| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to Julie Robison, PhD, Center on Aging, University of Connecticut Health Center, 263 Farmington Avenue, Farmington, CT 06030-6147. E-mail: jrobison{at}uchc.edu
 |
Abstract |
|---|
 TOP Abstract Design and MethodsResultsDiscussionReferences |
|---|
Key Words: Nursing home • Intervention • Family • Nursing staff • Dementia
From the perspective of nursing home staff, working successfully with residents' families is one particularly stressful aspect of their jobs (Cohen-Mansfield, 1995; Heiselman & Noelker, 1991; Pillemer, Hegeman, Albright, & Henderson, 1998). Staff members generally report having inadequate time for substantive or meaningful discussions with family members (Pillemer, 1996). These obstacles may be compounded by racial, cultural, and socioeconomic differences between staff and residents and their families that may result in negative stereotypes (Specht et al., 2000). Some families perceive the staff as untrustworthy (Krause, Grant, & Long, 1999; Tobin, 1995) and feel they need to monitor staff behavior (Duncan & Morgan, 1994; Stull, Cosbey, Bowman, & McNutt, 1997). Staff members believe families may have unrealistic expectations of nursing home care (George & Maddox, 1989; Heiselman & Noelker; Hertzberg, Ekman, & Axelsson, 2003).
Despite these challenges, evidence indicates that family involvement in the care of a relative living in a nursing home can be beneficial for residents, family members, and staff. Resident care may be optimized when communication between families and staff is consistently effective and working relationships are strong. Meaningful participation in the care of a relative living in a nursing home has been shown to improve family members' satisfaction, well-being, and experience of care (Gaugler, Andersen, Zarit, & Pearlin, 2004; Greene & Monahan, 1982; Maas et al., 2004; Penrod, Kane, & Kane, 2000; Tornatore & Grant, 2004). Nursing homes with policies that reflect a strong family orientation have been linked to positive outcomes such as family involvement and cooperative staff relationships (Friedemann, Montgomery, Maiberger, & Smith, 1997). Evidence indicates that good relationships with families are a key indicator of staff job satisfaction and intention to stay in the job (Robison & Pillemer, 2005).
Family and Staff Communication for Residents With Dementia
Effective communication between families and staff is critical for nursing home residents with dementia. Residents with cognitive impairments may be unable to express their concerns, needs, and preferences effectively. Although family members must depend on staff for information about their relative's life in the nursing home, they also possess important knowledge about residents' psychosocial, physical, and emotional histories that is essential for developing appropriate individualized care plans (Port et al., 2001). Unfortunately, research with both staff and families indicates that the sharing of detailed information about residents is often inadequate (Ekman & Norberg, 1988; Safford, 1989). Family support and involvement can assist staff in reducing residents' behavioral symptoms by helping to identify unmet medical, emotional, or social needs (Foley, Sudha, Sloane, & Gold, 2003; Radar & Tornquist, 1995). Substantive family participation also increases resident involvement in activities (Dobbs et al., 2005).
Interventions in Family and Staff Communication
Existing intervention models to enhance the quality of relationships among families and staff have generally focused exclusively on family members (Maas et al., 2004; Peak, 2000). For example, staff in one intervention study reported improved attitudes toward families after a family-based intervention (Maas et al.). Another intervention study, designed to help families partner with staff in caring for a family member with dementia, found that global deterioration of residents was slowed for those whose family member had participated in the training (Jablonski, Reed, & Maas, 2005). Few programs, however, seek to promote cooperation and communication through concurrent training and support of both families and staff (Davis & Buckwalter, 2001). Further, the critical role of institutional and administrative support has not been systematically examined.
Partners in Caregiving
Recognizing the importance of including nursing staff and administration in efforts to facilitate family participation in care, researchers from the Cornell Gerontology Research Institute, the Foundation for Long Term Care, and the Center on Aging at the University of Connecticut Health Center developed a program to promote family and staff cooperation, called Partners in Caregiving (PIC; Pillemer et al., 2003; Robison & Pillemer, 2005). The PIC program fosters positive relationships between families and staff through intensive training on effective communication, empathy development, and conflict resolution. PIC trains both families and staff, and then it guides families, staff, and facility administrators in a collaborative examination of facility procedures and policies that have an impact on family–staff interactions. A randomized, controlled evaluation of the PIC program demonstrated improved attitudes for both groups, a reduced intention to quit for staff, and less family conflict with staff, specifically for families of residents with dementia (Pillemer et al.).
