An innovative health systems approach to support early detection of cognitive impairment in primary care – the brain health navigator

Abstract

Background

Patients and providers experience barriers to early detection of mild cognitive impairment (MCI) and dementia. We developed a new primary care-based role, the Brain Health Navigator (BHN), who is trained to assess patients for cognitive impairment, identify addressable causes, suggest appropriate diagnostic testing, connect patients to resources, and assist patients in accessing disease-modifying treatments and dementia care management. This study describes the BHN role, the feasibility of implementing the BHN in Primary Care (PC) clinics, and the initial patients’ outcomes for those who saw the BHN.

Methods

Patients ≥65 years were screened with a Digital Cognitive Assessment (DCA) in seven PC clinics from June 1, 2022, to May 31, 2023. Patients who scored likely cognitively impaired or borderline for impairment were eligible for referral to the BHN. Clinics and providers could determine if referrals were automatic or on a patient-by-patient basis. The BHN encounter included a comprehensive assessment of standardized tools and suggested laboratory and imaging studies to facilitate the diagnostic process and connect patients to resources, care and treatment, and research opportunities.

Results

466 of 861 patients with likely impaired or borderline impaired DCA results were referred to the BHN.More patients with likely impaired scores (62.7%) were referred to the BHN compared to those with borderline scores (47.6%). Of the 466 referred patients, 28.9% with likely impaired scores and 23.5% with borderline scores completed a BHN visit. Patients who were seen by the BHN had a significantly higher likelihood of receiving a new diagnosis of MCI than patients who did not see the BHN and were more likely to have orders for diagnostic tests, such as vitamin B12 thyroid function and Magnetic Resonance Imaging of the head and neck. Referrals to both neurology and neuropsychology were significantly more common among patients who completed the BHN visit.

Conclusions

A BHN enhances follow-up care and monitoring for patients with abnormal cognitive screening tests. A BHN visit increases the rate of evidence-based diagnostic evaluation for MCI and dementia in PC.

Trial Registration

This study was designated as exempt by the Indiana University Institutional Review Board (#15281).

Background

In the United States, Medicare Annual Wellness Visits (AWVs) are often cited as a mechanism to increase early detection of Alzheimer’s Disease and related dementias in primary care [1]. A key component of AWVs is the detection of cognitive impairment through clinician observation, family-reported observations, or the use of brief cognitive assessment tools [2]. However, the uptake and effectiveness of the AWV has been limited due to underutilization, lack of a recommended cognitive assessment tool, and lack of follow up of abnormal cognitive assessments [3–7].

One of the most cited barriers to early detection of Alzheimer’s Disease and Related Dementias (ADRD) in primary care is the lack of a structured process for diagnostic follow-up when a patient screens as cognitively impaired or presents with subjective cognitive complaints. This is compounded for primary care providers by barriers such as limited time, access to informants who know the patient well, and a lack of comfort or confidence in talking about cognition alone or within the context of other chronic conditions [8, 9]. From the perspective of the patient, barriers to early detection include lack of knowledge about the differences between symptoms of early dementia versus normal aging [6, 7], stigma around dementia and fear of potential loss of autonomy, and uncertainty about navigating the healthcare system following an abnormal cognitive screening. These barriers create gaps for patients that lead to underdiagnosis, delayed diagnoses, reduced access to treatment, decreased involvement in clinical research, worse behavioral and psychological symptoms of dementia, and poor management of comorbidities [10, 11]. As a result, patients with dementia frequently have higher healthcare utilization and worse health outcomes, which may lead to an increased burden and higher stress of family caregivers [12, 13].

In other models for detecting disease, processes for assisting patients to navigate the complex interface between primary care and specialty care have been shown to have positive patient and provider outcomes. For example, in cancer care, nurse navigators have been shown to successfully increase patient rates of adherence to follow-up care after an abnormal screening, increase patient satisfaction, and decrease anxiety [2, 14–17]. Patient navigation in primary care emerged in the 1990 s and has grown as a method to facilitate access to care, bridge gaps in care for complex patients caused by healthcare fragmentation, assist in coordinating care across the healthcare system continuum, and connect patients and their families and caregivers to community resources [18]. Nurse navigator models in primary care have been shown to reduce barriers to specialty care for patients with chronic diseases, have high provider and patient satisfaction [19], and significantly improve adherence to recommended care [20].

The recommended evaluation following a positive screening test for cognitive impairment in primary care [21] has many of the complex elements that are ideal for care navigation. For example, further initial workup may include a detailed review of comorbidities, reconciliation and review of prescription and over-the-counter medications, screening for depression, identification of laboratory abnormalities, an assessment of structural brain changes such as vascular disease, stroke, or atrophy, and patient education. Increasing the reliability and effectiveness of a primary care-based diagnostic evaluation process through the use of care navigation has the potential to improve the diagnostic journey for patients and providers and by facilitating appropriate referrals and lessening primary care provider burden for supporting post-diagnosis care. Additionally, a brain health care navigator in primary care could help to ensure access to dementia treatments and care management for the right patients at the right time.

