Person-Centered Care Plans for Nursing Home Residents With Behavioral and Psychological Symptoms of Dementia

Abstract

Little literature exists examining the development and implementation of person-centered care (PCC) plans focused on behavioral and psychological symptoms of dementia (BPSD). The current study aimed to describe BPSD documented in nursing home (NH) residents’ care plans, the types of approaches staff document in addressing those symptoms, and whether resident and/or facility characteristics are associated with documentation of PCC approaches. The sample included 553 residents from 55 NHs in two East Coast states. Resistiveness to care (44.9%), agitation (42.2%), and aggression (42%) were most frequently documented in care plans. PCC approaches were documented in care plans in 21.3% to 62.7% of cases depending on BPSD type. Resident (e.g., younger age, lower functional ability, lower cognitive ability, longer length of stay, male gender) and facility (e.g., less certified nursing assistant staffing hours, greater percentage of residents taking antipsychotic medications, non-profit status) characteristics were associated with increased odds of PCC approaches being documented. Optimal PCC planning is discussed, and a sample PCC plan is provided.

There are approximately 1.4 million individuals residing in nursing homes (NHs) in the United States and approximately one half have dementia (Harris-Kojetin et al., 2019). Behavioral and psychological symptoms of dementia (BPSD), such as aggression, agitation, and apathy, are common in NHs, with up to 97% of people with dementia experiencing at least one symptom (Cloak & Al Khalili, 2020; Kales et al., 2015). Preventing and managing BPSD are critical for reducing caregiver burden; promoting a pleasant, therapeutic environment; and improving residents’ quality of life (Gerlach & Kales, 2018; Smalbrugge et al., 2017). The Centers for Medicare & Medicaid Services (2019) emphasize the use of nonpharmacological, person-centered dementia care practices that are defined as those that use resident values, preferences, and interests when addressing BPSD (Abbott et al., 2018). These practices, however, are often difficult for NHs to implement due to challenges at the facility (e.g., resources, staffing) and the resident/family level (Banaszak-Holl et al., 2015; Smalbrugge et al., 2017). Research also shows that staff lack knowledge of how to implement evidence-based person-centered care (PCC) approaches for BPSD; yet, staff are responsible for care planning and surveyors are trained to examine staff care plans for evidence of PCC delivery (Matthews et al., 2018).

An interdisciplinary care planning process with individualized written care plans is a federal regulatory requirement for all NHs intended to support individuals with dementia to live their best possible life (Dellefield, 2006; Molony et al., 2018). Although guidelines exist for person-centered assessment and care planning for persons with dementia and there is empirical evidence that person-centered dementia care models decrease BPSD and psychotropic medication use, there has been limited research around the care planning process and content (Li & Porock, 2014; Molony et al., 2018). A recent integrative review that focused on how care plans have changed after intervention implementation (e.g., staff education) found that care plans generally did not have personalized interventions, although some postintervention care plans were more individualized (Mariani et al., 2017). Most studies included in the review were not conducted in the United States and did not focus on BPSD specifically.

To expand on what is currently known regarding the development, content, and impact of PCC plans for BPSD, the purpose of this exploratory study was to describe: (a) the BPSD that staff include in their written care plans; (b) the types of approaches staff document in the care plans for addressing those BPSD (i.e., person-centered vs. non-specific); and (c) whether resident characteristics (e.g., age, gender, race, marital status, education, length of stay, cognitive ability, functional ability) and/or facility characteristics (e.g., profit status, star [quality] rating, number of beds, percent of residents taking antipsychotics, RN staffing hours, certified nursing assistant [CNA] staffing hours) are associated with the documentation of PCC approaches. These findings can be used to help staff improve PCC plans for BPSD.

METHOD

The current study was a secondary analysis using baseline data from an ongoing pragmatic clinical trial (Resnick, Kolanowski, et al., 2018). In the trial, the effectiveness of an implementation strategy for improving staff uptake of nonpharmacological management for BPSD without causing decline in function and physical activity is being tested. The study was approved by a university institutional review board and the protocol has been published (Resnick, Kolanowski, et al., 2018).

