Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jul 1.
Published in final edited form as: J Appl Gerontol. 2023 Feb 4;42(7):1414–1423. doi: 10.1177/07334648231155444

The Association between Primary Care Practices’ Structural Capabilities and Hospitalizations among Persons Living with Dementia

Vaneh E Hovsepian 1, Tatiana Sadak 2, Amelia E Schlak 3, Jianfang Liu 4, Lusine Poghosyan 4
PMCID: PMC10257735  NIHMSID: NIHMS1867460  PMID: 36738162

Abstract

Background:

Persons living with dementia (PLWD) are more likely to be hospitalized than individuals without dementia. Little is known about key features (i.e., structural capabilities) in primary care practices where PLWD receive care. This study assessed the relationship between structural capabilities (i.e., care coordination, community integration, and reminder systems) and hospitalizations among PLWD.

Methods:

We conducted a secondary analysis of cross-sectional data from 5,001 PLWD in 192 practices and used three datasets: nurse practitioner surveys, Medicare claims, and Minimum Data Set. Using generalized estimating equations, we evaluated the association between structural capabilities and hospitalizations.

Results:

PLWD who received care from practices with care coordination were less likely to have hospitalizations (OR = 0.62, p < .05). No statistically significant associations were observed between community integration and reminder systems and hospitalizations.

Conclusion:

Primary care practices need to tailor structural capabilities to address the needs of PLWD to reduce hospitalizations.

What this paper adds:

The knowledge gained through this study can contribute to improving the quality of care for PLWD in primary care settings. More specifically, for the first time this study assesses the relationship between structural capabilities and hospitalization among PLWD who receive care from primary care practices employing nurse practitioners.

Applications of study findings:

The results of this study can inform clinicians, administrators, researchers, and policy makers to guide their efforts to enhance structural capabilities in primary care practices to potentially reduce hospitalizations.

In the United States (U.S.), persons living with dementia (PLWD) have almost twice the likelihood of being hospitalized and experience more extended hospital stays than older adults without dementia (Alzheimer’s Association, 2022; Lin et al., 2013). Some hospitalizations, called ambulatory care sensitive conditions (ACSC) hospitalizations, can be prevented if patients receive high-quality primary care (Billings et al., 1993; Gao et al., 2014). According to the Agency for Healthcare Research and Quality, ACSC hospitalizations are related to conditions such as pneumonia, diabetes, hypertension, asthma, congestive heart failure, and chronic obstructive pulmonary disease (Fingar et al., 2015; Lin et al., 2017; “Prevention Quality Indicators Technical Specifications,” 2020). Managing chronic conditions in PLWD remains a challenge, as these patients often face difficulties reporting symptoms, adhering to medication schedules, and complying with the treatment (Gual et al., 2018; Nelis et al., 2019).

Studies have shown that hospitalizations have adverse effects on PLWD, their caregivers, and the U.S. healthcare system. For PLWD, hospitalizations often result in hospital-induced delirium, falls, and discharge to nursing homes (Travers et al., 2014; Watkin et al., 2012). PLWD may also experience unmet needs such as poor pain management while in the hospital, as they may be unable to report pain levels or communicate clearly with hospital staff (Lichtner et al., 2016). PLWD hospitalizations can also result in caregiver strain because family caregivers often express anxiety and a sense of helplessness associated with their inability to positively influence their loved ones’ recovery (Boltz et al., 2015; Seidel & Thyrian, 2019). Additionally, after nursing home institutionalization, hospitalizations are the second most costly healthcare service utilization among PLWD (Alzheimer’s Association, 2022). Due to the myriad negative consequences of PLWD hospitalizations, it is vital to find preemptive methods to reduce hospital use.

Almost two-thirds of PLWD reside in the community and receive care in primary care settings (Alzheimer’s Association, 2022). Primary care providers identify most dementia cases, conduct comprehensive needs assessments, and coordinate care with relevant health care providers for PLWD (Spenceley et al., 2015). However, delivering high-quality dementia care is becoming increasingly challenging in primary care practices due to limited access to specialists and the increasing shortage of primary care physicians (Alzheimer’s Association, 2022). More specifically, the physician assistant and physician workforce is projected to increase by only 4.3% and 1%, respectively, for each year between 2016 and 2030; however, the nurse practitioner (NP) workforce will accelerate by nearly 7% in the same timeframe (Auerbach et al., 2018).

According to the American Association of Nurse Practitioners, NPs are licensed practitioners who have either masters or doctor of nursing practice degrees and are able to provide a variety of patient care services including assessment; ordering, performing, supervising and interpreting diagnostic and laboratory tests; and making diagnoses (American Association of Nurse Practitioners, 1993). NPs deliver high-quality and cost-effective care to older adults (Reilly et al., 2015; Woo et al., 2017). Yet, little is known about how to optimize the structural capabilities of primary care practices employing NPs who provide care to PLWD. Structural capabilities are defined as practice resources and features needed for delivering high-quality care (Friedberg et al., 2009). The availability of structural capabilities, such as care coordination, community integration, and reminder systems, reduce adverse patient outcomes among other populations (French et al., 2014; Johnson et al., 2013; Van’t Leven et al., 2013). Care coordination—the integration of personnel or activities used to manage patient care, including between primary and specialty providers (Bodenheimer, 2008)—reduces nursing home admissions among PLWD (Lee et al., 2020). Practices that incorporate community resources, such as referrals to the local Alzheimer’s Association, provide critical information about social services and support to improve the quality of life for both PLWD and their caregivers (Home and Community-Based Services for People With Dementia and Their Caregivers, 2013). Lastly, as many PLWD have multiple chronic conditions (Alzheimer’s Association, 2022; Phelan et al., 2012), reminder systems embedded within practices prompt clinicians to follow recommended guidelines for various chronic conditions such as asthma and diabetes. Reminder systems have also been linked with improved medication adherence (Tran et al., 2014).