New disease-modifying treatments for Alzheimer’s Disease (AD), specifically monoclonal antibodies targeting amyloid beta, are intended for individuals in the earliest stages of disease, primarily individuals who complete most activities independently. In the Phase 3 clinical trial for lecanemab, approximately 62% of treated individuals had Mild Cognitive Impairment (MCI) [22]. Therefore, it has been suggested that the ideal candidates for anti-amyloid drugs are individuals who are able to perform activities of daily life without assistance and may appear cognitively normal during casual encounters, including with physicians at routine office visits. Timely diagnosis of and consultation for patients who are optimal for treatment is low [20]. The majority of new AD treatments are intended for early symptomatic and MCI stages, indicating a clear trend in the future direction of AD management toward early detection and intervention. This further highlights the need for improving and enhancing primary care pathways for detection and patient care.

Biomarkers for diagnosing AD are becoming more important in the context of AD care and treatment. Prevention studies are treating individuals who have biomarker evidence of disease even before cognitive changes become apparent which some postulate may be the ideal time to treat AD [23, 24]. Blood-based biomarkers are increasingly available, and future diagnostic pathways will likely be facilitated by primary care, creating the need for primary care-based shared decision-making, patient education, and counseling based on biomarker testing and other results [25].

Adding the volume and complexity of these tasks to primary care practices is challenging. In the U.S. healthcare system, identified gaps exist in primary care provider (PCP) knowledge and comfort with diagnosing cognitive impairment, primary care-based management of dementia and related conditions, and PCPs’ belief in the effectiveness of therapeutic options [26–28] Health system structures including insufficient time, inadequate support resources, and lack of reimbursable billing options are cited as significant additional constraints by PCPs [29]. Collaborative models of dementia care have been explored, developed, and validated over the years to address and overcome barriers, providing system approaches that have improved care where these models are feasible [30–32].

Inadequate representation of people from racial and ethnic minorities in AD clinical trials has been identified as a significant concern among the scientific community. Recruitment for brain health research from primary care practices has been identified as a potential opportunity [33] particularly since older adults seeking primary care tend to be representative of the community [34]. However, lack of connections between primary care practices and AD research opportunities has historically been a barrier to leveraging this potential [35].

Given the numerous challenges facing patients experiencing cognitive decline and PCPs diagnosing dementia, we developed and implemented a patient care nurse (registered nurse) navigator—the Brain Health Navigator (BHN)—as support for the primary care team in the management of care for persons following an abnormal screening test. The role of the BHN was designed to assess cognitive impairment, facilitate the characterization of potential addressable causes, connect patients and their caregivers with community and research resources, and streamline entry to the clinical treatment pathway when needed. The purpose of this study is to demonstrate the feasibility of implementing a primary care-based BHN to improve the reliability and effectiveness of the diagnostic assessment process in primary care of patients who screen borderline for cognitive impairment or likely cognitively impaired.

Methods

The Indiana University (IU) and IU Health flagship site of the Davos Alzheimer’s Collaborative Healthcare System Preparedness Early Detection Program integrated the Linus Health Core Cognitive Evaluation Digital Cognitive Assessments (DCA) into seven IU Health primary care clinics in the greater Indianapolis area to screen patients 65 years and older who arrived for a primary care visit [36–38]. Under a waiver of consent and authorization for this pragmatic project, clinical data were collected for routine primary care for 12 months pre- and 90 days post-index DCA and 90 days post BHN visit. Screening results were presented to the patient’s primary care provider with a report that used a stop-light system: green, yellow, or red indicated unimpaired, borderline for impairment, or likely cognitively impaired respectively [39]. During the one-year project period from June 1, 2022, to May 31, 2023, patients who completed at least one DCA with a result with a borderline or likely impaired result were eligible for referral to the BHN. Referrals to the BHN were at the discretion of the PCP. Providers had the option to refer individual patients at the time of visit (individual-based referral), review a list of patients with borderline or likely impaired results and refer patients from the list, or opt to automatically refer patients with either or both borderline or likely impaired screen (rule-based referrals). The method of provider referral was tracked for each referred patient. Based on implementation of the DCA in the practices, no patients with unimpaired results were referred to the BHN. Providers who chose not to have a DCA result-based rule referring patients to the BHN were able to send patients to the BHN on a patient-by-patient basis.

Due to the pragmatic nature of this study, capacity for contacting and seeing patients referred to the BHN was limited in the early part of the study period as the position was being developed and implemented. To the greatest extent possible, efforts were made to reach out to all previously referred patients throughout the study period while maintaining outreach and capacity to recently referred patients. Referred patients were sent an introductory letter or message in the Electronic Health Record (EHR) patient portal asking them to call and schedule an appointment, followed by three phone call attempts by the BHN. Patients remained under the care of their PCP even if they were referred to or seen by the BHN at any point. BHN visits were conducted in the patient’s primary care clinic or in another participating primary care clinic if the location and schedule was more convenient. At the time of scheduling, patients were encouraged by the scheduler to bring a family member or a care partner to the BHN appointment. The same process was used for patients referred individually by their provider and patients referred based on a provider referral rule.

