Feasibility of the intervention
The feasibility of the intervention relates to the degree to which the participants enrol in, complete, and comply with the intervention [
46]. Despite concerted efforts, the 29% enrolment rate in this study was lower than the 45-64% rates reported in other similar studies [
3,
86,
90,
92], a difference that may be related to the nature of the intervention or the target population. Our lower enrolment rate could also be related to older adults’ reluctance to report depressive symptoms. In general, the reasons given for non-participation were not related to the nature of the intervention. The challenges of recruiting older adults with depression are well documented [
119,
120]. To address recruitment barriers, we used clear but simple communication, gave the client the time needed to decide, had a clear protocol for contacting potential participants and flexible scheduling, and educated family caregivers about the study. Our efforts confirm that, although recruitment was labour-intensive and difficult, it was achievable in this client group.
We had attrition rates of 31% at six months and 39% at one year, which are comparable to those reported in other similar studies [
86]. Client characteristics, organizational barriers, and client preferences contributed to attrition in our study. First, 16% of participants died over the course of the study, while 18% discontinued the study because of health problems or lack of interest, and the remainder were difficult to contact. Again, the reasons given for withdrawal were not related to the nature of the intervention.
Attrition has been recognized as a factor that threatens internal validity and reduces statistical power in a study. To minimize attrition, the study coordinator used a participant-tracking plan and the interviewers built rapport and trust with the participants and maintained between-assessment contact [
121,
122]. Participants were also compensated for their time ($15 for the baseline interview and $10 for the six-month interview). Case manager recruiters and interviewers met monthly with the principal investigator to clarify recruitment and data collection procedures, identify problems, and make suggestions for improvement. Attrition may contribute to self-selection bias when the characteristics of individuals who withdraw from the study differ from those of individuals who complete the study [
123]. In our study, drop-outs were a somewhat lower-functioning group than those who were retained in the study.
Compliance with the intervention was operationalized as: (1) the dose of the intervention, defined as the number of home visits and case conferences, and (2) the level of fidelity to treatment (i.e., the extent to which the RNs and PSWs adhered to the components of the intervention). About 88% of older adult participants received at least one home visit by either the study RN or PSW during the six-month intervention period. Of that number, 73% completed the six-month intervention and 27% withdrew. This completion rate is comparable to those reported in other similar studies [
3,
86,
90,
92]. The delivery rate of the components of the intervention ranged from 28% for development of an IP depression management plan to 78% for depression education. In addition, 71 to 84% of participants were screened for depressive symptoms, cognitive impairment, anxiety, and delirium. Variations in the delivery rate of these components may reflect tailoring of the intervention to individual clients, providers, and settings. The smaller time commitment to the development of an IP depression management plan could have been related to the smaller number of RNs compared to PSWs on the team. Indeed, the RNs reported heavy workloads and limited time as barriers to implementation of the intervention. Future research exploring the composition and distinct roles of the members of the nurse-led teams, including optimal RN:PSW ratios is warranted.
Several strategies were implemented to monitor and enhance fidelity of intervention implementation. These included: (1) monthly
audits of the study documentation to assess fidelity, (2) monthly
outreach visits with the intervention providers, and (3) scheduled
reminders and updates[
108]. These strategies proved to be effective in identifying problems related to implementation, clarifying the intervention protocol, and developing suggestions for improving fidelity of intervention implementation.
Acceptability of the intervention
Acceptability was defined as the older home care client and study RNs’ and PSWs’ perception of the intervention’s appropriateness, benefits, and convenience of implementation [
109]. Overall, the participants viewed the nurse-led intervention as highly acceptable and were able to describe a full range of benefits. Perceived benefits from the client perspective centred on the personal attributes of the providers (caring and competent) and positive results (decreasing depression, improving function, instilling hope, increasing confidence, increasing knowledge of depression, and facilitating access to other services and supports). Older adult participants also indicated that caring, emotional support, reassurance, and encouragement were key factors in the treatment of their depression. This was supported by the qualitative feedback from the providers. These findings suggest the need to include a relational measure of some kind in future studies to quantify the impact of this important aspect of the intervention on the outcomes. Family caregivers indicated that they also valued the recognition and support they received as a result of the intervention. This finding suggests that future IP nurse-led DCM interventions should include a family satisfaction measure of some kind to capture the impact of the intervention on family caregivers.
Providers highlighted the benefits of: (1) regular in-home visits, (2) IP collaboration and teamwork, (3) increased knowledge of depression assessment and management, (4) the RN working collaboratively with the PSW, (5) improved communication and collaboration among home care, primary healthcare, and specialized mental health professionals and services, (6) improved recognition and management of depression, and (7) access to timely primary care and follow-up management.
Key barriers to implementing the nurse-led intervention included: (1) lack of home care provider knowledge and skills in depression assessment and management in older adults with complex chronic conditions, (2) addressing the stigma of mental illness, (3) heavy workloads and limited time, (4) communication barriers among the RN, PSW, home care CM, and PCP, and (5) limited access to personal support and other home care services. Recommended solutions to improve the sustainability of this approach to care delivery included: (1) expanding eligibility criteria, (2) providing additional staff education about depression care management, (3) improving communication among home care providers, and (4) developing a depression care pathway. These findings are similar to those of recent research on implementing evidence-based mental health programs in community settings [
89]. Our study suggests that future IP nurse-led DCM interventions should incorporate these strategies to improve the effective translation of this evidence-based approach to care.
