.038

^aP<.05

Abbreviations: DSD, delirium superimposed on dementia; SE, standard error.

We were also interested in understanding the effect that nursing staff hours per resident had on one’s ability to recognize dementia and delirium. Higher nursing hours per resident was associated with a greater likelihood of correct responses on the dementia case vignette (P=.02) and the hypoactive delirium case vignette (P=.007). In contrast, a higher number of nursing hours per resident was associated with lower rates of correct responses on the hyperactive delirium case vignette (P=.016). The DSD case vignettes were not affected by the number of nursing hours per resident.

In Table 3, we present the three most common responses to the case vignette assessments for each case. Licensed nurses and nursing assistants most commonly reported Alzheimer’s disease, dementias, and the aging process as the causes for the dementia case vignette. On the case vignettes describing hypo- and hyperactive delirium with no underlying dementia, licensed nurses cited up to 23 different causes, most frequently noting UTIs and medications, whereas nursing assistants identified 22 different causes, naming UTIs, depression, and medications. Licensed nurses identified approximately 20 different underlying causes of hypo- and hyperactive DSD, most commonly urinary tract infections (UTIs) and Alzheimer’s disease and other dementias, whereas nursing assistants identified 24 different causes, more commonly cited Alzheimer’s disease and other dementias, aging process, and medications.

Table 3.

Most Common Responses of Licensed Nurses and Nursing Assistants in all Testing Groups for Each Case Vignette Assessment

Case Vignette	Three Most Common Responses (% of Participants)
	Licensed Nurses	Nursing Assistants
1 (Dementia)	AD & Dementias (31.9%)	Aging Process (26.25%)
	Aging Process (25.84%)	AD & Dementias (17.35%
	Progression of Disease (6.38%)	Medications (6.07%)
2 (Hypoactive delirium)	UTI (55.62%)	UTI (19.96%)
	Medications (15.20%)	Depression (14.32%)
	Depression (14.59%)	Medications (12.59%)
3 (Hyperactive delirium)	UTI (30.09%)	UTI (13.67%)
	Medications (20.67%)	Medications (12.36%)
	Environmental Changes (16.11%)	AD & Dementias (9.33%)
4 (Hyperactive DSD)	UTI (20.97%)	AD & Dementias (16.49%)
	2) AD & Dementias (20.67%)	Aging Process (8.24%)
	Environmental Changes (14.59%)	UTI (8.24%)
5 (Hypoactive DSD)	UTI (21.28%)	UTI (21.28%)
	AD & Dementias (15.81%)	AD & Dementias (15.81%)
	Medications (11.55%)	Medications (11.55%)

Abbreviations: AD, Alzheimer’s disease; DSD, delirium superimposed on dementia; UTI, urinary tract infection.

Discussion

This is one of the largest studies to date to examine nursing home staff recognition of delirium across time in both licensed nurses and nursing assistants. Approximately half of the participants recognized dementia, but the recognition rates dropped when the same staff members were asked about the four case vignettes describing delirium and DSD. The case vignette describing hyperactive DSD was the most frequently recognized case vignette by both licensed nurses and nursing assistants, and licensed nurses recognized hyperactive delirium and hyperactive DSD more frequently than nursing assistants. Conversely, hypoactive DSD was recognized least often. These results are similar to past studies that reported that hypoactive delirium was the least-recognized subtype of delirium among registered nurses.^6,18 In the current study, licensed nurses identified hypoactive DSD at similar rates to the nursing assistants.

Our findings support the work of previous authors reporting on staff delirium recognition. Voyer and colleagues,¹ for example, found that long-term care nurses did not perceive changes in the cognition or behavior of their patients with dementia to be important and therefore did not document those changes. In our study, the case vignette describing dementia was recognized correctly by the most participants of both staff positions, who most commonly cited the cause of the case to be “Alzheimer’s disease and other dementias” or “the aging process”. This supports the notion that nursing home staff members are most familiar with dementia and the variations in behavior common to dementia; however, this knowledge does not translate into knowledge related to cognitive changes that are common in persons with dementia, such as delirium. Our findings also were similar to those of McCrow et al.,⁸ who reported that the dementia case vignette was correctly identified the most often.

Our study extended our prior work¹⁴ by specifying licensed nursing staff and nursing assistant levels of delirium recognition—a unique perspective when compared with other studies measuring delirium recognition. When Fick and colleagues⁶ initially used case vignettes to assess delirium recognition, acute care nurses recognized hypoactive DSD only 21% of time. In the present study, both licensed nurses and nursing assistants performed similarly.