The association between structural capabilities and hospitalizations has not been studied in primary care practices employing NPs and providing care to PLWD. This study assesses the relationship between the presence of structural capabilities (i.e., care coordination, community integration, and reminder systems) in primary care practices employing NPs and hospitalizations (i.e., ACSC and all-cause) among PLWD.

Methods

Study Design and Data Sources

This study was a secondary analysis of cross-sectional data produced from a large study on reducing racial and ethnic disparities among older adults with chronic conditions, including cardiovascular disease, asthma, hypertension, diabetes, or dementia. The parent study collected NP surveys and linked it with Medicare claims data to accomplish its aim. The NP surveys included information on the structural capabilities of primary care practices employing NPs, while the Medicare claims data contained patient-level information on patient characteristics and patient outcomes. Participation in the parent study was voluntary, and researchers informed participants that they may omit any questions if they felt uncomfortable answering. Researchers in the parent study also informed the NPs in detail how the data would be de-identified, used, and kept secure in the consent form. Because the study involved analyzing secondary data and data were de-identified, no informed consent was needed. Details of the parent and the present studies are described below. The Institutional Review Boards at the Columbia University School of Nursing approved both the parent and current studies completed.

Parent Study

NP surveys.

The parent study collected information about primary care practices including information on practice structural capabilities through NP surveys. The study identified NPs from a database called SK&A OneKey, which contained data on provider and practice names, practice locations, contact information, and National Provider Identifiers (NPIs) in the U.S. (DesRoches et al., 2015). NPs were recruited from six states: California, Pennsylvania, New Jersey, Washington, Texas, and Arizona.

The NP survey took place in 2018–2019. NPs working in primary care practices completed the questionnaires either on paper or online. Primary care practices in the parent study included NPs with family medicine, general practice, internal medicine, internal medicine/pediatrics, internal medicine/preventive medicine, general preventive medicine, or geriatrics specializations. Only primary care practices that employed at least one NP were included. The study used a Dillman approach for mixed-mode surveys to maximize the response rate (Dillman et al., 2014). Three separate postcard reminders and a second survey mailing were sent to nonrespondents to increase the response rate (Harrison et al., 2021). Nonrespondents were called three times to encourage them to complete the survey. Overall, 1,244 NPs across six states completed the survey. The study achieved a response rate of 21.9% (Harrison et al., 2021), which is similar to other large-scale surveys of nurses (Brooks Carthon et al., 2020; Lasater et al., 2019).

Medicare claims data.

The parent study obtained Medicare claims data from the six states for beneficiaries older than 65 with the following chronic conditions: asthma, chronic obstructive pulmonary disease, hypertension, congestive heart failure, cardiovascular disease, diabetes, and dementia. In Medicare claims, the researchers defined chronic conditions using the International Classification of Diseases Tenth Revision (ICD-10) diagnostic and procedure codes. Patient demographics (i.e., age, sex, and race) were attained from the Medicare Beneficiary Summary File (MBSF). Information about patient outcomes such as hospitalizations was obtained from the Medicare claims data.

Attribution.

The parent study used a common attribution approach (Mehrotra et al., 2010). The researchers attributed beneficiaries to clinicians by NPI and the beneficiaries to practices by the practice identifier available in the OneKey database. First, they calculated the proportion of primary care evaluation and management (E&M) paid amounts provided to a given Medicare beneficiary by each primary care provider that submitted a claim for that Medicare beneficiary in 2018–2019. Next, the beneficiary was assigned to the clinician who provided the highest proportion of E&M paid amounts as long as that provider accounted for 30% or more of E&M expenditures (Mehrotra et al., 2010). This minimum threshold was set to ensure that the beneficiary and provider had a reasonably strong patient-provider relationship. Medicare beneficiaries who did not have a dominant provider were not included in this analysis. Next, beneficiaries were linked to practice in the survey through their dominant provider’s NPI. Each beneficiary’s dominant primary care practice was determined by linking the dominant physician or NP NPI to a primary care practice associated with an NP in the survey. More specifically, we used both physician and NP patients receiving care from the practices that were included in this study. A practice was randomly selected for rare cases (<3%) when clinicians had multiple practices in the survey.

Current Study

To accomplish the aims of our study, we used three datasets: NP surveys, Medicare claims, and the Minimum Data Set (MDS). We selected primary care practices in California since California has the highest number of PLWD and the second-most Medicare spending per capita for this population ( Alzheimer’s Association, 2022). In total, 222 NPs who provided care to PLWD completed the surveys across 192 unique practices in California. The population of interest for this study was community-residing Medicare beneficiaries 65 years and older who had at least one service claim billed by NPs or physicians for Alzheimer’s Disease or Related Dementias (ICD-10 codes F0150, F0151, F0280, F0281, F0390, F0391, F04, G138, F05, F061, F068, G300, G301, G308, G309, G311, G312, G3101, G3109, G94, R4181, and R54). Community-residing Medicare beneficiaries were identified as patients who had less than 100 nursing home (Amjad et al., 2019). We used the MDS data for admission and discharge records to identify community-residing beneficiaries. Then, we linked the MDS data with Medicare claims data to identify PLWD using the ICD-10 codes. Based on the criteria above, we identified 5,001 PLWD across 192 practices in California to be the sample in the current study.

Measures

Independent variables.

As part of the survey, NPs provided information on structural capabilities—the study’s independent variables—by completing the validated Primary Care Practice Site Survey (Friedberg et al., 2010). Structural capabilities included in this study were community integration, reminder systems, and care coordination.