Partners in Caregiving in the Special Care Unit Environment
On the basis of the positive outcomes in the original PIC program for families of dementia residents, researchers adapted the PIC program for use on dementia special care units. Partners in Caregiving in the Special Care Unit Environment (PIC-SCU) draws heavily from the original program, utilizing the same structure and considerable overlap in curricular content. Program modifications in PIC-SCU include a new module on understanding behavioral symptoms as well as numerous case studies focused on residents with dementia.
Like the original program, the PIC-SCU program has two primary sequential components. First, parallel training sessions are provided to both family and staff, designed to enhance communication techniques and develop conflict-resolution skills and empathy for the other group. Communication techniques focus on developing active listening skills and providing constructive feedback. The training includes a section on how cultural, racial, and other differences (e.g., socioeconomic status) can affect communication. Four primary training methods are used throughout the workshops: "mini-lectures," case discussions, brainstorming sessions, and role plays. The program is detailed in a comprehensive training manual that contains directions for facilitating each of the sessions, descriptions of training activities, and master copies of handouts. The staff and family training sessions are between 4 and 5 hours in length. Finally, upon completion of the training, a meeting with families, staff, and nursing home administrators is held to set concrete goals for the unit and facility regarding procedures and policies that affect families. This 2-hour session brings staff and families together to discuss issues of concern with the facility administrators. The joint meeting is carefully structured and includes opportunities for sharing ideas, as well as prioritizing policy changes (see Table 1).
|
|
Design and Methods |
|---|
|
TOPAbstractDesign and MethodsResultsDiscussionReferences |
|---|
Participants
The PIC-SCU intervention study included 20 facilities in three Connecticut counties, which researchers randomly selected from all skilled nursing facilities with special care units or dementia programs in those counties. Sampling 20 facilities provides generalizability to a larger population of facilities and gives researchers the ability to estimate facility variances with acceptably small sampling variances. It also provides a number of individuals within facilities for a sample with sufficient statistical power in the mixed models.
Research staff conducted brief telephone interviews with all skilled nursing facilities in the identified geographic area to determine which facilities had dementia programs. Twenty-eight facilities met the criteria; researchers randomly selected 20 of these for recruitment. Three of these facilities declined to participate; researchers then randomly selected 3 more facilities, which agreed to participate.
Researchers randomly assigned the 20 participating facilities to control and intervention groups, stratified by proprietary status. Three project trainers, all with extensive group leadership experience, conducted the PIC-SCU workshops in the 10 intervention facilities. After researchers had collected all evaluation data, they provided program materials and training to the control facilities.
Researchers recruited all members of the nursing staff (i.e., registered nurses or RNs, licensed practical nurses or LPNs, and certified nurse's aides or CNAs) and one family member for each resident identified by the facility social work staff to complete interviews and, in the treatment facilities, to attend the workshops. Researchers included all eligible relatives, many of whom lived out of state or visited infrequently. In a few cases in which multiple family members for a single resident asked to attend the workshops, only one person per resident completed interviews for this study.
Research staff conducted telephone interviews with nursing staff and families at the same three time points (baseline and 2 and 6 months after training). Because of the unit-based nature of the intervention, research staff attempted to have interviews with all families and staff members on the unit, whether or not they participated in the workshop.
Staff members who left employment in the nursing home after Time 1 received an abbreviated Time 2 interview and were not interviewed at Time 3. If they left after Time 2, they did an abbreviated Time 3 interview. Family members whose relative died or left the facility between Time 1 and Time 2 also received an abbreviated Time 2 interview and were not interviewed at Time 3. If they left or died after Time 2, they completed an abbreviated Time 3 interview.
Facility staff completed behavioral symptom checklists for all unit residents at baseline and at the 2-month follow-up time. Research staff did not do a resident follow-up at the 6-month time point because of potential confounding from disease progression over this time. Facility administrators completed a structured questionnaire about the facility at baseline and at the 6-month follow-up. Finally, exploratory, qualitative interviews with 9 of the treatment facility administrators after 6 months assessed their success in meeting the goals established at the joint meeting.