Setting and Participants

To increase external validity, NH and resident exclusion criteria were kept to a minimum. NHs from two East Coast states were invited to participate if they: (a) agreed to actively partner with the research team on an initiative to change practice; (b) had at least 100 beds or 50 beds if the facility had a dedicated dementia care unit; (c) could identify a staff member to be an internal champion and work with the research team in the implementation process; and (d) were able to access email and websites via a phone, tablet, or computer. Invitations were posted on relevant websites (e.g., state-based long-term care organizations), mailed to eligible facilities (i.e., sufficient bed size) across the two states, and followed up with telephone calls and site visits. Approximately 100 NHs in each state were approached for participation.

Residents were eligible if they: (a) were living in a participating NH; (b) were age ≥55 years; (c) had a diagnosis of dementia and a score of 0 to 12 on the Brief Interview of Mental Status (BIMS) (Saliba et al., 2012), indicating moderate to severe cognitive impairment; (d) had a history within the past 1 month of exhibiting at least one BPSD; (e) were not enrolled in hospice; and (f) were not in the NH for short-stay rehabilitation. A list of all potentially eligible residents was obtained. These residents were approached for assent by trained study research assistants (RAs). If the resident did not assent, there was no further contact. If a resident did assent, an evaluation of their decisional capacity was performed using the Evaluation to Sign Consent (Resnick et al., 2007). Residents’ written/verbal consent was obtained if they demonstrated decisional capacity. If decisional capacity was impaired, a resident’s legally authorized representative was approached for consent. Approximately 12 to 13 residents per NH were recruited.

Procedures

At baseline, resident demographic information was obtained from the medical chart by RAs. Resident cognitive ability was obtained by RAs who interviewed the resident, and functional ability was determined by staff interview. RAs were baccalaureate or master’s prepared RNs who had at least 5 years of experience working with NH residents. They were instructed on study measures and data collection procedures during a 2-day training session. RAs evaluated participants’ most recent care plan for evidence of BPSD and PCC approaches for BPSD. Care plan approaches were designated as person-centered if they incorporated the personal characteristics, preferences, and/or interests of the resident when planning interventions.

Measures

Facility Characteristics.

For each participating NH, facility characteristics were obtained, including profit status (non-profit/for-profit) and NH size (number of beds). Overall star quality rating (1 to 5; higher scores indicate higher quality rating), RN and CNA staffing hours, and antipsychotic medication use (percent of residents who received an antipsychotic medication) data were taken from the Nursing Home Compare web-site (access https://www.medicare.gov/nursinghomecompare). RN and CNA staffing hours were measured by the average number of hours per resident per day of care.

Resident Characteristics.

Demographic characteristics were recorded for each resident, including age, race, gender, marital status, length of stay (in days), and education. Cognitive ability was assessed using the BIMS (Saliba et al., 2012), which includes recall and orientation questions with scores ranging from 0 to 15. Higher scores indicate greater cognitive ability. Functional ability was measured using the Barthel Index (BI), a 10-item measure of performance of activities of daily living (Mahoney & Barthel, 1965). Items are weighted to account for the amount of assistance required. Scores range from 0 (complete dependence) to 100 (complete independence). The BIMS and BI have been used with older adults in NHs and have prior evidence of reliability and validity. Interrater reliability in the current study ranged from 0.98 to 1.00.

Care Plans.

The Checklist for Evidence of Person-Centered Approaches for Behavioral and Psychological Symptoms of Dementia in Care Plans is an investigator-developed tool that includes eight potential BPSD: agitation, aggression, sexually inappropriate behavior, wandering, inappropriate or disruptive vocalizations, repetitive behavior, apathetic behavior, and resistiveness to care. Each consented resident’s care plan was reviewed by a RA for the presence of one or more of these BPSD documented by staff. For each BPSD exhibited by the resident, a RA determined whether any listed intervention for that behavior was a PCC approach. Either there was an intervention in the care plan appropriately addressing the behavior or there was not (1 = yes, 0 = no). If the item was not relevant (i.e., the resident did not exhibit BPSD), this was marked as “not applicable.” Interrater reliability of the Checklist was excellent (correlations between two RAs: r = 0.93, p = 0.001) and Cronbach’s alpha was 0.96 in the current study. Further evidence of reliability and validity has been published (Resnick, Galik, et al., 2018).