Prior to the BHN visit, the BHN reviewed the patients’ EHR for all comorbidities, and medications and previous brain related testing, labs, and primary care notes. During the BHN visit, they reviewed the patients’ DCA results with them and answered their questions about the test results. Additional data collection at the visit included a Montreal Cognitive Assessment (MoCA) [40] to further characterize the patient’s cognitive status. The BHN evaluated addressable causes of cognitive impairment using a variety of structured assessment tools. The Patient Health Questionnaire-9 (PHQ-9) [41] and Generalized Anxiety Disorder-7 (GAD-7) [42] were used to assess depression and anxiety. Vascular risk was assessed with the Atherosclerotic Cardiovascular Disease risk score (ASCVD) [43]. After a complete medication reconciliation of prescription and over-the-counter medications, an Anticholinergic Burden Scale (ACBS) [44] was used to calculate the anticholinergic burden. Any medication that contributed to an elevated anticholinergic burden was highlighted in a communication to the PCP. Following the visit, the BHN reviewed the patient’s EHR again and proposed additional laboratory tests or brain imaging to the PCP based on the assessed findings from the evaluation for addressable causes. Order proposals included tests such as vitamin B12 level, thyroid stimulating hormone (TSH), radiologic brain imaging, and, when indicated by an elevated sleep apnea screening score [45], a sleep study.

At the following visit, the BHN provided education on community resources, based on a repository assembled by a social worker and a registered nurse, including local and national resources (e.g., Alzheimer’s Association) and a list of local dementia caregiver support groups. The BHN also initiated and engaged in discussions with each patient and their care partner about establishing power of attorney, elder care law services, and connecting patients with home and community-based services such as area agencies on aging, as appropriate. Both electronic and paper-based resources were provided to patients and care partners during the BHN visit.

Patients referred to the BHN were ultimately triaged into categories based on the results of their workup: those with cognitive symptoms from potentially reversible or intervenable causes, those with complex medical conditions likely to require more comprehensive testing and care, and those with suspected neurodegenerative diseases. When appropriate, the BHN proposed laboratory, imaging, and referral orders for specialists to the PCP. Expedited referral workflows were developed and implemented with neurology, geriatrics, social work, and other specialists. This referral process did not differentiate referral options by level of cognitive impairment.

As part of the BHN protocol, patients were offered to participate in research opportunities through either a handoff to a clinical research coordinator or to a research recruitment registry, depending upon their interest and eligibility. As part of an embedded pilot research study, which was approved by the Institutional Review Board and will be reported separately, the BHN offered patients the opportunity to undergo blood-based biomarker testing to assist in their diagnostic process and have their results disclosed to them. A board-certified behavioral neurologist (first author JRB) provided a one-hour educational video to train PCPs on biomarker disclosure.

Patient demographic data, comorbidities, medications, referrals to specialists, diagnostic testing for cognitive impairment, cognitive assessment scores, new diagnoses of MCI or ADRD, and referrals to dementia-related clinical trials were collected at the time of DCA screen [37], at the BHN visit, and 90 days post the BHN visit from each patient’s electronic health record. Patients who screened borderline or likely impaired were analyzed, respectively, by those who had versus did not have a BHN referral from their PCP and also by those who were referred and had versus did not have a BHN visit. All referrals to the BHN were tracked by source and date. The Area of Deprivation Index (ADI) [46] and Charlson comorbidity index [47] were calculated to further characterize the cohort of patients with abnormal screenings and to determine if disparities existed in referrals or outcomes.

Statistical analysis

Two-sample T-tests or Wilcoxon rank-sum test were performed to compare whether patients differed by referral status on demographic and clinical characteristics. Fisher’s exact tests were used for categorical variables. Similar analyses were conducted for the following comparisons: referred patients seen by the BHN versus referred patients not seen by the BHN; likely impaired patients versus borderline patients; and outcomes by referred patients seen by the BHN versus referred patients not seen by the BHN.

Due to differences in patient panel characteristics at each of the seven clinics, clinic-adjusted p-values were calculated for differences in patient characteristics. Both overall and clinic-adjusted p-values are presented when a material difference between overall and clinic-adjusted values are present. When no material difference is present, clinic-adjusted values are presented.

Because DCA screening rates at each of the seven clinics varied, the relationship between screening rate and referral rate at each clinic was explored. The effect of a scheduling-induced lag time between the DCA index visit and the BHN index visit was explored. The relationship of race and ethnicity and acceptance of research referral was explored.

All analyses were performed using SAS v9.4.

