Rural-Urban Disparities in Diagnosis of Early-Onset Dementia

IMPORTANCE Limited access to appropriate specialists and testing may be associated with delayed diagnosis and symptom management for patients with early-onset Alzheimer disease and related dementias (ADRDs). OBJECTIVES To examine rural vs urban differences in diagnostic and symptom management service use among patients with early-onset ADRDs. DESIGN, SETTING, AND PARTICIPANTS This cross-sectional study was conducted using commercial claims from 2012 to 2017. Included patients were those with early-onset ADRDs aged 40 to 64 years, including new patients, defined as those with no claims of ADRDs for 36 months before the first ADRD diagnosis. The likelihood of receiving diagnostic and symptom management services was estimated, with adjustment for individual-level variables associated with health care use. Data were analyzed from February 2021 to March 2022. EXPOSURES Rural residence. MAIN OUTCOMES AND MEASURES Among patients with new, early-onset ADRDs, use of psychological assessment and neuropsychological testing performed at the initial diagnosis (index date) or 90 days or less after the index date and use of brain imaging during the 180 days before the index date were collected. Access to different clinicians on the index date or 90 days or less after the index date was also collected, including visits to primary care physicians and nurse practitioners (PCPs) and specialty visits to psychologists, neurologists, and psychiatrists.


Introduction
Alzheimer disease and related dementias (ADRDs) are debilitating conditions that cause loss of memory, cognition, and independent functioning. While most patients with ADRDs are older than age 65 years, 1 approximately 6% have early-onset ADRDs between ages 30 and 65 years. 2 Earlyonset ADRDs are associated with a more aggressive clinical course, with accelerated cognitive and functional declines compared with later-onset ADRDs. [3][4][5][6] As a result, patients with early-onset ADRDs may experience rapid cognitive decline and become highly dependent on unprepared family caregivers. Additionally, because of the younger age at inception with early-onset ADRDs, most patients are still in the workforce and continue to have notable family and social responsibilities.
Thus, timely diagnosis and management of early-onset ADRDs may help patients preserve daily functioning longer. 1 Because of atypical symptoms and the young age at beginning stages of disease, diagnosis of early-onset ADRDs can often be challenging. For instance, patients with early-onset ADRDs often present with multidomain cognitive impairments with relative sparing of memory function. 2,[7][8][9][10] This can be associated with unfortunate delays in diagnosis. For example, a study 6 reported a mean delay of 1.6 years in the diagnosis of early-onset ADRDs compared with late-onset disease. Because of those challenges, early-onset ADRDs may require expertise from relevant specialists, such as neurologists, psychologists, and psychiatrists, for timely diagnosis. These specialists may perform neuropsychological and behavioral assessments that demand special training for administration and interpretation. While primary care physicians and nurse practitioners (PCPs) may play crucial roles in the diagnosis and care of these patients, they often need consultative guidance from specialists regarding diagnostic and symptom management approaches. 11 Patients with cognitive impairment who are evaluated by specialists also have shorter times to diagnosis and incur lower costs after diagnosis. 12 Diagnostic assessments can also include advanced imaging studies to rule out medically reversible causes of dementia. 13 Accurate and early diagnosis of ADRDs can facilitate the development of symptom management plans, allowing caregivers and families to arrange appropriate accommodations for care. Appropriate and timely care plans may help mitigate some of the onerous disease burden from patients and families. 2,6 If access to specialists with expertise in ADRDs and diagnostic testing is limited, early-onset ADRDs may be misdiagnosed or there may be diagnostic delays. 14,15 Access to specialty clinicians with requisite expertise in the evaluation of cognitive impairment may be especially challenging in rural areas, where clinician shortages have been well described. [16][17][18] However, there is a dearth of studies of the use of specialists or the deployment of appropriate tests for early-onset ADRDs in rural areas compared with urban areas. Thus, we used a nationwide claims database from a privately insured population to examine rural-urban differences in use of imaging studies and neuropsychologic assessments at the time of diagnosis of early-onset ADRDs. Our study may provide evidence on rural-urban differences in patterns of care and access to specialist clinicians for the diagnosis and management of early-onset ADRDs. Our findings may be especially relevant for exploring policies to enhance the availability of a specialist workforce and allocate necessary resources for timely diagnosis and symptom management among patients in rural areas with earlyonset ADRDs.

Methods
The Ohio State University Institutional Review Board exempted this cross-sectional study from review and waived informed consent because it used deidentified data. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Study Design
We conducted a retrospective cross-sectional study using the 2012 to 2018 IBM Watson MarketScan Commercial Claims and Encounters Database. 19 This nationwide claims database includes information regarding diagnoses, health service procedures, clinician types, and care settings for commercially insured enrollees. The database also contains monthly insurance enrollment information, insurance plan type, age, sex, and rural residency data.