Measures
Table 2 displays the reliability coefficient (alpha), range, mean, and standard deviation for each of the family and staff measures used in the analyses.
|
Staff Measures
Using questions identical to those asked of family members, researchers measured two staff outcomes: the Interpersonal Conflict Scale, with the word family substituted for staff, and the CES-D scale. An additional measure for staff was the Family Behaviors Scale, which measures staff perceptions of how often family members treat them with respect, act rudely, smile and greet them, or ignore them (Pillemer et al., 1998). The Family Empathy Scale asks staff the degree to which family members understand how much time it takes to do the staff members' job, are mostly concerned about their own needs, and are sensitive to staff members' feelings, on a scale of never, rarely, sometimes, or almost always (Pillemer et al., 2003). Researchers measured job burnout by using the Depersonalization subscale of the Maslach Burnout Inventory (Maslach, 1982), modified by Pillemer and Moore (1989) for use with nursing home staff. This measure includes six statements and answer categories of strongly agree, agree, disagree, or strongly disagree. Staff completed an adapted version of the Generic Job Satisfaction Scale (MacDonald & MacIntyre, 1997), with 11 statements and answer categories of strongly agree, agree, disagree, or strongly disagree. Researchers measured the likelihood that staff would leave employment in the nursing home by means of a single item, which asked the following: "Thinking about the next 12 months, how likely do you think it is that you will decide to quit your job?" (very likely, likely, not too likely, or not at all likely). Another single item measured overall perception of job stress (very, a little, somewhat, or not too stressful).
Resident Measures
Nursing staff completed the short form of the Cohen-Mansfield Agitation Inventory (CMAI; Werner, Cohen-Mansfield, Koroknay, & Braun, 1994) for each resident in the unit to assess 14 behavioral symptoms of dementia. In addition, researchers tracked the resident transfers off the unit and out of the facility after baseline.
Facility Measures
Each facility administrator completed a structured questionnaire reporting facility-level factors that could be affected by the intervention at Times 1 and 3. These measures included the number of specific family programs in the facility and whether or not they provided specialized training to staff in working with family members. Administrators rated the unit staff's level of training and skill in dementia care on a 5-point scale and estimated the percentages from 0 to 100 of very involved families as well as inactive families. Finally, for the past 6 months, they reported RN or LPN and CNA turnover rates, whether they had any reported serious problems or negative incidents involving family members, and the number of family complaints for the unit.
In addition to having the administrators complete the structured facility-measures questionnaire, researchers conducted qualitative, open-ended interviews with intervention facility administrators at Time 3 to characterize their perspectives on the impact of the program at the institutional level. The administrators reported on changes to facility programs and policies made in response to the prioritized group goals that were set during the program's joint meeting between families, nursing staff, and administrative staff.
Statistical Analyses
Analyses focused on treatment group differences for the family, staff, resident, and facility outcome measures as already described. We examined the effects of the PIC-SCU program by comparing change in the intervention group from before the workshops (baseline) to after the workshops (2 or 6 months) to change in the control group over the same time period. A 2 x 3 classification structure (treatment by time) forms the core of the statistical models for evaluation. The key tests of whether the intervention has had an effect are the test of the Time x Treatment interaction and specific, preplanned contrasts partitioned from that interaction.
For the family and staff outcomes, which were assessed at all three time points, we evaluated the program in terms of its effects at each of the two follow-up assessments and for effects averaged over those two assessments. Because the program is of short duration and moderate intensity, effects might be expected to be strongest at the first follow-up, nearest to the time of intervention. The examination of the second follow-up gives an indication of the extent to which effects continue over a longer period of time. It is also possible that certain effects may take a longer time after involvement with the program to be manifested and will be seen only at Time 3. The test of effects averaged over the two follow-up assessments provides the best evidence of beneficial consequences of the program that are sustained over time.
We carried out all analyses in general linear mixed models (Henderson, 1982; Searle, 1971), with time included as levels of a fixed factor and individuals as levels of a random factor. We included facilities in the model as levels of an additional random classification factor, taking into account variance associated with facilities and that families and staff members are grouped by facilities. It allows inferences to the population of facilities from which we sampled (see Pillemer et al., 2003, 2006, for examples). We performed our analysis with SAS 9.0 software.
We were interested not only in detecting any overall effects of the intervention but also in whether the effects of the intervention were stronger for groups with particular needs or stressors. For families, we hypothesized that spouses might be affected differently by the intervention than adult children. We therefore tested for an interaction effect of the family member's relationship to the patient (spouse or sibling vs adult child or grandchild). Relationships with family members are likely to differ for nurses and CNAs. Thus, for staff, we examined the effect of being a nurse (RN or LPN) versus being a CNA on staff outcomes.
Our analysis of the resident and facility outcomes, which were each assessed at two of the time points, tested differences between intervention and control facilities from baseline to the post-test. These models used a 2 x 2 repeated measures design (Intervention x Time).