Results

 A total of 861 patients who were screened from June 1, 2022, to May 31, 2023, screened likely impaired or borderline impaired on the DCA. Of those, 236 (27.4%) screened likely impaired and 625 (72.6%) screened borderline impairment. A total of 446 (51.8%) patients (148 likely impaired and 298 borderline) were referred to the BHN. Table 1 presents demographic and clinical characteristics by BHN referral status. Unadjusted results show that referred patients were more like to screen likely impaired (vs. borderline) (p < 0.001), were younger (p = 0.01), had less years of education (p < 0.001), were more likely to be Black (p < 0.001), and had higher ADI scores (p < 0.001). Referred patients had significantly higher rates of hypertension (p = 0.03) and hyperlipidemia (p < 0.001). After adjusting for the clinic that the patient was associated with and was screened in, only result on the DCA, age and hyperlipidemia were significant. The relationship between screening rate and referral was not significant (not shown).

Table 1.

Descriptive variables of patients by brain health navigator referral status

	Patients not referred to the BHN (n = 415)	Patients referred to the BHN (n = 446)	p-value	Clinic adjusted p-value
Digital cognitive assessment result, n (%)			< 0.001	0.038
Likely cognitively impaired	88 (21.2)	148 (33.2)
Borderline for cognitive impairment	327 (78.8)	298 (66.8)
Age, mean (SD)	75.4 (6.3)	74.3 (6.7)	0.014	0.087
Age category, n (%)			0.010	0.039
65–69 years	82 (19.8)	128 (28.7)
70–74 years	131 (31.6)	130 (29.2)
75–79 years	87 (21.0)	95 (21.3)
80–84 years	80 (19.3)	55 (12.3)
85–89 years	26 (6.3)	24 (5.4)
90 + years	9 (2.2)	14 (3.1)
Sex, n (%)			0.130	0.198
Female	223 (53.7)	263 (59.0)
Male	192 (46.3)	183 (41.0)
Race, n (%)			< 0.001	0.833
Asian	10 (2.4)	8 (1.8)
Black or African American	46 (11.1)	183 (41.0)
White	358 (86.3)	252 (56.5)
Other reported	1 (0.2)	3 (0.7)
Ethnicity, n (%)			0.071	0.260
Hispanic	13 (3.2)	26 (5.8)
Non-Hispanic	400 (96.8)	420 (94.2)
Years of education, mean (SD)	14.1 (2.7)	13.4 (3.0)	< 0.001	0.697
Area Deprivation Index, median (25%ile, 75%ile)	49 (28, 71)	71 (43, 91)	< 0.001	0.456
Charlson Comorbidity Index, median (25%ile, 75%ile)	1 (0, 2)	1 (0, 2)	0.100	0.877
Comorbidities, n (%)
Hypertension	332 (80.0)	382 (85.6)	0.030	0.096
Hyperlipidemia	203 (48.9)	297 (62.6)	< 0.001	0.014
Chronic Kidney Disease	83 (20.0)	104 (23.3)	0.248	0.440
Diabetes	142 (34.2)	173 (38.8)	0.179	0.429
Congestive Heart Failure	44 (10.6)	60 (13.4)	0.211	0.301
COPD	74 (17.8)	86 (19.3)	0.560	0.928
Obstructive Sleep Apnea	50 (12.0)	50 (11.2)	0.750	0.164
Obesity	82 (19.8)	84 (18.8)	0.795	0.191
Cancer	44 (10.6)	40 (9.0)	0.424	0.596
Primary Care Clinic			< 0.001
Clinic 1	116 (27.9)	21 (4.7)
Clinic 2	2 (0.5)	13 (2.9)
Clinic 3	160 (38.6)	49 (11.0)
Clinic 4	27 (6.5)	253 (56.7)
Clinic 5	102 (24.6)	83 (18.6)
Clinic 6	7 (1.7)	3 (0.7)
Clinic 7	1 (0.2)	24 (5.7)

BHN  Brain Health Navigator, COPD Chronic Obstructive Pulmonary Disease

One of the seven clinics elected to send all patients at the practice with a borderline or likely impaired DCA screening to the BHN. The six other clinics chose to have individual provider-based decisions about rule-based referrals to the BHN. Fourteen of 42 (33.3%) providers at those clinics chose to refer patients to the BHN with a likely impaired DCA result based on a rule. Eleven of 42 (26.2%) providers at the individual provider decision clinics decided to additionally refer patients with a borderline DCA result to the BHN based on a rule. All remaining providers at the six clinics made referrals without a rule on a patient-by-patient basis. Table 2 presents the breakdown of rule-based versus individual provider-based referrals to the BHN.

Table 2.