Effects of the intervention
This study provides initial evidence for the feasibility, acceptability, and sustained effects of the intervention in improving client outcomes, reducing use of expensive health services, and improving clinical practice behaviours of home care providers. Our results extend those in the literature and current translation of evidence-based depression care in several ways.
First, the IP nurse-led mental health promotion intervention proved to be feasible and effective in reaching our target group - older home care recipients with depressive symptoms. The baseline rate of clinically significant depressive symptoms of 56% in the present sample greatly exceeds the 8.5-47% rates reported for representative samples of older home care recipients [
1,
4,
8,
25,
61,
92]. Our study deliberately recruited only seniors who had an increased risk for depressive symptoms; in fact, their rate of depressive symptoms was closer to the rate reported among institutionalized older adults [
124]. A key issue is not simply the high prevalence of depressive symptoms, but rather the combination of depressive symptoms and high rates of co-morbidities. Almost all the older adults in our study had multiple (3 or more) chronic conditions and 68% had six or more. This situation is cause for concern, given that depression in the context of multiple chronic conditions is associated with increased medical symptom burden, functional impairment, and poor adherence to treatment, increasing the probability of adverse health outcomes and increased healthcare utilization and costs [
125]. Overall, these findings underscore the important role of home care in the screening, early identification, and management of depression in this vulnerable population.
Of the 142 eligible consenting older home care clients, 103 (73%) had clinically significant depressive symptoms or were taking an antidepressant medication. Of that number, only 22% were adequately treated, a rate consistent with the 20-30% rates observed in other studies [
1,
8,
12]. These findings are noteworthy, given that untreated or under-treated depression is associated with greater morbidity and dependency, functional decline, diminished HRQoL, pain [
14], poor adherence to medical treatment, increased demands on family caregivers, premature nursing home admissions [
16], increased use of healthcare services [
2,
3,
9,
15‐
18], and increased risk of premature death from suicide and other medical conditions [
19].
A key strength of our study was that the two-step screening and recruitment process identified many older adults who would not normally have received any treatment for their depression. This finding highlights the importance of incorporating depression screening, using a validated screening tool, into routine clinical practice for older adults requiring PSS. Based on the results of this study, screening should include evaluation of risk factors for depression, particularly anxiety, stressful life events, and a history of depression.
Second, the results of this study add to the growing evidence for the effectiveness of an IP nurse-led depression care management intervention for community-living older adults in reducing depressive symptoms [
41,
66,
69,
72,
76,
79,
81‐
84,
86] and improving HRQoL [
41,
42,
70,
76,
79]. As we hypothesized, the IP nurse-led mental health promotion intervention was effective in reducing depressive symptoms at the six-month follow-up, with a small additional improvement six months after the intervention. The 20% (3.5 point) reduction in the mean CES-D depressive symptom score at one year is comparable to that reported in other DCM trials involving community-living older adults [
41,
42,
66,
69,
72,
76,
79,
81‐
85],[
87,
91,
92] and is clinically meaningful [
110]. This difference in the CES-D score translated into an impressive 62% reduction over the study period in the proportion of clients with clinically significant depressive symptoms.
Our study enrolled older adults with any level of depression severity, cognitive impairment, and other comorbid health conditions, including people who are often excluded from community-based studies. Almost one-third (31%) of the sample had dementia. Thus, this study makes an important contribution by providing knowledge of the effectiveness of a nurse-led mental health promotion intervention among a more vulnerable group of older home care clients. The use of less restrictive selection criteria increased the heterogeneity of the sample, reflecting the variability in older home care recipients seen in everyday practice. It also enhanced the generalizability and clinical applicability of the research findings to include older adults at risk of, suffering from, or recovering from depression [
46,
126].
An important finding of this study was that it provides preliminary evidence for the effectiveness of the intervention among older home care clients with dementia. This finding is particularly noteworthy given that depression in clients with dementia frequently remains undiagnosed or the depression is considered to be an inevitable and untreatable consequence of dementia. Our findings are consistent with those of previous studies that have shown that dementia in clients with depression does respond to treatment, and appropriate therapy can improve the well-being of these patients [
127]. These findings suggest that future IP nurse-led DCM interventions should target older home care clients with dementia.
A novel finding of this study was the long-term maintenance effect of the intervention in reducing depressive symptoms. This result is particularly meaningful, given the chronic and recurrent nature of depressive symptoms in this population. Most studies on the effectiveness of DCM interventions have analysed only the immediate effects of the intervention. Our findings suggest that clinical benefits continue to accrue well beyond the intervention period. Other studies have reported similar positive long-term effects on depressive symptoms [
41,
69] but they involved a nurse working in collaboration with an IP team, not leading an IP DCM strategy.