Consistent with previous studies, we created practice-level scales for each structural capability (Friedberg, Coltin, et al., 2009; Friedberg, Safran, et al., 2009; Martsolf et al., 2018). The care coordination scale refers to whether there was a designated staff member for care coordination services. The community integration scale refers to whether practices employing NPs had agreements with community services and/or a referral system linking their patients to community programs. The reminder systems scale refers to whether the practices had a system to follow recommended guidelines for chronic diseases such as hypertension and diabetes.

All items in each scale were operationalized as binary variables so that the presence or absence (yes/no) was denoted for each structural capability. Subsequently, practices were determined to have a particular structural capability if they had more than half of the relevant structural capability items. The corresponding survey items for each structural capability evaluated in this study appear in Table 1.

Table 1.

Corresponding Survey Items for Each Structural Capability

Structural Capabilities
Care Coordination
Creating and managing patient problem lists
Providing resources to assist patients in self-management
Medication management
Contacting patients when they are due for needed services
Helping patients access community and social services
Helping patients schedule appointments with various providers
Working with other clinicians to better coordinate patient care
Community Integration
Agreements with community service agencies (e.g., health departments) to enhance services for patients
Referral system for linking patients to community programs
Reminder Systems for Patients with:
Asthma/COPD
Cardiovascular disease
Hypertension
Congestive heart failure
Diabetes
Open in a new tab

Abbreviations. COPD: chronic obstructive pulmonary disease

Dependent variables.

This study had two dependent variables: all-cause hospitalizations and ACSC hospitalizations. All-cause hospitalizations were defined as any length of stay of more than one day in a given 12-month period as recorded in the Medicare inpatient claims. ACSC hospitalizations were determined by satisfying one or more criteria from the Agency for Healthcare Research and Quality “Prevention Quality Indicators Technical Specifications” Version 2020 criteria (“Prevention Quality Indicators Technical Specifications.,” 2020). Both ACSC and all-cause hospitalizations were operationalized as binary variables in this study.

Covariates.

We controlled for patient age, sex, race, comorbidities, practice type, and geographical classification as covariates in our analyses. We obtained patient demographics from the MBSF referenced in the parent study section. Age was operationalized as a continuous variable, sex as a binary variable, and race as a categorical variable with the following categories: non-Hispanic white, non-Hispanic black, Hispanic, Asian/Pacific Islander, American Indian/Alaska Native, and other. We identified all chronic conditions listed in Table 2 with ICD-10 diagnostic codes in Medicare claims data and operationalized them as binary variables. Practice-type information was obtained from NP surveys. For the geographic classification (i.e., urban and rural) of the practice, we used ZIP code Version 3.1 of the Rural-Urban Community Area (RUCA) codes (University of North Dakota, 2014). RUCA categories are rural-urban classifications based on population density and work commuting patterns (U.S. Department of Agriculture Economic Research Service,2019).

Table 2.

Patient Characteristics

Patient Characteristics Overall (n = 85,962) PLWD (n = 5,001) Patients without Dementia (n = 80,961) P
Age, M (SD) 75.27 (7.65) 83.04 (8.24) 74.79 (7.35) <.0001
Age, n (%) <.0001
 65–74 46,132 (53.67) 852 (17.02) 45,280 (55.93)
 75–84 27,886 (32.44) 1,910 (38.15) 25,976 (32.09)
 85≤ 11,944 (13.89) 2,244 (44.83) 9,700 (11.98)
Sex, n (%) <.0001
 Female 51,265 (59.64) 3,217 (64.26) 48,048 (59.35)
 Male 34,697 (40.36) 1,789 (35.74) 32,908 (40.65)
Race, n (%) <.0001
 White or Caucasian 60,065 (69.87) 3,217 (64.26) 56,848 (70.22)
 Black or African American 3,589 (4.18) 276 (5.51) 3,313 (4.09)
 Hispanic 10,727 (12.48) 697 (13.92) 10,030 (12.39)
 Asian 8,531 (9.92) 689 (13.76) 7,842 (9.69)
 Other (Native Hawaiian or other Pacific Islander, 3,050 (3.55) 127 (2.54) 2,923 (3.61)
 American Indian or Alaska Native, and other)
Chronic Conditions, n (%)
 Hypertension 61,866 (71.97) 4,026 (80.42) 57,840 (71.45) <.0001
 Peripheral vascular disease 15,283 (17.78) 1,592 (31.80) 13,691 (16.91) <.0001
 Diabetes 24,403 (28.39) 1,643 (32.82) 22,760 (28.11) <.0001
 Cerebrovascular disease 11,260 (13.10) 1,515 (30.26) 9,745 (12.04) <.0001
 Depression 13,479 (15.68) 1,446 (28.89) 12,033 (14.86) <.0001
 CVD 17,870 (20.79) 1,434 (28.65) 16,436 (20.30) <.0001
 Renal disease 14,632 (17.02) 1,380 (27.57) 13,252 (16.37) <.0001
 CHF 10,330 (12.02) 1,158 (23.13) 9,172 (11.33) <.0001
 COPD 13,008 (15.13) 1,022 (20.42) 11,986 (14.81) <.0001
Health care utilization, n (%)
 All-cause hospitalizations 12,411 (14.44) 1,532 (30.60) 10,879 (13.44) <.0001
 ACSC hospitalizations 1,720 (2.00) 266 (5.31) 1,454 (1.80) <.0001
Practice type, n (%) <.0001
 Physician practice 42,805 (49.80) 2,352 (46.98) 40,453 (49.97)
 Community health center 15,822 (18.41) 755 (15.08) 15,067 (18.61)
 Hospital-based clinic 9,429 (10.97) 741 (14.80) 8,688 (10.73)
 Other (retail or urgent care center, nurse-managed clinic, and others) 17,906 (20.83) 1,158 (23.13) 16,748 (20.69)
Geographic classification, n (%) <.0001
 Urban 78,424 (91.23) 4,657 (93.03) 73,767 (91.12)
 Rural 7,538 (8.77) 349 (6.97) 7,189 (8.36)
Open in a new tab

Abbreviations. PLWD: persons living with dementia, CVD: Cardiovascular disease, COPD: chronic obstructive pulmonary disease, CHF: chronic heart failure

Note. P-values were generated from chi-square tests for categorical variables and t tests for continuous variables.