Analyses

To address our first two research questions, means, standard deviations, frequencies, and percentages were used to describe the demographic and clinical characteristics of the resident sample, facility characteristics, and BPSD and PCC approaches documented in resident care plans. To address our third research question regarding the association of resident and facility characteristics with the presence of PCC strategies in care plans, we first recoded marital status (0 = not married, 1 = married) and race (0 = non-Caucasian, 1 = Caucasian) to dichotomous indicators to ease interpretation in models. We then used multi-level logistic regression to determine the predictive association of Level 1 resident characteristics (i.e., age, gender, race, marital status, education, length of stay, cognitive ability, and functional ability) and Level 2 facility characteristics (i.e., profit status, star rating, number of beds, percent of residents taking antipsychotics, RN staffing hours, and CNA staffing hours) with PCC approaches documented for each of the eight types of behaviors noted in care plans (i.e., resistiveness to care, agitation, aggression, wandering, apathy, disruptive vocalizations, repetitive behaviors, and sexually inappropriate behaviors). Each outcome was coded as 1 = yes, PCC approach documented or 0 = no, PCC approach not documented; where a specific behavior was not observed, data were set as missing. For each of the eight outcomes, we ran an unconditional model to examine variance in the proportion of PCC plans across NHs. We then added resident level characteristics (Level 1) and facility characteristics (Level 2) to the models. Then, a model was revised to include only the significant factors in favor of parsimony and improved model fit (i.e., Akaike Information Criterion). Models were built in SPSS Statistics 26 using the generalized linear mixed models approach.

RESULTS

Participants were drawn from 55 NHs across the states of Maryland (n = 32) and Pennsylvania (n = 23). Approximately two thirds of NHs were of for-profit status (61.8%). NH size and quality ratings ranged from small to large (62 to 412 beds) and low to high quality (1 to 5 stars), respectively. From 5.3% to 29.2% of residents were prescribed antipsychotics. On average, residents received <1 hour (48.6 minutes) of RN and 2.3 hours (136.6 minutes) of CNA staffing hours per day (Table 1).

TABLE 1.

Nursing Home Facility Characteristics (N = 55)

Characteristic	n (%)
State
Maryland	32 (58.2)
Pennsylvania	23 (41.8)
Profit status
Non-profit	21 (38.2)
For profit	34 (61.8)
	Mean (SD) (Range)
Size (no. of beds)	150.27 (78.03) (62 to 412)
Quality measures
Total star (quality) rating	3.44 (1.29) (1 to 5)
Residents taking antipsychotics (%)	12.87 (4.64) (5.3 to 29.2)
Staffing hours (per resident/per day in minutes)
RN	48.56 (18.86) (16 to 106)
CNA	136.56 (23.04) (91 to 198)

Note. CNA = certified nursing assistant.

Most residents (n = 553) were widowed (47%), Caucasian (75.8%), female (72%), and 56 to 108 years old (mean age = 85.87, SD = 10.75 years). Average length of stay was 2.5 years (SD = 2.4 years). Mean cognitive ability (BIMS) was 4.12 (SD = 3.5; range = 0 to 12), and average BI functional ability score was 9.04 (SD = 30.87; range = 3 to 100) (Table 2).

TABLE 2.

Sample Demographic and Clinical Characteristics (N = 553)

Characteristic	n (%)
Gender
Female	398 (72)
Male	155 (28)
Race
Caucasian	419 (75.8)
Black	133 (24.1)
Mixed	1 (0.2)
Marital status
Widowed	260 (47.0)
Married	100 (18.1)
Never married	97 (17.5)
Divorced	57 (10.3)
Separated	3 (0.5)
Education level
Less than high school	20 (3.6)
Some high school	27 (4.9)
Graduated high school	99 (17.9)
Trade school	13 (2.4)
Some college	25 (4.5)
Graduated college	30 (5.4)
Postgraduate studies	11 (2.0)
	Mean (SD) (Range)
Age (years)	85.87 (10.75) (56 to 108)
Cognitive abilitya	4.12 (3.5) (0 to 12)
Functional abilityb	39.04 (30.87) (3 to 100)
Length of stay (days)	907.89 (891.59) (0c to 7,111)

^aAssessed with the Brief Interview of Mental Status cognitive inventory.

^bAssessed with the Barthel Index.

^cJust admitted.