We compared all family and staff workshop participants from the intervention facilities to all respondents from the control facilities. We did not include intervention facility families and staff who completed interviews but did not attend workshops in these analyses. Thus we used an efficacy subset analysis, rather than intention to treat. This methodology is appropriate for this program, which is still in a preliminary testing phase (Gross & Fogg, 2004). In this case, it is appropriate to report results for the group that shows the most significant effect of the intervention and it is clear that inferences are limited to the group that actually participates in the intervention. The impact of the intervention was in fact stronger for those individuals who actually attended workshops.
|
Results |
|---|
|
TOPAbstractDesign and MethodsResultsDiscussionReferences |
|---|
Interviews were attempted with 491 family members, resulting in 388 respondents at Time 1, for a completion rate of 79%. The retention rate between Time 1 and Time 2 for family members was 92%, resulting in 355 Time 2 interviews. Time 2–Time 3 retention was 87%, resulting in 325 Time 3 family interviews (Table 3).
|
|
|
|
|
When we examined generational differences for the family participants, spouses and other same-generation family members in the intervention group increased their facility involvement in the short term, whereas control family involvement declined. The family involvement of the adult children and other younger generation relatives did not differ significantly between the treatment and control groups.
Effect of the Intervention on Staff Outcomes
Staff intervention participants benefited from the PIC-SCU program on two outcomes, when compared with control staff. Whereas conflict with families decreased for intervention staff after the workshops, conflict rose for control staff, in the short term (Table 8). Furthermore, by 2 months, treatment staff showed a trend to a decline in feelings of depression, but control staff's depression symptoms increased. By 6 months, control and intervention groups no longer differed on either measure.
|
Impact of the Intervention on Resident Behaviors Over Time
Nursing staff assessed resident behaviors at baseline (n = 561) and then 2 months after the training (n = 559) by using the short form of the CMAI. Five of the 14 behaviors improved significantly (p 
.05) more for the treatment than for the control group: cursing or verbal aggression; other aggression, self-abuse, or sexual advances; inappropriate dress or disrobing; constant requests for attention or help; and wandering. Two additional behaviors, grabbing people or destroying property and being restless, showed a trend (p <.10) toward more improvement for the treatment group (Table 9). In 5 of the remaining 7 behaviors, the treatment group showed more improvement than the control group, but not at a statistically significant level. There were no differences in the number of transfers off the unit or out of the facility between treatment and control residents over the 6-month follow-up (data not shown).
|
Intervention Workshop Evaluations
All intervention participants completed an evaluation of the program immediately following the PIC-SCU workshops. Combined family and staff ratings of the program were very positive. The PIC-SCU training focuses in part on resolving conflict in the nursing home. Respondents rated how well prepared they felt to tackle difficulties in working with the other group after the workshop: 79% indicated they now felt well-prepared and the remaining 21% felt at least somewhat well prepared. Almost all respondents gave a positive overall evaluation to the program, with 63% describing it as excellent and 35% as good. Nearly all participants (98%) reported that the materials covered in the workshop relate to their experiences in the nursing home. Finally, 100% of participants indicated that they would recommend the program to others.
PIC-SCU Joint Meeting Goals
Follow-up calls to nine of the intervention facility administrators gathered qualitative data on their perceptions of program or policy changes on their special care unit, which they attributed to their participation in the PIC-SCU program. All nine administrators felt that they had either met or exceeded the top priority goals set by the group (families, staff, and administrators) at the joint meeting. Three representative examples of their responses follow:
We added a new family handbook; we have improved communication through better notification of events at (the facility); use of community bulletin board and notices on units; collaborative effort to draft a new newsletter; increased use of name tags; more weekend activities; better coordination of education programs for families; improved communication of education programs; added a new orientation program for new families; at least 2 education programs for families a month; added an evening support group; added more meetings with families to raise general concerns (flexible meeting dates and times).We expanded the lounge to double size to facilitate programs and dining; implemented more dementia-specific programs; put in place a more stimulating decor; notify families of availability of telephone conferences which are now being used; started a family monthly newsletter and a staff newsletter.
Staff recognition (program) for those nominated by families; formal family invites to resident care conferences; laundry reclaim day scheduled to recover lost clothing; initiation of "color scapes" dementia training for staff; additional recreational hours added to the SCU.