Provider brain health navigator referral type

	All Providers		Providers at an Individual Provider Decision Clinic
	Red DCA referral	Yellow DCA referral	Red DCA referral	Yellow DCA referral
BHN Rule Based Referral	79 (73.8%)	76 (71.0%)	14 (33.3%)	11 (26.2%)
Individual Patient BHN Referral	28 (26.2%)	31 (29.0%)	28 (66.7%)	31 (73.4%)
Total Providers	107		42

Of the 446 patients referred to the BHN, a total of 100 (22.4%) had a BHN visit. Reasons for patients not seen by the BHN include patient declined (n = 133, 29.8%), BHN was unable to contact patient (n = 160, 35.9%), and patient was not able to be approached (n = 53, 11.9%). Comparisons of those patients who had a BHN visit versus those who did not are presented in Table 3. In unadjusted analyses, patients seen by the BHN were more likely to have a diagnosis of Chronic Obstructive Pulmonary Disease (COPD) (p < 0.01) and sleep apnea (p < 0.05). Patients seen by the BHN were more likely to be referred by an individual-based referral decision method than a rule-based referral decision method (34.0% vs. 21.4%; p < 0.05). The effect of a scheduling-induced lag time between the DCA index visit and the BHN index visit was not significant. After adjusting for the clinic that the patient was associated with and was screened in, patients having COPD, sleep apnea, and hyperlipidemia were the only significant predictors of having a BHN visit.

Table 3.

Descriptive variables of patients who had an encounter with the brain health navigator

	Not Seen by BHN (n = 346)	Seen by BHN (n = 100)	p-value	Clinic adjusted p-value
Digital cognitive assessment result, n (%)			0.185	0.286
Likely cognitively impaired	237 (68.5)	61 (61.0)
Borderline for cognitive impairment	109 (31.5)	39 (39.0)
Age, mean (SD)	74.4 (6.7)	73.8 (7.1)	0.464	0.307
Age category, n (%)			0.110	0.062
65–69 years	90 (26.0)	38 (38.0)
70–74 years	107 (30.9)	23 (23.0)
75–79 years	79 (22.8)	16 (16.0)
80–84 years	43 (12.4)	12 (12.0)
85–89 years	16 (4.6)	8 (8.0)
90 + years	11 (3.2)	3 (3.0)
Sex, n (%)			1.000	0.918
Female	204 (59.0)	59 (59.0)
Male	142 (41.0)	41 (41.0)
Race, n (%)			0.638	0.702
Asian	5 (1.5)	3 (3.0)
Black or African American	142 (41.0)	41 (41.0)
White	196 (56.6)	56 (56.0)
Other reported	3 (0.9)	0 (0.0)
Ethnicity, n (%)			0.473	0.426
Hispanic	22 (6.4)	4 (4.0)
Non-Hispanic	324 (93.6)	96 (96.0)
Years of education, mean (SD)	13.2 (3.1)	13.9 (2.4)	0.053	0.011
Area Deprivation Index, median (25%, 75%)	71 (43, 89)	73 (49, 93)	0.322	0.417
Charlson Comorbidity Index, median (25%, 75%)	1 (0, 2)	1 (0, 3)	0.178	0.186
Comorbidities, n (%)
Hypertension	294 (85.0)	88 (88.0)	0.519	0.628
Hyperlipidemia	220 (63.6)	59 (59.0)	0.414	0.752
Chronic Kidney Disease	82 (23.7)	22 (22.0)	0.789	0.677
Diabetes	129 (37.3)	44 (44.0)	0.245	0.311
Congestive Heart Failure	42 (12.1)	18 (18.0)	0.136	0.334
Chronic Obstructive Pulmonary Disease	57 (16.5)	29 (29.0)	0.009	0.007
Obstructive Sleep Apnea	32 (9.2)	18 (18.0)	0.019	0.033
Obesity	61 (17.6)	23 (23.0)	0.246	0.253
Cancer	27 (7.8)	13 (13.0)	0.115	0.057
Provider authorization method for BHN referral, n (%)			0.012	0.073
Individual based referral decision method	73 (21.4)	34 (34.0)
Rule based referral decision method	269 (78.6)	66 (66.0)
Days between DCA and attempted contact by BHN, n (%)			0.409	0.177
0–28 Days	174 (58.8)	64 (64.0)
≥29 Days	122 (41.2)	36 (36.0)
Primary Care Clinic			0.011
Clinic 1	11 (3.2)	10 (10.0)
Clinic 2	8 (2.3)	5 (5.0)
Clinic 3	34 (9.8)	15 (15.0)
Clinic 4	199 (57.5)	54 (54.0)
Clinic 5	71 (20.5)	12 (12.0)
Clinic 6	2 (0.6)	1 (1.0)
Clinic 7	21 (6.1)	3 (3.0)

BHN Brain Health Navigator, DCA Digital Cognitive Assessment

Among the patients who screened likely impaired and borderline on the DCA, Table 4 shows clinic adjusted differences in outcomes for the patients who had a BHN visit versus those who did not (both not referred and referred but not seen). The BHN intervention led to several statistically significant outcomes. Patients who were seen by the BHN had a significantly higher likelihood of receiving a new diagnosis of MCI than patients who did not see the BHN (5.0% vs. 1.3%, p < 0.05). Patients who saw the BHN were also more likely to have new order for medication for dementia (3.0% vs. 0.8%, p < 0.05), more likely to have orders for diagnostic tests, such as vitamin B12 (26.0% vs. 7.4%, p < 0.001) and TSH (24.0% vs. 9.1%, p = 0.001). Magnetic Resonance Imaging (MRI) orders were higher among those seen by the BHN (25.0% vs. 3.4%, p < 0.001), and referrals to both neurology and neuropsychology were significantly more common. Notably, 43% of all patients referred to the BHN expressed interest in research participation, and when determined to be eligible, these individuals were connected to a research coordinator for recruitment screening visits.