BPSD documented in written care plans, from most to least frequent, were: resistiveness to care, agitation, aggression, wandering, apathy, disruptive vocalizations, repetitive behaviors, and sexually inappropriate behaviors (Table 3). Eighty-three percent of residents (n = 460) had at least one BPSD documented in their care plan during the time of study. When these BPSD were noted, PCC approaches ranged from being present in 21.3% for sexually inappropriate behavior, up to 62.7% for resistiveness to care, with rates varying based on the behavior type (Table 3).

TABLE 3.

Behavioral and Psychological Symptoms of Dementia Noted in Resident Care Plans and Use of Person-Centered Care (PCC) Approaches

	n (%)
Behavior Noted in Care Plan	Yes	Noa	PCC Approach Used in Care Plan	Example PCC Approach
Resistiveness to care	244 (44.9)	299 (55.1)	153 (62.7)	Use of person’s preferences related to personal care (e.g., bathing time preferences, type of bathing)
Agitation	229 (42.2)	314 (57.8)	127 (55.5)	Use of environmental preferences (e.g., appropriate levels of stimulation)
Aggression	228 (42)	315 (58)	120 (52.6)	Use of person’s preferences for activity/distraction and to avoid boredom
Wandering	211 (38.9)	332 (61.1)	109 (51.7)	Safety plan in care plan to allow for safe wandering (e.g., access to locked open area)
Apathy	210 (38.7)	333 (61.3)	111 (52.9)	Inclusion of resident preferences in facilitating daily activity (e.g., inclusion of pets, holiday-related activities)
Disruptive vocalization	198 (36.5)	345 (63.5)	114 (57.6)	Guidelines for how to communicate during care and other interactions when being disruptive
Repetitive behavior	95 (17.5)	448 (82.5)	22 (23.2)	Guidelines for how to provide care in a way that will optimize function and include individual in care activity yet avoid repetitive behavior
Sexually inappropriate behavior	89 (16.4)	454 (83.6)	19 (21.3)	Plan for how to react/communicate and respond to episodes of inappropriate sexual behavior

Note. Care plan data are missing for 10 participants; therefore, percentages refl ect the proportion of care plans of 543 completed care plans.

^aBehavior not relevant.

After accounting for the nesting of participants within facility, we found that individual resident characteristics (i.e., age, gender, length of stay, cognitive ability, functional ability) and facility-based characteristics (i.e., profit status, percentage of residents taking antipsychotics, and CNA staffing hours) were associated with the documentation of PCC plans. The associations were dependent on the type of BPSD considered (Table 4).

TABLE 4.

Association of Resident and Facility Characteristics With Person-Centered Care Approaches for Behavioral and Psychological Symptoms of Dementia