Data analysis.

We calculated the descriptive statistics, including the frequency, mean, and standard deviation of PLWD characteristics (i.e., age, sex, race, comorbidities, practice type, and geographical classification), and compared them with the same characteristics of patients without dementia. To identify significant differences between the characteristics of PLWD and those without dementia, we used a t-test for the continuous variable of age and chi-square tests for the remainder of the variables, which were all categorical. To determine the prevalence of structural capabilities across the practices that provided care to PLWD, we calculated the presence of each structural capability.

To assess for multicollinearity, we calculated variance inflation factors and found a low risk of collinearity among our variables. We used generalized estimating equations in all of our logistic regression models to account for the clustering effect of patients within the practices. First, we developed unadjusted models to assess the bivariate relationship between structural capabilities (i.e., availability of care coordination, community integration, and reminder systems) and ACSC and all-cause hospitalizations. Then, we calculated fully adjusted models accounting for patient demographics, the chronic conditions included in Table 2, practice type, and geographical classification. We conducted data analyses using Statistical Analysis System (SAS) Version 9.4.

Results

Patient Characteristics

Table 2 shows the characteristics of PLWD and those without dementia. Our sample included 5,001 PLWD and 80,961 older adults without dementia. There were notable differences between the characteristics of PLWD compared to older adults without dementia. For instance, PLWD were older (M = 83.04) compared to those without dementia (M = 74.79; p < .0001). The majority of both PLWD and those without dementia were female and White. PLWD had more chronic conditions compared to older adults without dementia. For instance, over 80% of PLWD had hypertension compared to 71% of older adults without dementia (p < .0001). Healthcare utilization was also higher among PLWD compared to older adults without dementia. For instance, over 30% of PLWD had at least one all-cause hospitalization compared to 13% of older adults without dementia (p < .0001). Most of the PLWD (46.98%) and older adults without dementia (49.97%) received care from physician-led practices. Lastly, over 93% of PLWD and 91% of patients without dementia received care from clinics in urban areas.

Prevalence of Structural Capabilities and the Association with Hospitalizations

Table 3 shows the frequency of structural capabilities in PLWD-serving practices. Approximately 35% of PLWD-serving practices had care coordination, 57% had reminder systems, and 64% had community integration. Regression models show the association between structural capabilities and hospitalizations in PLWD practices. Only one regression model showed statistically significant findings. Indeed, an adjusted logistic regression model showed that PLWD who received care from practices with care coordination were less likely to have ACSC hospitalization (OR = 0.62, p < .05). The rest of our adjusted and unadjusted logistic regression analyses showed no association between the availability of community integration and reminder systems in primary care practices and ACSC or all-cause hospitalizations among PLWD (see Table 4). We also conducted sensitivity analysis and found similar findings (see Appendix, Table 1).

Table 3.

Frequency of Structural Capabilities in PLWD-Serving Practices

Practices Providing Care to PLWD (n =192)
Care Coordination n (%) 68 (35.42)
Community Integration n (%) 122 (63.54)
Reminder Systems n (%) 110 (57.29)
Open in a new tab

Abbreviations. PLWD: persons living with dementia

Table 4.

The Association Between Structural Capabilities and Hospitalization Among PLWD

Care Coordination
Outcome Model 1: Unadjusted Model 2: Adjusted
OR (95% CI) P OR (95% CI) P
All-cause hospitalization 0.93 (0.69, 1.25) 0.612 1.16 (0.91, 1.14) 0.231
ACSC hospitalization 0.71 (0.36, 1.37) 0.307 0.62 (0.40, 0.97) 0.035
Community Integration
Outcome Model 1: Unadjusted Model 2: Adjusted
OR (95% CI) P OR (95% CI) P
All-cause hospitalization 1.26 (0.96, 1.66) 0.100 1.10 (0.90, 1.34) 0.351
ACSC hospitalization 1.26 (0.76, 2.09) 0.373 0.92 (0.64, 1.32) 0.637
Reminder Systems
Outcome Model 1: Unadjusted Model 2: Adjusted
OR (95% CI) P OR (95% CI) P
All-cause hospitalization 1.01 (0.73, 1.40) 0.941 0.99 (0.80, 1.22) 0.916
ACSC hospitalization 1.06 (0.59, 1.91) 0.848 1.08 (0.76, 1.54) 0.657
Open in a new tab

Abbreviations. PLWD: persons living with dementia, OR: odds ratio, CI: confidence interval, ACSC: ambulatory care sensitive conditions

Note. The reference group was practices with less than half structural capabilities.

Discussion

This study assessed the relationship between structural capabilities and ACSC and all-cause hospitalizations among PLWD who receive care from primary care practices that employ NPs. Our findings showed that PLWD who received care from practices with care coordination were less likely to have ACSC hospitalizations than PLWD who received care from practices that did not have care coordination. Evidence also shows that care coordination for older adults, including PLWD, can improve the quality of care and quality of life (Hughes et al., 2017). Therefore, care coordination, which involves the integration of personnel or activities used to manage patient care, including primary and specialty, can address the multidisciplinary needs of PLWD and reduce ACSC hospitalizations.