Table 4.

Outcomes for patients by brain health navigator (BHN) visit status at 90 days

	Not Seen by BHN n = 761	Seen by BHN n = 100	Clinic adjusted p-value
New Dementia Related Diagnosis, n (%)
Alzheimer’s Disease or Related Dementia (ADRD)	8 (1.0)	0 (0.0)	0.455
Mild Cognitive Impairment (MCI)	10 (1.3)	5 (5.0)	0.017
New antidementia medication ordered, n (%)	6 (0.8)	3 (3.0)	0.015
Order for labs, n (%)
Vitamin B12	56 (7.4)	26 (26.0)	< 0.001
Thyroid Stimulating Hormone (TSH)	69 (9.1)	24 (24.0)	0.001
Order for imaging of head, neck, n (%)
Computed Tomography (CT)	40 (5.3)	5 (5.0)	0.578
Magnetic Resonance Angiography (MRA)	2 (0.3)	0 (0.0)	0.753
Magnetic Resonance Imaging (MRI)	26 (3.4)	25 (25.0)	< 0.001
Order for Referral to Specialist, n (%)
Neurology	29 (3.8)	16 (16.0)	< 0.001
Geriatrics	4 (0.5)	0 (0.0)	0.544
Neuropsychology	3 (0.4)	3 (3.0)	0.016
Psychiatry	4 (0.5)	0 (0.0)	0.544
Deceased, n (%)	6 (0.8)	0 (0.0)	0.753

Regarding the BHN impact on research referrals, there was no difference in acceptance of a referral to research based on race (Asian 66.7% n = 2 vs. Black of African American 46.3% n = 19 vs. White 37.5% n = 21, p = 0.48) or ethnicity (Hispanic or Latino 75.0% n = 3 vs. Not Hispanic or Latino 40.6% n = 39, p = 0.31). Results not shown. Patients who were seen by the BHN were significantly more likely to consent to AD blood-based biomarker testing and results disclosure compared to those patients who were not referred and to those who were referred but did not have a BHN visit (1.2% vs. 2.9% vs. 11.0% p < 0.001). Table 5.

Table 5.

Enrollment in AD blood-based biomarker research study

	Not Referred n = 415	Referred, not seen by BHN n = 346	Referred, Seen by BHN N = 100	P-Value
Patient consented for AD blood-based biomarker test and results disclosure, n (%)	5 (1.2)	10 (2.9)	11 (11.0)	< 0.001

DISCUSSION

Enhanced primary care evaluation

To our knowledge, this is the first use of a nurse navigator in primary care to support the care of patients with identified cognitive impairment. We believe this model could have an important impact on early identification of patients with cognitive impairment and streamline referral and treatment pathways, which will contribute to better health outcomes, improve healthcare utilization, and reduce burden on PCPs.

As a member of the primary care team, the BHN provides more comprehensive assessments, connection to resources, and care navigation for patients with cognitive concerns and borderline or likely impaired cognitive screening results. By executing more comprehensive initial cognitive assessments, by reviewing all prescription and over-the-counter medications/supplements, and evaluating for addressable co-morbid medical and mental health conditions, the BHN helps identify contributing or alternate causes for cognitive impairment. For appropriate patients, the BHN proposes additional workup to the primary care providers, such as brain imaging to assess for vascular burden and atrophy patterns, laboratory workup, and in appropriate cases, biomarker assessments using blood-based biomarkers, allowing the PCP to have oversight of the care process, but reducing the direct burden of that process on the PCP. Referrals to both neurology and neuropsychology being significantly more common in patients who saw the BHN suggests enhanced follow-up care and monitoring in the BHN group. In addition, the BHN can focus on the detailed assessment of cognitive-related conditions and testing, allowing the PCP to continue to address and manage other concurrent conditions.

Thus, in this analysis, engaging with the BHN led to statistically significant improvements in care for patients. These findings are important because they address important gaps in the current rates of diagnosis and follow-up care of patients with MCI and dementia [48]. The BHN role parallels other nurse navigators in primary care, particularly those helping to manage chronic medical conditions, where the appropriate information and support significantly impact patient care [49].

Enhanced specialty care connection

In the initial development and implementation of the BHN role, the BHN workflow did not differentiate follow-up or referral testing options for patients with possible neurodegenerative disorders based on their level of cognitive impairment, which is consistent with our finding that the rate of referrals for patients with borderline impairment and likely cognitive impairment were not statistically different from one another. Since the end of the study period, we have evolved the BHN workflow with a goal to improve patient navigation so that the most appropriate patients are referred to the most appropriate specialists or enhanced care pathway for their level of impairment and/or care needs. Thus, we believe the BHN role continues to hold promise for more appropriate referrals and partially mitigate exacerbation of the bottlenecks in specialty referrals.