Identifying Early-Onset ADRD Population
We identified adults aged 40 to 64 years with any ADRD diagnosis in outpatient or inpatient claims in any year during the observation period. To identify ADRD diagnoses, we followed the algorithm that the Centers for Medicare & Medicaid Services (CMS) uses to identify ADRDs in the Medicare population using International Classification of Diseases, Ninth Revision (ICD-9) and International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes in claims data. 20 We required individuals to be continuously enrolled for the full calendar year with medical and prescription drug coverage each year.
We identified new patients as those having no claims of ADRDs for 36 months before the first ADRD diagnosis in claims (as the principal diagnosis). CMS uses the 36-month period as the reference window to identify ADRDs in Medicare. To apply this approach, we required new patients to be continuously enrolled during the 3-year look-back period, identifying new patients between 2015 and 2018. We defined the index date of a new patient as the date of the first observed ADRD diagnosis after 3 consecutive years without an ADRD claim. Using data from patients with newly diagnosed early-onset ADRDs, we conducted retrospective cohort analyses of health care use.

Measures
We measured the prevalence of early-onset ADRDs by determining the number of patients with We used another binary measure to examine diagnostic service use of brain imaging during 180 days prior to the index date, including computed tomography scans, magnetic resonance imaging, magnetic resonance angiogram scans, or positron emission tomography scans (CPT codes 70496, 70450, 70460, 70470, 0042T, 70544, 70545, 70546, 70551, 70552, 70553, 70554, 70555, 78608, and 78609). As part of sensitivity analyses, we alternated the observation period of imaging services to 12 months prior to the index date to investigate whether imaging studies were performed within a year prior to diagnosis.
To address access to specialty care, we captured whether a patient with a newly diagnosed early-onset ADRD visited a specialist with expertise in dementia on the index date or 90 days or less after the index date. We created a binary indicator separately for visits to psychologists, neurologists, or psychiatrists. Additionally, we investigated whether a patient with a new early-onset ADRD visited a PCP and none of 3 types of specialists on the index date or 90 days or less after the index date. We used this outcome to evaluate access to specialists and substitution of primary care for specialty care in the diagnosis and management of early-onset ADRDs.

Statistical Analysis
We summarized characteristics of patients with early-onset ADRDs using descriptive statistics. Then, we modeled the likelihood of receiving diagnostic and treatment services, adjusting for individuallevel variables associated with health care use. We estimated a logistic regression model separately for each dependent variable. All P values were 2-sided, and results were deemed statistically significant at P < .05 at 95% CI. Data were analyzed from February 2021 through March 2022. Data were analyzed using SAS statistical software version 9.4 (SAS Institute).
The key explanatory variable of the analysis was rurality, defined as residing in a nonmetropolitan statistical area (MSA). Thus, rural areas in our study included those outside of an urbanized area with a population of at least 50 000 individuals and its surrounding communities.      Index date indicates date of the first-observed Alzheimer disease and related dementia (ADRD) diagnosis after 3 consecutive years of no ADRD claim. All differences were statistically significant at 95% CI for unadjusted comparison between new patients in rural vs urban areas. vs urban areas in receiving psychological assessments or imaging studies on the index date or 90 days or less afterward. However, patients with early-onset ADRDs in rural areas were significantly less likely to receive neuropsychological testing compared with patients in urban areas (odds ratio [OR], 0.83; 95% CI, 0.70-0.98). Table 3 contains results of logistic regressions examining access to selected specialist clinicians.
We found that patients with newly diagnosed early-onset ADRDs who lived in rural areas were significantly less likely to have visited a psychologist compared with patients who lived in urban areas (OR, 0.72; 95% CI, 0.60-0.85). However, patients in rural areas with new early-onset ADRDs were  Index date indicates the date of first-observed Alzheimer disease and related dementia (ADRD) diagnosis after 3 consecutive years of no ADRD claim. All differences were statistically significant at 95% CI for unadjusted comparison between new patients in rural vs urban areas.