The hypoactive delirium case was the only one for which performance differed between testing groups, and only among licensed nurses, suggesting that staff members’ ability to recognize the conditions presented in the case vignettes was relatively constant over the 12-month study period. This finding is similar to that of Gesin and colleagues,¹⁹ who detected no significant changes in the knowledge or recognition of delirium among critical care registered nurses before the implementation of educational interventions.

Staff members at facilities with a greater number of nursing hours per resident performed better on the recognition of the dementia and hypoactive delirium cases compared with staff members at facilities with fewer nursing hours per resident. Surprisingly, staff members at facilities with more nursing hours did not perform as well on the hyperactive delirium case vignette. Although more nursing hours per resident may suggest greater exposure to residents,²⁰ our results suggest that this does not necessarily translate to improved recognition of cognitive impairment. Our results are mixed, however, and suggest further study is needed.

When comparing the narrative responses on the case vignette assessments, we found similar answers from the licensed staff and nursing assistants to those reported by Hosie and colleagues.^5,7 For the case vignette describing hypoactive delirium, the authors reported that licensed nurses cited infection, hypoxia, and medications as causes, whereas the licensed nurses in our study also cited UTIs and medications as causes. Licensed nurses and nursing assistants most often identified the cause of hyperactive and hypoactive delirium (no dementia) to be a UTI. The next most common staff-identified cause was medications, which was similar to the findings of previous studies in which staff were asked to identify the causes of behavioral changes related to a delirium diagnosis.^21,22 Because nursing home staff are in a position to recognize subtle changes in their patients, experience may have taught them that certain factors (eg, infections, medications) lead to these subtle changes.

Nursing home staff members are most familiar with dementia and the variations in behavior common to dementia.

Overall, nursing assistants made observations about mental status changes that were as accurate as licensed nurses. This finding was not anticipated, but has important implications for scope of nursing assistant practice and education. Two studies found significant value in the observations of nursing assistants and suggest promoting the value of their input with the healthcare team by expanding their role to improve the quality of geriatric care. Liu²³ examined the nursing assistant’s role in the process of pain management in cognitively impaired individuals; results demonstrated that nursing assistants played several roles in the pain management process (ie, pain assessor, pain reporter, subordinate to implementing prescribed medications, instigator of implementing nonpharmacologic interventions). Further, Lawhorne and colleagues²⁴ found nursing assistants to be invaluable in the care of long-term care residents with urinary incontinence. In a 2014 mixed methods study in which researchers interviewed registered nurses about how their work environment influences care, Cline et al.²⁵ found that nursing assistants were frequently mentioned as being critical to the quality of acute care. McMullen and colleagues²⁶ recently examined certified nurse assistant (CNA) scope of practice state by state and argued that “allowing CNAs to provide expanded tasks for which they are adequately trained may optimize care and prevent adverse events.”

Given our study findings and the need to improve resident care in nursing homes, we recommend that consideration be given to this role expansion for nursing assistants. There is growing evidence and support highlighting the importance of including nursing assistants as integral team members,²⁷ especially in the nursing home setting. We also recommend that multifaceted educational interventions^28–30 to promote the knowledge and recognition of delirium be included in the mandatory education hours specified for all nursing home staff, including nursing assistants. Educational sessions should be interactive, should include case studies and live videos, and should engage participants and nursing home leadership.²⁷

The interpretation of our study’s findings is limited by the use of a convenience sample and the lack of demographic and other information about participants that may have been useful for interpreting our results. We were not able to determine whether the same or different staff members were completing the case vignettes over the three time points. Also, although the case vignettes describe clinical cases, there may be additional contextual aspects that would assist staff in delirium recognition in an actual situation. Despite these limitations, the study has several important advantages over previous work, including a large sample size, the inclusion of staff from eight different nursing homes across rural and urban settings, assessment at multiple time points, the use of previously validated case vignette measures for delirium recognition, and the differentiation between staff positions within the sample.

Conclusion

In this study, delirium and its subtypes were recognized by both licensed nurses and nursing assistants less than 53% of time. Nursing assistants performed similarly to the licensed nurses, indicating that all staff members require further education to recognize delirium—especially the delirium subtype hypoactive DSD. Considering the dire consequences of delirium in the dementia population, it is vital that staff members that are closest to these vulnerable individuals accurately recognize delirium to institute treatment measures. Our findings suggest that nursing assistants—who performed as well as licensed nurses in our study—may be able to assess mental status changes with additional education.