At the time of the initiation of this study, no Federal Drug Administration (FDA)-approved disease-modifying treatments for AD was widely available. Therefore, when designing the BHN role and their triage function, patients suspected of having AD were referred to neurology, while those with significant vascular risk and comorbidities were referred to geriatrics or, if the patient preferred, back to primary care. Our updated practice allows the BHN to determine if a patient may be eligible for disease-modifying treatments (based on an agreed-upon checklist) and to refer them directly to a neurology treatment nurse coordinator instead of to the neurology referral pool, which further expedites patients’ access to disease-modifying medications. The BHN’s close collaboration with the neurology treatment team enhances the pathway process, as the BHN is familiar with treatment requirements, including contraindicated medications and relevant medical conditions.

While we were unable to measure the appointment lag of referrals in this study due to the limited follow up period, we believe that the BHN’s involvement can accelerate the diagnostic process, as BHN visits are available much sooner than traditional PCP referrals to neurology and geriatrics in our system. Subsequently, the warm hand off to the neurology treatment team for patients who had been thoroughly assessed by the BHN, are interested in treatment evaluation, and meet key treatment inclusion and exclusion criteria, allow patients to have faster evaluation of their appropriateness for disease modifying therapies.

In published literature reporting additional analysis of the disease-modifying treatment donanemab, treatment was found to be more effective in earlier disease-stage individuals, making early identification and early referral a key component of treatment success [50]. In appropriate patients, the BHN also follows up on imaging results and may potentially suggest screening or definitive biomarker assessments. The embedded blood-based biomarker pilot research study holds promise for the feasibility of the BHN offering blood-based biomarker tests and interpretation as part of the cognitive impairment assessment. In our model, the BHN disclosed biomarker results to patients and families with careful consideration, following extensive training on the nuances and the potential psychological impact of these discussions. Our finding that patients who saw the BHN were more likely to have a blood-based biomarker performed indicates that the BHN may play an important role in facilitating the use of blood-based biomarkers in primary care. More research is needed in this area.

We found that primary care patients at an older age were more likely statistically to have likely impaired DCA screening results, suggesting that routine longitudinal primary care-based screening may be helpful in identifying MCI earlier than in current healthcare systems models. Patients referred to the BHN were younger, further suggesting that routine screening in primary care may be successful at improving rates of early detection and follow-up when combined with a primary care-based BHN visit for navigating next steps. Further research is needed in this area.

Connection to community resources and research

In addition to the triage and referral processes, the BHN often served as a patient’s initial point of contact for connection to community organizations and social work services, which hold potential to have an impact on reduction of caregiver stress and burnout. Advance care planning discussions like the ones initiated by the BHN at an early stage have been shown to improve outcomes in patients and families living with dementia, ultimately leading to improved patient autonomy in decision-making and reductions in healthcare utilization [51].

In our study, the large clinic with standardized screening practices and rule-based referrals had the greatest number of patients screened and referred to the BHN. This clinic had the highest rates of barriers to social determinants of health, suggesting that clinic factors could have contributed to the trend toward significance of ADI for those screened as likely cognitively impaired. However, previous literature has demonstrated higher rates of abnormal cognitive assessments in patients with low socioeconomic status [52]. Furthermore, in our study, the trend toward significance of the ADI with those who screened likely cognitively impaired demonstrates the importance of community resource referral as a key function for the BHN. These factors highlight the need for further research on the role of primary care in addressing health disparities of early detection and treatment.

We demonstrated that over 40% of patients with borderline impairment or likely cognitive impairment were willing to accept a referral for research, including 46% of Black or African American patients who saw the BHN. This finding supports previous literature suggesting that primary care-based recruitment for AD clinical trials has potential to enhance clinical trial diversity and that when given the opportunity to participate equitably, disparities in AD research participation may be significantly reduced.

Although a relatively small number of people participated in a research sub-study to collect blood-based biomarkers, a significantly higher percentage of patients volunteered to consent when they interacted with the BHN. We believe that this data demonstrates the BHN playing a crucial role in engaging patients in a primary care setting in neurologic research, which may be beneficial in improving the representation of people from racial and ethnic minorities in AD clinical trials.

Our finding that 43% of real-world primary care patients with borderline and likely cognitively impaired screenings who saw the BHN were interested in research referrals offers potential insight into the promise of pathways from primary care early detection into clinical trials as a mechanism to enhance diversity of dementia-related clinical trials. Additionally, it demonstrates the BHN role as successful in connecting patients with opportunities that are appropriate to their individual health and goals, as well as connecting historically fragmented pieces of the AD research community. Primary care’s contributions to AD team science hold the promise of yielding greater societal benefits.