While community integration and reminder systems are associated with improvements in the outcomes of other populations, our findings were not consistent with the literature regarding the care for PLWD. Our results showed that the presence of such structural capabilities is not associated with hospitalizations among PLWD. There may be several reasons why our findings for PLWD differ from the research on structural capabilities in other chronic diseases. First, because of deficits in memory, language, and judgment, PLWD often face difficulties reporting symptoms, adhering to medications, reporting medication side effects, and complying with treatment and follow-up recommendations (Gual et al., 2018; Nelis et al., 2019). Cognitive impairment complicates providing care to PLWD, so structural capabilities should be tailored to address PLWD needs in primary care settings. For instance, providing PLWD with continuous long-term care from a single provider, addressing the nonmedical needs of PLWD (e.g., transportation), and offering comprehensive caregiver support (e.g., educational resources) in primary care settings can potentially lead to fewer hospitalizations of PLWD (Bass et al., 2015; Larson & Stroud, 2021).

Another possible explanation is that providers, PLWD, and caregivers do not optimally utilize the available structural capabilities in primary care practices. For example, primary care providers may not view referrals to community resources as a priority due to the time constraints of a busy practice (Sadarangani et al., 2020). Research shows that when providers collaborate with community organizations instead of simply using them as a referral source, benefits to patients and their caregivers may increase exponentially (Sadarangani et al., 2020). Lastly, such services are not broadly adopted because they are poorly reimbursed (Boustani et al., 2019). Evaluating alternative incentive structures and payment models that improve the availability and quality of tailored structural capabilities could improve patient outcomes and may ultimately lead to fewer hospitalizations of PLWD. Instead of only focusing on the availability of structural capabilities, future studies should explore whether providers, PLWD, and their caregivers are using available structural capabilities to their full potential. In addition, as policymakers are creating legislation and healthcare guidance focused on developing alternative payment models and incentives for utilizing structural capabilities, future studies should evaluate if alternative payment models improve patient outcomes among PLWD.

Future studies should also focus on the stages of dementia, as each stage requires special medical attention and referral to stage-specific psychosocial services. Because dementia has mild, moderate, and severe stages (Alzheimer’s Association, 2022), PLWD in different stages have different care requirements. PLWD with moderate and severe dementia have more complex medical needs than those with mild dementia. They face more difficulties communicating and performing routine tasks, including activities of daily living and managing their comorbidities (Alzheimer’s Association, 2022), which puts them at a higher risk of hospitalization. Therefore, to accurately assess how structural capabilities are associated with hospitalizations of PLWD, future studies should include information about the stage of the disease in their analysis.

Additionally, research shows that high caregiver strain is prevalent among those who care for PLWD with moderate and advanced stages (Boltz et al., 2018). As dementia progresses, caregivers experience increasing strain and burnout (Alzheimer’s Association, 2022). High levels of caregiver strain are associated with worse patient outcomes, including hospitalizations (Boltz et al., 2018). Thus, primary care practices should design structural capabilities that address caregiver needs. To accomplish this goal and reduce adverse patient outcomes, future research should include qualitative studies that gather data on optimizing primary care structural capabilities to address the needs of PLWD and caregivers.

Lastly, we also investigated the characteristics of PLWD and those without dementia. Our findings are consistent with the literature. For instance, the PLWD in our study are older than persons without dementia (Tom et al., 2015). Dementia incidence also increased after the age of 85 and was greater among women than men (Beama et al., 2018). Focusing future research on this population may be beneficial in discovering the optimal structural capability designs that have the most significant association with outcomes such as hospitalizations. Our findings also showed that PLWD have more chronic conditions than persons without dementia (Clague et al., 2017; Phelan et al., 2012). This is notable because PLWD with comorbidities have increased rates of polypharmacy, hospitalizations, and cognitive and functional decline (Clague et al., 2017; P.-J. Lin et al., 2013; Melis et al., 2013). Therefore, it is crucial to design structural capabilities that manage comorbidities to reduce hospitalizations of PLWD.

Limitations

This study has some limitations. First, we used secondary data analyses to accomplish our aims, the limits of which have been discussed previously (Cheng & Phillips, 2014). For instance, we used datasets from the parent study that were produced for other purposes, and some covariates (e.g., patient educational level and socioeconomic status) were not available for the analysis in our study, which is a significant weakness. Lower educational levels and socioeconomic status have been associated with a higher risk of a dementia diagnosis and worse patient outcomes (Maccora et al., 2020; Petersen et al., 2021). Thus, not having access to such covariates limits our understanding of their influence on hospitalizations. Furthermore, dementia is a progressive neurodegenerative disease with distinct stages and corresponding care requirements. We could not assess the stages of dementia because Medicare claims do not contain this information. Thus, we could not address how each stage may affect the outcome of our study. Lastly, because the study has a cross-sectional design, we could not test for causality.

Conclusion

Findings from this study generated important evidence about the association between structural capabilities and ACSC and all-cause hospitalizations among PLWD. We found that care coordination – a service that integrates and improves access to medical and nonmedical services, while providing information, coaching, and support for managing chronic conditions at home – was the only structural capability that affected the reduction of ACSC hospitalizations. Our findings highlight the crucial importance of incorporating care coordination services in primary care practices that serve PLWD and tailoring other structural capabilities to meet the unmet needs of this vulnerable population.

Funding Statement

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institute on Aging grant number 1R36AG071946–01 (Hovsepian, PI), the National Institute on Minority Health and Health Disparities grant number R01MD011514–03S1 (Poghosyan, PI), the National Institute of Nursing Research grant number T32-NR007104 (Aiken, McHugh, Co-PIs) and the National Institute of Nursing Research T32-NR014205 (Poghosyan, Stone, Co-PIs) ).