	Resistiveness to Care (N = 244)		Agitation (N = 229)		Aggression (N = 228)		Wandering (N = 211)
	B (SE)	OR	B (SE)	OR	B (SE)	OR	B (SE)	OR
Level 1–Resident Characteristics
Age	−0.002 (0.02)	—a	0.003 (0.03)	—	0.02 (0.02)	—	−0.01 (0.02)	—
Gender (1 = female)	−0.08 (0.33)	—	−0.31 (0.44)	—	0.00 (0.39)	—	0.31 (0.37)	—
Length of stay	0.00 (0.00)	—	0.00 (0.00)	—	0.00 (0.00)	—	0.00 (0.00)	—
Cognitive ability	−0.04 (0.05)	—	−0.02 (0.06)	—	0.03 (0.06)	—	−0.02 (0.05)	—
Functional ability	−0.002 (0.01)	—	0.003 (0.01)	—	−0.01 (0.004)*	0.99	0.002 (0.01)	—
Level 2–Facility Characteristics
Profit status (1 = for profit)	−1.49 (0.60)*	0.23	−0.71 (0.69)	—	−1.11 (0.76)	—	−1.48 (0.74)*	0.23
% of residents taking antipsychotics	0.08 (0.07)	—	0.18 (0.09)*	1.20	0.24 (0.09)**	1.27	0.11 (0.08)	—
CNA staffing hours	−0.05 (0.01)***	0.96	0.001 (0.02)	—	−0.02 (0.02)	—	−0.02 (0.02)	—
Intercept	7.54 (2.69)**		−1.31 (3.94)		−0.34 (3.58)		3.62 (3.28)
Random effect (intercept)	1.86 (0.78)*		3.43 (1.21)**		4.20 (1.46)**		4.26 (1.47)**
Model characteristics
AIC	1199.81		1139.61		180.41		1068.03
% correctly classified	82.6		88.4		88.5		90.9
	Apathy (N = 210)		Disruptive Vocalization (N = 198)		Repetitive Behavior (N = 95)		Sexually Inappropriate Behavior (N = 89)
	B (SE)	OR	B (SE)	OR	B (SE)	OR	B (SE)	OR
Level 1–Resident Characteristics
Age	0.01 (0.02)	—	−0.04 (0.02)*	0.97	−0.05 (0.04)	—	0.002 (0.04)	—
Gender (1 = female)	0.03 (0.46)	—	−0.28 (0.38)	—	−1.15 (0.66)	—	−2.16 (0.69)**	0.12
Length of stay	0.001 (0.00)**	1.001	0.00 (0.00)	—	0.00 (0.00)	—	0.00 (0.00)*	1.00
Cognitive ability	0.05 (0.07)	—	−0.07 (0.05)	—	−0.18 (0.08)*	0.84	0.09 (0.10)	—
Functional ability	−0.01 (0.01)	—	−0.001 (0.01)	—	0.01 (0.01)	—	0.02 (0.01)*	0.98
Level 2 – Facility Characteristics
Profit status (1 = for profit)	−1.31 (0.91)	—	−1.71 (0.93)	—	−4.28 (1.10)***	0.01	−1.02 (1.31)	—
% of residents taking antipsychotics	0.23 (0.11)*	1.25	0.09 (0.10)	—	0.26 (0.13)	—	0.25 (0.11)*	1.28
CNA staffing hours	−0.03 (0.02)	—	−0.01 (0.02)	—	−0.04 (0.02)	—	−0.02 (0.03)	—
Intercept	1.62 (3.13)		5.92 (4.15)		9.11 (4.67)		−0.85 (5.65)
Random effect (intercept)	5.94 (2.18)**		5.88 (2.17)**		2.63 (2.04)		5.30 (3.03)
Model characteristics
AIC	1126.58		1031.90		592.88		555.49
% correctly classified	90.3		90.8		92.6		97.8

Note. SE = standard error; OR = odds ratio; CNA = certified nursing assistant; AIC = Akaike Information Criterion. Models also tested effects of marital status, race, education, overall star rating, number of beds, and number of RN hours, but these characteristics were not significant and were dropped from final models.

^aOnly significant ORs are reported.

^*p < 0.05;

^**p < 0.01;

^***p < 0.001.

BPSD and Presence of a PCC Plan

Resistiveness to Care.

The odds of having a PCC plan for residents displaying resistiveness to care increased in non-profit NHs and in NHs with lower CNA staffing hours.

Agitation.

The odds of having a PCC plan for residents displaying agitation increased as the percentage of residents taking antipsychotics in the NH increased.

Aggression.

The odds of having a PCC plan for residents displaying aggression increased for residents with lower functional ability and as the percentage of residents taking antipsychotics in the NH increased.

Wandering.

The odds of having a PCC plan for residents displaying wandering increased in non-profit NHs.

Apathy.

The odds of having a PCC plan for residents displaying apathy increased as the resident’s length of stay increased and as the percentage of residents taking antipsychotics in the NH increased.

Disruptive Vocalizations.

The odds of having a PCC plan for residents displaying disruptive vocalizations increased as the age of the resident decreased.

Repetitive Behaviors.

The odds of having a PCC plan for residents displaying repetitive behaviors increased as the cognitive ability of the resident decreased and the resident resided in a non-profit NH.

Sexually Inappropriate Behaviors.

The odds of having a PCC plan for residents displaying sexually inappropriate behaviors increased for residents who were male, for residents with lower functional ability, and as the percentage of residents taking antipsychotics in the NH increased.

DISCUSSION

The current study aimed to describe BPSD that NH staff documented in care plans and identify whether PCC interventions to address these BPSD were documented. We also examined resident and facility characteristics associated with documentation of PCC plans. Overall, BPSD were common, but care plans for residents displaying these BPSD were not always person-centered as defined by our study criteria (i.e., care plans that use resident values, preferences, and interests when addressing BPSD). We also found that several resident and facility characteristics were associated with documentation of PCC plans.