Billing and revenue

Although the BHN model confers a cost to the healthcare system, its potential benefits in reducing the burden of other healthcare providers and accelerating patient triage and workup are critical in managing patients who may be candidates for disease-modifying therapies. While financial data on the cost of the BHN is not clearly understood, BHN billing may be possible using appropriate documentation. The Current Procedural Terminology (CPT) Evaluation and Management billing code 99211 can be used for billing outpatient nurse encounters, but the services provided by the BHN far exceed the typical services provided under this billing code. However, in our clinic with the current infrastructure, the services provided by the BHN meet the Center for Medicare and Medicaid Services Chronic Care Management service requirements [53]. These more comprehensive codes better account for the services provided by the BHN particularly in patients who require multiple encounters for diagnostics, connection with community-based services, and workup for disease modifying treatments. Although these codes are unlikely able to fully fund the service, they can offset cost substantially depending on patient volumes. Further policy-based recognition of this role and additional funding mechanisms could support broader implementation. Anecdotal feedback from BHN nurses within our healthcare system has been overwhelmingly positive, with high job-satisfaction, which is consistent with other primary care-based nurse navigation models [19].

Limitations

Several limitations to this study should be noted, including the lack of financial data and survey results related to biomarkers and patient satisfaction, which may impact the generalizability and conclusions of our findings.

Limited rates of patient completion of BHN appointments

Another limitation to our study is that the number of patients who ultimately saw the BHN during the study period was relatively low. In this study, the BHN was only able to conduct a visit with 22.8% of referred patients. This was partially due to the delays in establishing the role and development of the approach. The majority of patients did not respond to the BHN contacts, with a maximum of three attempts made per subject. Approximately one-third of patients declined the BHN visit. Patient refusal to see the BHN may result from stigma related to concerns over brain health or denial of symptoms, and discerning these two is an aim for future research. It is worth noting that the BHN role was not established and implemented at the same time as the DCA screening in primary care, leading to substantial gaps in the lag time from DCA screening to BHN outreach and appointments for some patients.

We evaluated two factors retrospectively to better understand the low BHN completion rate: the type of referral and timing to first contact. We found that patients who were individually referred by their PCP were more likely to complete a BHN appointment. This suggests that providers likely conveyed their perceived value of the BHN referral during their encounter with the patient. Our opinion is that more immediate emphasis on BHN follow-up by the provider and timeliness of the referral to the BHN after DCA screening may play a role in the rate at which patients see the BHN. We hypothesize that more rapid contact by the BHN would lower rates of patients declining to see the BHN [54]. While our data analysis did not show significance in a retrospective analysis, further research on this factor is warranted, as this type of practice leads to better visit adherence in the experience of the authors and in previously published work [55]. With that in mind, our updated workflows are designed with a goal of scheduling a BHN appointment on the same day the DCA is completed.

Conclusions

As a member of the primary care team, the BHN can reduce burden from the PCP by providing more comprehensive assessments, patient education, and care navigation for patients with cognitive concerns and borderline or abnormal cognitive screening results. The BHN role leads to higher completion of appropriate medical evaluation of addressable causes of cognitive impairment, which can include blood biomarkers, and more frequent assignment of a diagnosis, enhancing the success of early diagnosis to treatment clinical pathways. The BHN may also play an important role in connection with clinical research. Future research is needed to understand how the BHN role can be integrated and sustained in different systems.

Abbreviations

ACBS

Anticholinergic Burden Scale

AD

Alzheimer’s Disease 

ADI

Area Deprivation Index 

ASCVD

Atherosclerotic Cardiovascular Disease 

AWV

Annual Wellness Visit 

BHN

Brain Health Navigator 

COPD

Chronic Obstructive Pulmonary Disease 

CPT

Current Procedural Terminology 

DCA

Digital Cognitive Assessment 

EHR

Electronic Health Record

FDA

United States Federal Drug Administration 

GAD-7

Generalized Anxiety Disorder-7 

MCI

Mild Cognitive Impairment 

MoCA

Montreal Cognitive Assessment 

MRI

Magnetic Resonance Imaging 

n

Number 

PCP

Primary Care Provider 

PHQ-9

Patient Health Questionnaire-9 

SAS

previously Statistical Analysis System, no longer considered an abbreviation 

SD

Standard Deviation 

TSH

Thyroid Stimulating Hormone

%

Percent 

Authors’ contributions

JB: development, implementation, manuscript preparation, writing, and editing. NRF and DRW: development, implementation, data management, manuscript writing and editing. DBH, DS, KS: development, implementation, manuscript writing, and editing. AJP: data management, analysis, manuscript writing, and editing. APH: data management, manuscript writing, and editing.

Funding

This research was supported by the Davos Alzheimer’s Collaborative Healthcare System Preparedness Program, Linus Health, and C2N Diagnostics.

Data availability

The data for this article are not a publicly available dataset. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Exemption from ethics approval and waiver of informed consent to participate granted: The ethics committee of Indiana University Institutional Review Human Subjects Review Office waived the need for ethics approval and the need to obtain consent for the collection, evaluation, analysis and publication of the anonymized data for this non-interventional, quality improvement study. (Application 15281). This study has been performed in accordance with the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.