Risks to Human Subjects and Adequacy of Protection Against Risks Human Subjects Involvement, Characteristics, and Design

This study is a secondary data analysis of existing data produced from nurse practitioner (NP) surveys and Medicare claims data on all-cause and potentially avoidable hospitalizations for older adults with Alzheimer’s disease and Alzheimer’s disease-related dementias (AD/ADRD). This study will be accomplished using data from an R01 grant and its affiliated Alzheimer’s-focused administrative supplement (PI: Poghosyan, R01 MD011514–03S1). This study focuses on the influence of structural capabilities in primary care practices employing NPs on all-cause and potentially avoidable hospitalizations among patients with AD/ADRD. NP surveys include information about structural capabilities of primary care settings that employ NPs and Medicare claims data contain information on hospitalizations and patient characteristics. The collected human subjects data (i.e., NP surveys and Medicare beneficiaries) are de-identified. Thus, we expect minimal risks involved with this study. As mandated by CUMIC’s IRB training requirements for ethical conduct of research, the authors have completed training regarding the protection of human subjects.

Appendix

Table 1.

Sensitivity Analysis for the Association Between Structural Capabilities and Hospitalization Among PLWD

Care Coordination
Outcome Model 1: Unadjusted Model 2: Adjusted
OR (95% CI) P OR (95% CI) P
All-cause hospitalization 0.93 (0.68–1.26) 0.628 0.91 (0.71–1.16) 0.455
ACSC hospitalization 0.71 (0.40, 1.28) 0.307 0.65 (0.45, 0.95) 0.026
Community Integration
Outcome Model 1: Unadjusted Model 2: Adjusted
OR (95% CI) P OR (95% CI) P
All-cause hospitalization 1.26 (0.96, 1.66) 0.099 1.10 (0.90, 1.35) 0.351
ACSC hospitalization 1.26 (0.76, 2.09) 0.373 0.92 (0.64, 1.32) 0.637
Reminder Systems
Outcome Model 1: Unadjusted Model 2: Adjusted
OR (95% CI) P OR (95% CI) P
All-cause hospitalization 1.02 (0.75, 1.39) 0.902 0.90 (0.81, 1.20) 0.900
ACSC hospitalization 1.05 (0.58, 1.87) 0.848 1.05 (0.74, 1.49) 0.801
Open in a new tab

Abbreviations. PLWD: persons living with dementia, OR: odds ratio, CI: confidence interval, ACSC: ambulatory care sensitive conditions

Note. The reference group was practices with less than one-third of structural capabilities.

Footnotes

Informed Consent and Assent

Because the study involves analyzing secondary data and data are de-identified, no informed consent is needed.

Conflict of Interest Statement

The authors confirm that there are no relevant financial or non-financial competing interests to report.

