Community resources training showed the greatest uptake in our fidelity analysis as well as the greatest difference in being addressed in conversation and documentation when comparing trained and control NCMs. It could be that some aspects of the education and coordination components involved in the community resources protocol were similar to tasks that NCMs were already preforming. Therefore, NCMs successfully incorporated the protocol into their daily tasks when they were trained on various resources available for PLWD and caregivers. Utilization of community resources has been shown to increase caregivers’ confidence and support in caring for PLWD .
The behavior management training showed the smallest uptake at 12% of records showing evidence of core protocol components and action plans. However, in comparison to NCMs who had not completed training, trained NCMs showed a significant increase in their identification of behavior issues. Behavioral symptoms in dementia are commonplace, with a recent study finding the prevalence of depression, apathy, and agitation/aggression to each be over 30% . For this reason, the difference in behavior identification between groups is likely not due to a difference in occurrence of behavioral symptoms but to increased attention to the symptoms by the NCM. This improvement not only provides an opportunity for NCMs to better manage these complications with supervision and further training, but also to refer to specialists for the most challenging behaviors. Additionally, misunderstanding the meaning of behaviors has been identified as a key theme contributing to caregivers’ challenges with behavioral symptoms . Naming and understanding the cause of a behavior may aide the caregiver in coping and approaching challenges at home. The difference between groups was not seen when it came to management of these behaviors. The complexity of understanding the causes of behavioral symptoms in dementia and the application of specific pharmacological and non-pharmacological therapies is challenging, time consuming, and new to NCMs. Furthermore, fidelity analysis review only occurred in the first six months after training; discussing and addressing behavioral issues might take more time to navigate. Regardless, behavior management is made possible by behavior identification and therefore we believe this to be a step forward in this aspect of dementia care.
Conversations with caregivers about safety followed a similar story, with significantly more records written by trained NCMs showing evidence of identifying safety risks but not of safety planning. Therefore, the management aspect of both behavior and safety could be an area to expand training or provide more consultation with dementia care specialists. Of note, we also observed a trend of trained NCMs being more likely to identify a back-up caregiver. Alternate caregivers are crucially important for safety, but a suggestion like this could also ease caregiver burden and strengthen caregivers’ perception of support.
The decision making and advance care planning protocol showed an uptake of 31% within trained NCMs. This protocol required understanding of when to initiate the advanced care planning conversation, expertise to identify current plans and potential needs, and knowledge to provide appropriate resources and referrals. The breadth of medical, legal, and financial planning questions allows for comprehensive care planning; however, the detailed nature could also be a barrier to completing the protocol for either the NCM or the caregiver-patient dyad. Advance care planning also has an emotional element that some families or NCMs may not be prepared to address . Because these conversations may take more time and careful effort, it may not be realistic to assess the implementation of this protocol in a six month period.
The caregiver wellbeing training was implemented at a lesser rate of 23% but did show promising indications of recognizing the importance of the care partner in comprehensive dementia care. As mentioned previously, most care management works directly with patients. However, due to changing decision-making capacity and nature of Alzheimer’s disease and related dementias, the caregiver is a critical effector of care coordination in dementia care. Although studies have shown poor outcomes on mental and physical health for caregivers  as well as greater self-efficacy among caregivers who have more support , a successful method to include them in a dementia care model has not been widely implemented. The CRESCENT caregiver wellbeing training focused on screening for caregiver strain, asking empathic questions, engaging in collaborative goal setting, and providing resources. The lower protocol implementation could be related to NCMs prioritizing more medically pressing and time sensitive issues, so something that seems less urgent could be delayed. It is also possible that caregivers were not as interested in engaging in conversations about themselves, and perhaps wanted to focus on issues directly related to the PLWD. Interestingly, trained NCMs were more likely to document when there were multiple caregivers involved. Access to caregiver contact information is often unclear from the medical record or other documentation available. Sharing caregiving responsibility, which is quite common in dementia care, can complicate this further. We also observed a trend in increased discussions regarding caregiver financial wellbeing and caregiver physical health. These are promising steps toward recognizing care partners as important members and receivers of dementia care.
There were no differences between the trained and untrained groups in discussing topics related to medications. Medication reconciliation and management is a core function of many high-risk care management programs . Because of this, we would not necessarily expect to see differences in implementation of the medication management protocol between the trained and control groups. However, tailoring training to cover specific medications and management issues of PLWD, such as attending to medication anticholinergic burden, may enhance care. We also did not observe a difference between groups in conversations and documentation of care coordination. This is also expected as NCMs see care coordination as a core part of their job.
Given the diverse and personalized needs of PLWD and family caregivers, more care may not necessarily mean better care. However, CareEco has shown improved outcomes including increased quality of life for PLWD, better caregiver outcomes, and decreased emergency department visits by incorporating dementia care training protocols [11, 18]. Therefore, we believe that adherence to a great number of training protocols in CRESCENT may yield similar outcomes representing high quality dementia care.
As far as we know, this is the first intervention with a focus on training already existing NCMs in a primary care system to deliver dementia specific care. In a narrative review of dementia care models within primary care, nurses, social workers, or unlicensed dementia care managers were utilized primarily for care coordination and were not provided with training in dementia care . Other studies have explored the benefits of providing dementia care training to personnel such as primary care physicians, inpatient staff, and home care workers [20,21,22].
Furthermore, our study is unique in that we assessed documentation of interactions between NCMs and caregivers to better understand the effect of our training intervention. As our healthcare system relies on these individuals heavily in the context of dementia care, future studies are needed to expand our understanding and optimization of the NCM-family caregiver partnership.
This study was limited in its recruitment and care of a racially and ethnically diverse patient population. Despite choosing primary care practices with greater racial and ethnic diversity prior to randomization, most NCMs and PLWD were still white, non-Hispanic, and English speaking. Notably, 14.8% spoke a language other than English. Further, the study period was limited to six months, and it is possible that greater or different effects would be observed over a longer observation period. Additionally, due to the short time period, we cannot comment on the sustainability of our training intervention. Finally, as we were relying on the documentation of clinical encounters, it is possible that some conversation topics were missed if they were not documented in the medical records.