Our first finding was that 83% of residents in the sample (n = 553) had at least one BPSD documented in their care plan, and resistiveness to care (44.9%), agitation (42.2%), and aggression (42%) were most common. This high rate of BPSD in NH residents is not a new finding and has been reported repeatedly in the literature (Kales et al., 2014). Resistiveness to care, however, has typically been reported at lower rates (17% to 27%) compared to agitation (17% to 67%) and aggression (11% to 77%) (Ishii et al., 2012). Our finding may differ from prevalence rates in the literature due to some behavioral assessment tools not having an individual category for resistiveness to care and this BPSD then being recorded as agitation or aggression. As noted in our study, the second and third most common BPSD noted in care plans were agitation and aggression. Repetitive and sexually inappropriate behaviors were noted much less frequently. We do not know if these behaviors actually occurred less frequently, or if the low rate of documentation indicates that staff were uncomfortable identifying these behaviors (i.e., in the case of sexually inappropriate behavior) and/or lacked knowledge of how to manage these behaviors (Torrisi et al., 2017). More research is needed on accurate recognition and follow through with documenting BPSD in care plans.

Our second finding was that although BPSD were ubiquitous in the NHs, care plans for all BPSD were not, in general, person-centered. Depending on the BPSD, up to 79% of care plans had no PCC approaches documented. This finding is consistent with an integrative review that found care plans to be generally lacking in personalization and recommended staff education on how to develop individualized care plans so they can guide care in a way that is meaningful for the resident (Mariani et al., 2017; Selbaek et al., 2013).

Our third finding was that several resident and facility characteristics were associated with documentation of PCC plans. These characteristics varied by the BPSD and begin to shed light on at-risk residents and NHs where PCC plans need to be improved.

Resident Characteristics

For a given behavior, older female residents with more functional and cognitive abilities living in the NH for shorter periods of time were found to have fewer PCC approaches documented in their care plans. It is not clear why older residents and those with better cognitive and physical ability would be less likely to have a personalized care plan for BPSD. Further, BPSD often occur during assistance with direct care, which, in turn, occurs more often for residents who have functional impairments. It may be that as the frequency of direct care assistance increases, so does the burden of these behaviors, triggering documentation in the care plan. This type of situation would contrast with residents who are more functionally able, less likely to need assistance, and therefore less likely to exhibit BPSD. A scoping review of determinants of BPSD found that younger age and lower functional ability were associated with more common BPSD (e.g., agitation, aggression) (Kolanowski et al., 2017). In the current study, the personal characteristics of older age and better functional and cognitive ability were associated with the less frequently documented behaviors (i.e., disruptive vocalizations, repetitive behaviors, and sexually inappropriate behaviors). Compared to agitation and aggression, there is less evidence of efficacious approaches for the management of some of these BPSD. Lack of staff documentation in the care plans underscores the need for research on these less common but possibly difficult to address BPSD.

Being female was associated with decreased odds of having PCC approaches for sexually inappropriate behaviors documented within care plans. Males display more physical manifestations of sexually inappropriate behavior, whereas women display more verbal manifestations (De Giorgi & Series, 2016). Verbal sexual behaviors are less likely to pose a threat/challenge to staff, accounting for less frequent strategies being documented for females.

Finally, a shorter length of stay in NHs was associated with less documentation of PCC approaches. PCC requires knowledge of the person’s needs, values, and goals, and communication of that knowledge to staff caring for the resident (Kolanowski et al., 2015; Van Haitsma et al., 2020). This information may take time to accumulate, especially in NHs that do not have care planning processes that include input from residents, family, and direct care workers starting at the time of admission to the NH (Kolanowski et al., 2015). Care plans are timelier when all stakeholders are involved and staff have efficient tools to document person-centered goals and interventions (Mariani et al., 2017).

Overall, our findings point to the need for staff education on how to develop PCC plans for all residents’ BPSD regardless of their age, functional ability, gender, or length of stay. Some BPSD may be neglected in written care plans because of the lack of evidence for appropriate approaches. More research is needed to provide staff with evidence-based approaches for those behaviors.