References

  1. Alzheimer’s Association. (2022). 2022 Alzheimer ‘s disease facts and figures. Alzheimer’s & Dementia, 15, 321–387. [Google Scholar]
  2. American Association of Nurse Practitioners. (1993). Discussion paper: Scope of Practice for Nurse Practitioners. 10.1136/ip.8.suppl_2.ii9 [DOI]
  3. Amjad H, Carmichael D, Austin AM, Chang C, & Bynum JPW (2019). Continuity of Care and Health Care Utilization in Older Adults With Dementia in Fee-for-Service Medicare. JAMA Internal Medicine, 21224(9), 1371–1378. 10.1001/jamainternmed.2016.3553 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Auerbach DI, Staiger DO, & Buerhaus PI (2018). Growing Ranks of Advanced Practice Clinicians — Implications for the Physician Workforce. New England Journal of Medicine, 378(25), 2358–2360. 10.1056/nejmp1801869 [DOI] [PubMed] [Google Scholar]
  5. Bass DM, Judge KS, Maslow K, Wilson NL, Morgan RO, McCarthy CA, Looman WJ, Snow AL, & Kunik ME (2015). Impact of the care coordination program “Partners in Dementia Care” on veterans’ hospital admissions and emergency department visits. Alzheimer’s & Dementia (New York, N. Y.), 1(1), 13–22. 10.1016/j.trci.2015.03.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Billings J, Zeitel L, Lukomnik J, Carey TS, Blank AE, & Newman L (1993). Impact Of Socioeconomic Status On Hospital Use In New York City: by Billings John, Zeitel Lisa, Lukomnik Joanne, Carey Timothy S., Blank Arthur E., and Newman Laurie. Health Affairs. [DOI] [PubMed] [Google Scholar]
  7. Bodenheimer T (2008). Coordinating Care — A Perilous Journey through the Health Care System. New England Journal of Medicine, 358(10), 1064–1071. 10.1056/nejmhpr0706165 [DOI] [PubMed] [Google Scholar]
  8. Boltz M, Chippendale T, Resnick B, & Galvin JE (2015). Anxiety in family caregivers of hospitalized persons with dementia: Contributing factors and responses. Physiology & Behavior, 29(3), 236–241. 10.1097/WAD.0000000000000072 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Boltz M, Lee KH, Chippendale T, & Trotta RL (2018). Pre-admission functional decline in hospitalized persons with dementia: The influence of family caregiver factors. Archives of Gerontology and Geriatrics, 74(September 2017), 49–54. 10.1016/j.archger.2017.09.006 [DOI] [PubMed] [Google Scholar]
  10. Boustani M, Alder CA, Solid CA, & Reuben D (2019). An alternative payment model to support widespread use of collaborative dementia care models. Health Affairs, 38(1), 54–59. 10.1377/hlthaff.2018.05154 [DOI] [PubMed] [Google Scholar]
  11. Brooks Carthon JM, Brom H, Poghosyan L, Daus M, Todd B, & Aiken L (2020). Supportive Clinical Practice Environments Associated With Patient-Centered Care. Journal for Nurse Practitioners, 16(4), 294–298. 10.1016/j.nurpra.2020.01.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Beama C, Kaneshiroa C, Janga J, Reynoldsb C, Pedersena N, and M. G. (2018). Differences Between Women and Men in Incidence Rates of Dementia and Alzheimer’s Disease Christopher. Journal of Alzheimer’s Disease : JAD, 64(4), 1077–1083. 10.1016/j.physbeh.2017.03.040 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cheng HG, & Phillips MR (2014). Secondary analysis of existing data: opportunities and implementation. Shanghai Archives of Psychiatry, 26(6), 371–375. 10.11919/j.issn.1002-0829.214171 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Clague F, Mercer SW, Mclean G, Reynish E, & Guthrie B (2017). Comorbidity and polypharmacy in people with dementia: Insights from a large, population-based cross-sectional analysis of primary care data. Age and Ageing, 46(1), 33–39. 10.1093/ageing/afw176 [DOI] [PubMed] [Google Scholar]
  15. DesRoches CM, Barrett KA, Harvey BE, Kogan R, Reschovsky JD, Landon BE, Casalino LP, Shortell SM, & Rich EC (2015). The results are only as good as the sample: Assessing three national physician sampling frames. Journal of General Internal Medicine, 30, 595–601. 10.1007/s11606-015-3380-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Dillman D, Smyth J, & Christian L . (2014). Internet, Phone Mail, and mixed-mode surveys: The tailored design method.
  17. Fingar KR, Barrett ML, Elixhauser A, Stocks C, & Steiner CA (2015). Trends in potentially preventable inpatient hospital admissions and emergency department visits. Healthcare Cost and Utilization Project (HCUP) Statistical Briefs, May 2015, 1–20. http://www.ncbi.nlm.nih.gov/pubmed/26741014 [PubMed]
  18. French DD, LaMantia MA, Livin LR, Herceg D, Alder CA, & Boustani MA (2014). Healthy aging brain center improved care coordination and produced net savings. Health Affairs, 33(4), 613–618. 10.1377/hlthaff.2013.1221 [DOI] [PubMed] [Google Scholar]
  19. Friedberg MW, Coltin KL, Safran DG, Dresser M, & Schneider EC (2010). Medical home capabilities of primary care practices that serve sociodemographically vulnerable neighborhoods. Archives of Internal Medicine, 170(11), 938–944. 10.1001/archinternmed.2010.110 [DOI] [PubMed] [Google Scholar]
  20. Friedberg MW, Coltin KL, Safran DG, Dresser M, Zaslavsky AM, & Schneider EC (2009). Associations between structural capabilities of primary care practices and performance on selected quality measures. Annals of Internal Medicine, 151(7), 456–463. 10.7326/0003-4819-151-7-200910060-00006 [DOI] [PubMed] [Google Scholar]
  21. Friedberg MW, Safran DG, Coltin KL, Dresser M, & Schneider EC (2009). Readiness for the patient-centered medical home: Structural capabilities of massachusetts primary care practices. Journal of General Internal Medicine, 24(2), 162–169. 10.1007/s11606-008-0856-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gao J, Moran E, Li YF, & Almenoff PL (2014). Predicting potentially avoidable hospitalizations. Medical Care, 52(2), 164–171. 10.1097/MLR.0000000000000041 [DOI] [PubMed] [Google Scholar]
  23. Gual N, Morandi A, Pérez LM, Brítez L, Burbano P, Man F, & Inzitari M (2018). Risk factors and outcomes of delirium in older patients admitted to postacute care with and without dementia. Dementia and Geriatric Cognitive Disorders, 45(1–2), 121–129. 10.1159/000485794 [DOI] [PubMed] [Google Scholar]
  24. Harrison JM, Germack HD, Poghosyan L, D’Aunno T, & Martsolf GR (2021). Methodology for a six-state survey of primary care nurse practitioners. Nursing Outlook, 0, 2–9. 10.1016/j.outlook.2021.01.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Home and Community-Based Services for People with Dementia and their Caregivers. Single Point of Entry Systems (2013), Alzheimer’s Association. (January), 1–12.
  26. Hughes S, Lepore M, Wiener JM, & Gould E (2017). Research on Care Coordination for People with Dementia and Family Caregivers Background Paper. April.
  27. Johnson DK, Niedens M, Wilson JR, Swartzendruber L, Yeager A, & Jones K (2013). Treatment outcomes of a crisis intervention program for dementia with severe psychiatric complications: The kansas bridge project. Gerontologist, 53(1), 102–112. 10.1093/geront/gns104 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Larson EB, & Stroud C (2021). Meeting the challenge of caring for persons living with dementia and their care partners and caregivers. In Meeting the Challenge of Caring for Persons Living with Dementia and Their Care Partners and Caregivers. 10.17226/26026 [DOI] [PubMed]