The joint meeting goals were diverse and facility specific, responding to the nature of the issues raised by families and staff during that meeting. The meeting served to organize and prioritize policy and program changes, specifying the people responsible for overseeing the changes and a timeline for achieving them.
|
Discussion |
|---|
|
TOPAbstractDesign and MethodsResultsDiscussionReferences |
|---|
The present study went beyond prior research on PIC by examining resident outcomes. Although not directly involved in the PIC-SCU intervention, residents appear to have derived benefit from the participation of their families and unit staff. Residents' behavioral symptoms clearly improved in the treatment facilities over the 2-month follow-up period. If confirmed by future research, this effect suggests additional benefits of the PIC model, beyond staff and family outcomes. One possible mechanism for the effects is changes at the facility level resulting from family and staff suggestions. Intervention facilities increased the number of programs they offered to families, whereas the control facilities had slightly fewer family programs in the same 6-month period. Alternatively, improved communication between families and staff may have led to better management of behavioral symptoms.
Evaluations from the groups who participated in the PIC-SCU program were consistently positive. Families and staff gave uniformly high assessments of the usefulness, relevance, and their overall satisfaction with the program. Finally, facility administrators who had participated in the joint meetings enumerated multiple individual policy and programmatic changes for their dementia units, which they attributed directly to their participation in the PIC-SCU program.
Although these findings are encouraging and suggest that continued adaptation and dissemination of the PIC-SCU model is worthwhile, there are several limitations to the study. Not surprisingly, treatment effects were not consistently demonstrated on all outcomes, and some effects were modest in size. In some instances, positive effects diminished over time, suggesting a need for adding periodic "boosters" to the program design. This was also the case for the original PIC program (Pillemer et al., 2003).
The intervention primarily affected only those family members who participated in the actual workshops, as shown in this efficacy subset analysis. If we include those with no exposure to the intervention in the analyses, as with intention to treat criteria, we see few significant effects. As Gross and Fogg (2004) suggest, conducting both intention to treat and efficacy subset analyses in the same study sheds light on the issue of adherence in predicting the program's impact. When advocating wider spread use of the PIC-SCU program, expected participation rates of approximately 50% for family members should be noted and attention should be paid to participant recruitment techniques.
Statistical power is somewhat limited in a sample of this size. It is not likely that larger sample sizes would have resulted in more significant relationships, however, because mean differences for nonsignificant outcomes are not large and test statistics did not approach significance.
Although most of the staff and family scales demonstrated excellent internal consistency, a few had relatively low alphas (e.g., the Family Empathy Scale had 
= 0.62). This evaluation might also have benefited from a "placebo" group, which would help to determine whether the specific contents of the intervention are effective or whether having any type of arrangement that includes staff and family meetings would have a similar effect.
Although ultimately resident care and quality of life are a desired outcome, we employed an indirect measure of effects on residents. Nevertheless, the CMAI is generally accepted as a reliable indicator of behavioral symptoms, and as a proxy for resident distress and unmet needs. Finally, we did not examine changes in facility policy or programs that arose as a result of the joint meeting planning process, but rather relied on administrator reports, which may be biased or incomplete.
An important next step in the refinement of the program evaluation is to develop the administrative and institutional policy components. Although the anecdotal evidence in this study suggests that the PIC-SCU program had positive influences on the organizational environment, this effect should be examined in a systematic way. Methods could include structured interviews with key administrative personnel and direct-care workers, review of institutional policies and procedures, content analysis of minutes from family council meetings, as well as other qualitative approaches that are well-suited for characterizing organizational change (Eccles, Grimshaw, Campbell, & Ramsay, 2003; Sofaer, 1999). A further step in future efforts is developing modifications of the PIC program for other environments, such as assisted living, home care, or group homes for adults with intellectual disabilities.
The Partners in Caregiving model represents one of only a handful of evidence-based interventions with demonstrated effectiveness both in general nursing home units and in the dementia-specific care environment. The program offers demonstrable benefits to families, staff, and residents. Participants find the program helpful and enjoyable. Complete, accessible program manuals are available. Wider dissemination could be accomplished in partnership with nursing home provider associations and advocacy organizations that support educational programming for families or nursing home staff.
|
Footnotes |
|---|
1 Center on Aging, University of Connecticut Heath Center, Farmington, CT.
2 Department of Human Development, Cornell University, Ithaca, NY.
Decision Editor: Linda S. Noelker, PhD
Received for publication July 24, 2006. Accepted for publication January 25, 2007.
|
References |
|---|
|
TOPAbstractDesign and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. Pillemer, R. Meador, C. Henderson Jr., J. Robison, C. Hegeman, E. Graham, and L. Schultz A Facility Specialist Model for Improving Retention of Nursing Home Staff: Results From a Randomized, Controlled Study Gerontologist, July 1, 2008; 48(suppl_1): 80 - 89. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