Facility Characteristics

The type of NH and its characteristics impacted the documentation of PCC. We found that residents in NHs with lower use of antipsychotics had fewer PCC approaches documented for apathy, sexually inappropriate behaviors, agitation, and aggression. This was an unexpected finding and one that is difficult to interpret given that low rate of antipsychotic use is designated as a NH quality indicator. Rate of antipsychotic medication use, however, does not necessarily translate into lack of PCC planning. A study by Bonner et al. (2015) found that NHs with high antipsychotic medication use involved consultant psychiatrists more often than those with low antipsychotic medication use. Staff in facilities with higher antipsychotic use for BPSD may be challenged with significant behavioral issues and sensitive to the need to also include nonpharmacological PCC approaches in their care plans. We did not measure type and number of consultations, but the effect on care planning deserves exploration in future studies.

For-profit NHs had fewer documented PCC approaches for repetitive behavior, resistiveness to care, and wandering. This finding is in line with prior literature that indicates non-profit NHs outperform for-profit NHs in terms of higher quality of care delivery, higher staffing levels, and lower staff turnover rates (Bos et al., 2017; You et al., 2016). This finding also raises the important issue about the need for greater investment in staff education in for-profit NHs as a way to improve resident outcomes.

Lastly, we found that having higher CNA staffing hours was associated with less documentation of PCC plans for resistiveness to care, the most common behavioral symptom in participants. CNAs provide most of the direct care to residents but their contributions to the care planning process have historically been omitted (Kolanowski et al., 2015). It is RNs who have the legal authority and responsibility for resident assessment and care planning (Burshnic et al., 2018). Although RN staffing hours was not found to be associated with greater documentation of PCC in the current study, other work has supported the importance of RNs who have leadership skills to elicit, share, and use relevant resident information from all staff to improve outcomes in NHs (Mueller et al., 2016).

LIMITATIONS

This secondary analysis was limited by the resident and facility characteristics available for statistical modeling because the parent study was not designed to specifically examine all factors that might impact care plan documentation. For instance, we did not collect data on staff training for developing PCC plans or staff turnover of those developing and updating the care plans. The sample size was large, although drawn from NHs in only two East Coast states. This study did not include a validation of what was found in the care plans; therefore, there is no way to know if the plans were implemented. In fact, there can be discrepancies between what is in the care plan and what care is actually delivered (Mariani et al., 2017). In addition, we included only those residents with moderate to severe dementia. Individuals with mild to moderate impairments may have had different expressions of BPSD and care plan documentation than those in this study.

IMPLICATIONS

This preliminary study has illuminated the need for additional research and staff education on how to develop PCC plans. Future research needs to consider whether the plans of care are actually implemented and evaluated for effectiveness at the resident level and what impact they have on residents’ quality of life. Our inquiry illuminates the need to educate NH staff on ways to develop PCC plans for BPSD. Experts have recommended the DICE intervention (Describe the behavior, Investigate its potential cause[s], Create a solution [approach/intervention] that addresses the cause[s], and Evaluate the response to the approach/intervention) for personalizing nonpharmacological BPSD care approaches (Kales et al., 2014).

The integration of idiographic and nomothetic approaches is ideal for PCC planning (Cohen-Mansfield & Mintzer, 2005). A nomothetic approach uses empirically based evidence and expert guidelines based on aggregate data to inform clinical judgments on intervention selection. Standardized care plans are largely based on this approach. An idiographic approach involves a multimethod assessment that informs an individualized care plan (Haynes et al., 2009; Mumma, 2001). This approach includes capturing the symptom(s) exhibited, monitoring change across time and context, and obtaining data from multiple sources to capture unique aspects of BPSD (Haynes et al., 2009; Mumma, 2001). Table A (available in the online version of this article) includes resources on care planning, and Table B (available in the online version of this article) is an example of a care plan developed by the authors to highlight the incorporation of PCC interventions.

CONCLUSION

BPSD are frequently documented in care plans but PCC approaches for these symptoms are less common. In general, care plans are more personcentered when the resident is younger, has lower functional ability, is male, and resides in the NH for a longer period of time. Non-profit NHs, those with lower CNA staffing hours, and a higher rate of antipsychotic use are more likely to have PCC plans developed for residents with BPSD. These study findings highlight the need for more staff education and research on documentation of PCC approaches for BPSD.