Discussion
Timely diagnosis and management of early-onset ADRDs can be difficult because patients often present with atypical symptoms and onset at a younger age compared with dementing illnesses in older cohorts. Diagnosis and symptom management among patients with early-onset ADRDs in rural settings can present an even greater challenge. First, in many rural areas, there is a dearth of specialists who have requisite skills to conduct and interpret special diagnostic assessments and supportive treatment for dementing illnesses. 11,15,17,22,23 Moreover, long travel distances for patients may create access barriers to subspecialty services in rural areas for accurate and timely diagnosis of early-onset ADRDs. 12 Findings from this cross-sectional study suggest that a more primary caredirected pattern of care occurred among patients in rural areas with newly diagnosed early-onset ADRDs compared with those in urban settings. We found that patients in rural areas were less likely to obtain neuropsychological tests compared with patients in urban areas. We also found that patients in rural areas were less likely to receive care from psychologists. Instead, patients with earlyonset ADRDs in rural areas relied more on PCPs for diagnosis and management of ADRDs.
Our study finding that rural residency was associated with lower likelihood of visiting psychologists at 90 days or less after diagnosis warrants special mention. This finding is consistent with the finding that patients in rural areas received fewer neuropsychological tests. These tests are designed to identify possible causes of apparent changes in cognitive function (eg, ADRDs or brain injury). Thus, the tests play a major role in the diagnosis, confirmation, and management of dementia. Neuropsychological tests are usually provided by neuropsychologists with specialized training, which includes preparation to assess levels of cognition, sensory perception, language abilities, abstract reasoning, and other aspects of learning and understanding. Most PCPs are not trained to administer neuropsychological evaluations to characterize behavioral and cognitive functions of patients with early-onset ADRDs. However, a rural practice often lacks access to a clinician with the training of a neuropsychologist. Nearly half of nonmetropolitan areas in the US have no psychologists. 23 This will likely exacerbate diagnostic delays for early-onset ADRDs in rural areas.
In addition, rural residents with early-onset ADRDs may face challenges in having symptoms appropriately managed by clinicians with expertise and experience in cognitive impairment.
While we found that there was limited access to neuropsychological tests among patients with early-onset ADRDs in rural areas, there were no significant differences in use of psychological testing.
These tests have notable differences. Neuropsychological tests are more specialized and are used to evaluate different cognitive abilities, including executive function and problem solving. Psychological tests are less specialized and detailed. Our study found fewer visits to psychologists among rural residents but statistically identical levels of psychological testing, suggesting the interesting possibility that other health professionals may have delivered these assessments. While it is highly unlikely that PCPs performed the tests, some rural practices may have clinically licensed social workers who could have been trained in the administration of psychological testing.
Although our study found a reliance on primary care among patients with early-onset ADRDs in rural areas, we did not find significant differences in access to neurologists between rural and urban residents. This was surprising given the more than 4-fold difference between urban and rural areas in the supply of neurologists 24 There are several potential reasons for this finding. First, our findings suggest that PCPs were referring patients in rural areas with early-onset ADRDs to neurologists or that patients were seeking specialty care on their own. Second, health plans may have maintained more adequate networks of neurologists than psychologists because reimbursements to psychologists have conventionally been low. Moreover, most patients in rural areas may have established the pattern of seeing specialist physicians, such as neurologists, in a medical center even when it required farther travel distances.
For unknown reasons, the prevalence of early-onset ADRDs has increased in recent years. 13 Given severe clinician shortages in most rural areas, this suggests that community health care leaders and policy makers should explore innovative solutions to deliver needed specialty care to patients with early-onset ADRDs. Increasing diagnosis rates may be associated with increasing costs of care and additional hurdles in the health care system. With a low number of available specialists to treat patients with early-onset ADRDs in rural areas, the most beneficial roles of specialists and PCPs may merit further research. While PCPs provide valuable care for patients in rural areas, these clinicians may need guidance from specialists regarding diagnostic approaches to cognitive impairment. 11 The recent substantial growth in telehealth during the COVID-19 pandemic represents a prospect to augment the ability to expedite access to clinical psychologists and neuropsychologists in rural areas.
Future improvement in broadband internet services may alleviate infrastructure problems, with associated improvements in opportunities to access specialists for residents in rural neighborhoods.
Telehealth access to behavioral clinicians may be especially important for communities with strong stigmas against mental health care. Furthermore, building on telehealth care, strategies such as Project Extension for Community Healthcare Outcomes (ECHO) may extend opportunities for PCPs to receive specialist training and consult with a team of specialists from various remote sites via regular teleconferencing. 25 Integrating behavioral consultation with primary care in rural areas may be associated with decreased disparities in diagnostic care access among patients with early-onset ADRDs in rural areas.

Limitations
There are several limitations worth noting in this study. First, the CMS algorithm used to identify patients with early-onset ADRDs was developed primarily for the Medicare population. This algorithm includes early-onset and non-early-onset diagnosis codes. We applied all diagnosis codes to our study population because clinicians may not always use early-onset-specific ICD-9 and ICD-10 codes. Second, definitions of patients with early-onset or new ADRDs were based on diagnoses in insurance claims data. Our study would miss patients with ADRDs who received medical care paid entirely out of pocket. Similarly, we could not observe income or education levels, factors that may be associated with health care use. Third, we could not identify neuropsychologists, a highly specialized group of psychologists, within all psychologists in our data. Fourth, we could not use a nuanced definition for rurality because the data did not contain geographic information more specific than the level of metropolitan statistical areas.

Conclusions
This cross-sectional study found that patients with early-onset ADRDs in rural areas had fewer neuropsychological tests and visits to clinical psychologists and were more likely to be diagnosed and treated by PCPs in the 90 days after diagnosis compared with those in urban areas. Given increasing rates of early-onset ADRDs, our findings suggest a need to improve access to neuroscience specialists in rural areas. Clinician training or consultative guidance to PCPs may be viable options to help overcome neuroscience workforce deficiencies in rural areas.