Another important finding of this study was the identification of three variables at baseline that predicted the severity of depressive symptoms at the six-month follow-up: anxiety, recent stressful life event, and history of depression. Little is known about the mechanism through which existing DCM interventions improve depression outcomes. Our study suggests that future IP nurse-led DCM interventions should give special attention to these factors as means of enhancing the effects of the intervention in reducing depressive symptoms. Our findings also suggest that limited home care resources may be used more effectively if targeted toward older adults with these characteristics.
Third, as expected, the intervention that reduced depressive symptoms also produced significant improvements in HRQoL. Given the lower level of HRQoL in older adult participants at baseline, this is a clinically important gain. Our findings are consistent with those of previous studies [
41,
42,
70,
76,
79]; however, a novel aspect of our study was that the intervention effects on HRQoL were sustained six months after the intervention period. The association between depressive symptoms and HRQoL is well documented in the literature [
70].
Fourth, a unique aspect of our study is that it included anxiety as an outcome, which is especially relevant in light of the high rate of co-morbidity between depression and anxiety [
84], 42% in our sample, and the finding that anxiety at baseline was a risk factor for depressive symptoms. The intervention resulted in a 49% reduction over the study period in the proportion of clients with anxiety. It might be beneficial for future interventions to specifically target anxiety as a means of reducing depressive symptoms and enhancing the HRQoL of this vulnerable population.
Fifth, our study showed that these improvements in client outcomes were achieved at no additional cost to society as a whole, thus making the intervention highly desirable, given its clinical benefits. Previous studies that included an economic evaluation focused only on the use of hospital, emergency room visits, home care services, antidepressant medications, primary care, and specialty mental health services as measures of cost [
73,
74,
86,
93,
95]. Our study is unique in that it measured use and costs of a full range of health services, from a societal perspective. Although our results showed no significant difference in the total mean costs of use of health services (including the nurse-led program costs), there were significant reductions in the costs of use of specific types of health services, such as acute hospitalization, ambulance services, and emergency room visits. Previous studies have also reported reductions in hospitalization with a collaborative IP DCM approach [
73,
74,
86,
95]. The $9,126 reduction in per-person costs of use of hospitalization by itself creates more than enough savings to pay for the intervention. In Canada, hospital costs constitute the largest component of healthcare expenditures for depression, at approximately $3.8 billion dollars per year [
128].
Sixth, the study results support and extend the literature regarding best practice guidelines for the prevention and management of depression in older adults with depressive symptoms. Given the multifactorial nature of depression, older adults with multiple chronic conditions are best served by an IP team of professionals and non-professional PSWs with complementary skills to address the biopsychosocial determinants of depression. Our results support the need for a chronic disease management approach to depression that (1) involves an IP team; (2) targets individuals at risk [
37]; (3) involves intensive RN case management and community navigation to facilitate timely access to services and supports [
45,
48]; (4) includes regular in-home visits by an RN and PSW; (5) encourages regular communication among home care providers, and (6) provides formal mechanisms for communication and collaboration between home care and primary healthcare providers and referral to specialized mental health services [
48‐
50]. Moreover, our findings demonstrate the role and value of PSWs working in collaboration with an RN and other health professionals in enhancing client outcomes. Other studies have also shown the effectiveness of using trained PSWs to improve the health outcomes of older adults in community-based [
7,
106] and institutional settings [
87,
96].
Seventh, the intervention was effective in improving clinical practice behaviours of home care providers. This finding is noteworthy, because most interventions that have successfully improved depression detection have not led to better clinical practice [
85]. The intervention had a significant effect on increasing antidepressant use among older home care participants at six months. Previous studies have also reported improvements in antidepressant use with a collaborative IP DCM approach [
129,
130]. It is possible that participants underreported use of mental health services, which was not increased during the study. Raina [
131] found that older adults tend to over-report contact with primary care practitioners and under-report contact with other medical specialists. A future qualitative study is warranted to learn more about the barriers and facilitators to accessing specialized mental health services for older home care recipients.
Positive changes in many other clinical practice behaviours of the study RNs and PSWs occurred over the intervention period. There was also an increase in the study RNs’ self-reported knowledge and confidence in caring for community-living older adults with depressive symptoms. These findings suggest that the study home care providers were successfully trained to provide and deliver this evidence-based intervention for depression to older adults as part of their caseload. The importance of non-mental health nurses conducting this intervention is considerable, given that previous effectiveness studies of collaborative IP DCM approaches used trained mental health nurses or advanced practice nurses. The nurse-led mental health promotion intervention in this study was provided by existing home care staff.
Overall, our findings suggest that the IP nurse-led mental health promotion intervention is a promising model that has the potential for moving the field toward greater dissemination of evidence-based depression care into real-world practice settings. The academic and community agency partnerships worked together successfully on this research. The participating organizations demonstrated shared commitment for planning, implementation, and evaluation; shared vision and objectives; infrastructure support; stakeholder engagement and buy-in; and strong leadership support in the development of this nurse-led model. These are all essential factors that contributed to the development of a practical, transferable, and sustainable practice model in this population. The research built capacity in depression care and fostered collaborative partnerships across the geriatric mental healthcare delivery system that further enhanced the sustainability of the intervention.