  29. Lasater KB, Jarrín OF, Aiken LH, McHugh MD, Sloane DM, & Smith HL (2019). A Methodology for Studying Organizational Performance: A Multistate Survey of Front-line Providers. Medical Care, 57(9), 742–749. 10.1097/MLR.0000000000001167 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lee DCA, Tirlea L, & Haines TP (2020). Non-pharmacological interventions to prevent hospital or nursing home admissions among community-dwelling older people with dementia: A systematic review and meta-analysis. Health and Social Care in the Community, 28(5), 1408–1429. 10.1111/hsc.12984 [DOI] [PubMed] [Google Scholar]
  31. Lichtner V, Dowding D, Allcock N, Keady J, Sampson EL, Briggs M, Corbett A, James K, Lasrado R, Swarbrick C, & Closs SJ (2016). The assessment and management of pain in patients with dementia in hospital settings: a multi-case exploratory study from a decision making perspective. BMC Health Services Research, 16(1), 427. 10.1186/s12913-016-1690-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Lin P-J, Fillit HM, Cohen JT, & Neumann PJ (2013). Potentially avoidable hospitalizations among medicare beneficiaries with Alzheimer’s disease and related disorders. Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association, 9(1), 30–38. 10.1016/j.jalz.2012.11.002 [DOI] [PubMed] [Google Scholar]
  33. Lin PJ, Zhong Y, Fillit HM, Cohen JT, & Neumann PJ (2017). Hospitalizations for ambulatory care sensitive conditions and unplanned readmissions among Medicare beneficiaries with Alzheimer’s disease. Alzheimer’s and Dementia, 13(10), 1174–1178. 10.1016/j.jalz.2017.08.010 [DOI] [PubMed] [Google Scholar]
  34. Maccora J, Peters R, & Anstey KJ (2020). What does (low) education mean in terms of dementia risk? A systematic review and meta-analysis highlighting inconsistency in measuring and operationalising education. SSM - Population Health, 12, 100654. 10.1016/j.ssmph.2020.100654 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Martsolf GR, Ashwood S, Friedberg MW, & Rodriguez HP (2018). Linking structural capabilities and workplace climate in community health centers. The Journal of Health Care, 55. 10.1177/0046958018794542 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Mehrotra A, Adams JL, Thomas JW, & McGlynn EA (2010). The effect of different attribution rules on individual physician cost profiles. Annals of Internal Medicine, 152(10), 649–654. 10.7326/0003-4819-152-10-201005180-00005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Melis RJF, Marengoni A, Rizzuto D, Teerenstra S, Kivipelto M, Angleman SB, & Fratiglioni L (2013). The influence of multimorbidity on clinical progression of dementia in a population-based cohort. PLoS ONE, 8(12). 10.1371/journal.pone.0084014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Nelis SM, Wu YT, Matthews FE, Martyr A, Quinn C, Rippon I, Rusted J, Thom JM, Kopelman MD, Hindle JV, Jones RW, & Clare L (2019). The impact of co-morbidity on the quality of life of people with dementia: Findings from the IDEAL study. Age and Ageing, 48(3), 361–367. 10.1093/ageing/afy155 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Petersen JD, Wehberg S, Packness A, Svensson NH, Hyldig N, Raunsgaard S, Andersen MK, Ryg J, Mercer SW, Søndergaard J, & Waldorff FB (2021). Association of Socioeconomic Status with Dementia Diagnosis among Older Adults in Denmark. JAMA Network Open, 4(5), 1–13. 10.1001/jamanetworkopen.2021.10432 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Phelan EA, Borson S, Grothaus L, Balch S, & Larson EB (2012). Association between incident dementia and risk of hospitalization. JAMA : The Journal of the American Medical Association, 307(2), 165–172. 10.1001/jama.2011.1964.Association [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Prevention quality indicators technical specifications. Rockville, Md.: U.S. Department of Health & Human Services, Agency for Healthcare Research and Quality, 2020. (https://www.qualityindicators.ahrq.gov/modules/pqi_resources.aspx#techspecs.) [Google Scholar]
  42. Reilly S, Miranda-Castillo C, Malouf R, Hoe J, Toot S, Challis D,OM (2015). Case management approaches to home support for people with dementia. Cochrane Database OfSystematic Reviews, 1. 10.1002/14651858.CD008345.pub2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Sadarangani TR, Salcedo V, Chodosh J, Kwon S, Trinh-shevrin C, & Yi S (2020). Redefining the Care Continuum to Create a Pipeline to Dementia Care for Minority Populations. 10.1177/2150132720921680 [DOI] [PMC free article] [PubMed]
  44. Seidel D, & Thyrian JR (2019). Burden of caring for people with dementia – Comparing family caregivers and professional caregivers. A descriptive study. Journal of Multidisciplinary Healthcare, 12, 655–663. 10.2147/JMDH.S209106 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Spenceley SM, Sedgwick N, & Keenan J (2015). Dementia care in the context of primary care reform: an integrative review. Aging & Mental Health, 19(2), 107–120. 10.1080/13607863.2014.920301 [DOI] [PubMed] [Google Scholar]
  46. Tom SE, Hubbard RA, Crane PK, Haneuse SJ, Bowen J, McCormick WC, McCurry S, & Larson EB (2015). Characterization of dementia and Alzheimer’s disease in an older population: Updated incidence and life expectancy with and without dementia. American Journal of Public Health, 105(2), 408–413. 10.2105/AJPH.2014.301935 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Tran N, Coffman JM, Sumino K, & Cabana MD (2014). Patient reminder systems and asthma medication adherence: A systematic review. Journal of Asthma, 51(5), 536–543. 10.3109/02770903.2014.888572 [DOI] [PubMed] [Google Scholar]
  48. Travers C, Byrne GJ, Pachana NA, Klein K, & Gray LC (2014). Prospective observational study of dementia in older patients admitted to acute hospitals. Australasian Journal on Ageing, 33(1), 55–58. 10.1111/ajag.12021 [DOI] [PubMed] [Google Scholar]
  49. U.S. Department of Agriculture Economic Research Service. (2019). Documentation: 2010 rural urban community area (RUCA) codes. https://www.ers.usda.gov/data-products/ruralurbancommuting-area-codes/documentation/
  50. Van’t Leven N, Prick AEJC, Groenewoud JG, Roelofs PDDM, De Lange J, & Pot AM (2013). Dyadic interventions for community-dwelling people with dementia and their family caregivers: A systematic review. International Psychogeriatrics, 25(10), 1581–1603. 10.1017/S1041610213000860 [DOI] [PubMed] [Google Scholar]
  51. Watkin L, Blanchard MR, Tookman A, & Sampson EL (2012). Prospective cohort study of adverse events in older people admitted to the acute general hospital: Risk factors and the impact of dementia. International Journal of Geriatric Psychiatry, 27(1), 76–82. 10.1002/gps.2693 [DOI] [PubMed] [Google Scholar]
  52. Woo BFY, Lee JXY, & Tam WWS (2017). The impact of the advanced practice nursing role on quality of care, clinical outcomes, patient satisfaction, and cost in the emergency and critical care settings: A systematic review. Human Resources for Health, 15(1). 10.1186/s12960-017-0237-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES