Linking Data to Determine Risk for 30-Day Readmissions in Dementia

ABSTRACT

Objective: The demand and the landscape of options for dementia care are growing. Standardization of care for persons with Alzheimer disease and related dementias (ADRD) lacks infrastructure across episodes of care. Use of electronic health records (EHRs) in practice settings yields valuable information that can enhance continuity of patient care. The objective of this study was to use EHR-derived variables to identify risk factors for 30-day readmissions in the ADRD population across episodes of care.

Study Design: Cross-sectional, retrospective study of older adults (aged ≥ 65 years) with ADRD discharged from a large urban academic medical center between October 1, 2018, and March 31, 2022.

Methods: Data extracted across episodes of care from the EHR included demographic characteristics, medical variables, and encounter variables.

Results: A total of 14,101 patients diagnosed with ADRD were included in the study. Factors associated with patients being more likely to experience 30-day hospital readmissions included advanced age, male sex, being a non-English speaker, having more severe comorbidities, staying in the hospital for more than 5 days, having had more than 1 surgical procedure in the prior 6 months, having had 3 or more inpatient admissions in the 6 months prior to index admission, having had more than 3 physician consultations in the prior 6 months, and having been discharged to settings other than home (all P < .05).

Conclusions: By utilizing the EHR to connect medical and encounter data across episodes of care, health care providers and administrators can gain valuable insight into identifying factors contributing to readmissions, which could be used to improve continuity of care for patients and caregivers, ultimately leading to better outcomes and reduced health care costs.

Am J Manag Care. 2025;31(12):In Press

Takeaway Points

This study linked electronic health record (EHR) data across episodes of care to determine risk for 30-day readmissions in older adults with Alzheimer disease and related dementias (ADRD).

• Utilizing the EHR to connect medical and encounter data across health care settings provides insight into the factors that contribute to readmissions.
• Using EHR data can improve continuity of care for patients and caregivers, which can lead to better outcomes and reduced health care costs.
• The identification of risk factors for 30-day readmissions in older adults with ADRD—such as being male, being a non-English speaker, having significant comorbidities, having more than 3 inpatient admissions in the 6 months prior to index admission, having more than 3 physician consultations in the prior 6 months, having had longer hospital stays, and having had at least 1 surgical procedure in the prior 6 months—provides insight into areas that need to be proactively addressed.

Health care systems in the US are shifting from a focus on quantity of care to a focus on quality of care, making the care process increasingly complex.1 Our society continues to age, and the number of older adults with dementia is increasing, resulting in a growing demand for dementia care and a rapidly expanding landscape of options. More than 6.7 million persons in the US are living with Alzheimer disease and related dementias (ADRD), which is projected to double by 2050.2 Reports indicate that approximately 10.7% of individuals 65 years and older have been diagnosed with ADRD.2 These individuals are twice as likely to be admitted to the hospital and have longer lengths of hospital stay, more readmissions, and higher mortality rates compared with older individuals without ADRD.3,4 Additionally, the readmission risk for individuals who speak languages other than English ranges from 15% to 25%, even when interpreter services are used at key points during the hospital stay.5-7 Findings of studies that address adverse outcomes have shown that dementia is associated with increased risk of mortality and 30-day readmissions.8,9 One factor that may contribute to these observations is the lack of standardization in dementia care within health care systems. Managing the care of patients with dementia poses several challenges for caregivers10 and health care providers.11,12 Individuals diagnosed with ADRD often experience fragmented care, resulting in a lack of coordination in treatment and inadequate measures to prevent readmissions. This impacts hospitals because readmissions are part of quality measurement, transparency, and improvements needed for value-based care in the inpatient setting.

Hospital readmissions are one of the key measures used for evaluating the quality of inpatient acute care and are the measure used in the Hospital Readmissions Reduction Program, which is part of a Medicare value-based purchasing program. CMS defines a hospital readmission as an admission to an acute care hospital within 30 days of discharge from the same or another acute care hospital. It uses an all-cause definition in that the readmissions do not need to be related to the cause of the initial hospitalization. This methodology was adopted by the National Quality Forum, which endorsed an all-cause hospital readmission measure for 30 days after hospital discharge. This measure is publicly reported as part of quality reporting programs.13

The electronic health record (EHR) can be a valuable tool for detecting risk factors for patients with ADRD, such as those relevant to hospital readmission. EHRs aid in analyzing past usage of hospital services, including previous emergency department (ED) visits and surgeries. By examining comprehensive information from the EHR, health care professionals can identify factors associated with potentially avoidable adverse outcomes to advance efficient and effective clinical care for this population. With the widespread implementation of EHRs across practice settings, this rich source of information can enhance our understanding of patient care and its continuity, ultimately leading to better outcomes and reduced health care costs. The value of EHR data lies in their ability to reflect real-world practice scenarios in multiple care settings.14,15 The objective of this study was to identify risk factors for 30-day readmission in the ADRD population across episodes of care using EHR-derived variables.

METHODS

Inclusion/Exclusion Criteria

This retrospective study included adults 65 years and older with ADRD who were discharged between October 1, 2018, and March 31, 2022, from a large urban academic medical center in Los Angeles, California. Data extracted from the EHR included demographic characteristics, medical data, and encounter variables. Patients with ADRD were identified based on all hospital encounters through Epic16 with the following International Statistical Classification of Diseases, Tenth Revision (ICD-10) codes: A81.00, F01.50, F01.51, F02.80, F02.81, F03.90, F03.91, F10.27, G23.1, G30.0, G30.1, G30.8, G30.9, G31.01, G31.09, G31.83, G31.84, G91.0, R41.81, and R54, which were logged in the problem list or past medical history. We also included patients who had any of the following FDA-approved ADRD medications in their inpatient medication list: donepezil (Aricept), memantine (Namenda), rivastigmine (Exelon), galantamine hydrobromide (Razadyne), or donepezil and memantine hydrochlorides (Namzaric). This ADRD identification method was validated in a previous study17 and is increasingly being used to define dementia status18,19 due to underdiagnosis of dementia in the patient medical record.20,21 Patients were excluded if they were flagged as opting out of research (requested not to include their data in any research initiative), were not diagnosed and coded with ADRD-relevant ICD-10 codes, or did not use ADRD-relevant medications.

Outcome Variable

The outcome was whether patients 65 years and older with identified ADRD had all-cause 30-day hospital readmissions (yes or no) after discharge from the acute care hospital. If the patient was readmitted to the acute care hospital within 30 days for any reason, it was coded as a dichotomous variable (yes or no) for readmission. This was based on the CMS definition of 30-day all-cause readmissions.

Predictor Variables

Demographic, encounter, and medical variables for patients with and without all-cause 30-day hospital readmissions were examined. Demographic data included age at the time of index admission, sex (male, female, unknown), self-reported ethnicity (Hispanic or Latino, non-Hispanic, unknown), self-reported race (White; Black; Asian; others including American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, or others; unknown), marital status (single, widowed, married, unknown), language (English, non-English [which included Russian, Farsi, Spanish, Korean, and others]), and employment (employed, not employed, disabled, retired). Encounter variables included type of admission (elective, nonelective), admission source (home, facility), admission department (medical, surgical, others), intensive care unit (ICU) visits (total ICU, medical ICU, surgical ICU, unknown), discharge disposition (home, inpatient rehabilitation facility [IRF], skilled nursing facility [SNF], died, others [which included critical access hospitals, federal health care facilities, acute hospitals, inpatient psychiatric hospitals, long-term care, and other facilities]), ED visits 6 months prior to index admission (0 visits, 1 visit, 2 visits, ≥ 3 visits), inpatient admissions 6 months prior to index admission (0 visits, 1 visit, 2 visits, ≥ 3 visits), and payer (Medicare, non-Medicare). We determined these frequency categories of prior inpatient admission and ED visits based on clinical judgment and data distributions to represent different levels of postdischarge hospital readmission risk. Other continuous encounter variables included length of stay in days; days in the ICU; days in the hospital prior to mortality; total consults (0, ≤ 3, > 3); total number of physicians involved in patient’s care (1-11, > 11); and total number of other nonphysician consults, which included providers other than physicians who provided consults, such as physical therapists, occupational therapists, respiratory therapists, dietitians, social workers, or case managers (1-11, > 11). Medical variables included the Elixhauser Comorbidity Index score and surgery counts (1, > 1 [range = 1-25, with 9709 missing]).

Data Analysis

We first conducted descriptive analyses to examine the demographics of the study sample. We then used independent t tests for continuous variables and χ2 tests for categorical variables to assess the differences in demographic, encounter, and medical variables between patients with vs without hospital readmission (P < .05) because the data were normally distributed.

Bivariate analysis (t test and χ2 test) was used to assess the differences in demographic, encounter, and medical variables. We conducted a multicollinearity test to identify variables with high variance inflation factors (VIFs) if the VIF values exceed 5. Using this rule, we excluded 2 variables with high VIFs, physicians seen and nonphysicians seen, from our final logistic regression model. We used logistic regression models to determine risk factors associated with 30-day readmissions based on baseline demographic, medical, and encounter variables. Statistical analysis was performed using SAS 9.4 (SAS Institute Inc).This study was approved by the Cedars-Sinai Medical Center Institutional Review Board (IRB 00002192).

RESULTS

Demographics

Table 1 provides demographic characteristics and encounter experiences of patients with ADRD (N = 14,101). The majority were female (57.1%), non-Hispanic (91.7%), White (72.2%), widowed (37.0%), nonelectively admitted to the hospital (92.5%), covered by Medicare (83.28%), and admitted from home (79.0%); spoke English (70.9%); had 1 ED visit in the 6 months prior to index admission (47.4%); and had 1 inpatient admission in the 6 months prior to index admission (59.7%). The mean (SD) age of the sample was 84.24 (8.4) years, and the mean (SD) length of stay was 7.7 (11.0) days. The mean (SD) Elixhauser Comorbidity Index score was 12.9 (9.1).

The majority were discharged to home (47.9%), followed by SNFs (35.8%), other discharge sites (eg, critical access hospitals, federal health care facilities, acute hospitals, inpatient psychiatric hospitals, long-term care hospitals) (8.3%), and IRFs (7.96%). For patients with an ICU stay (n = 2336), 35.4% were in the medical ICU and 9.4% were in the surgical ICU; ICU type was unknown for 45.5%.

Patient Characteristics Associated With 30-Day Hospital Readmissions

Table 1 also illustrates patient characteristics and encounter experiences stratified by 30-day hospital readmissions. Patients with readmissions had significantly longer lengths of stay; had higher discharge ICD-10 code counts; had higher surgical counts; had more consultations from physicians and other professionals during the index admission; and were more likely to have had more than 1 ED visit and more than 1 inpatient admission 6 months prior to admission compared with patients without readmissions (all P < .05). Patients with readmissions vs those without readmissions were more likely to be 85 years and older (55.1% vs 51.5%), covered by Medicare (86.3% vs 82.6%), male (46.4% vs 42.0%), White (74.1% vs 71.7%), and married (37.0% vs 34.4%); were less likely to speak English (67.9% vs 71.6%); were more likely to have had nonelective admissions (96.5% vs 91.6%) through the medical admission type (94.9% vs 89.8%); were less likely to have been admitted from home (77.1% vs 79.5%); were more likely to have had more than 1 previous inpatient admission (51.2% vs 36.0%) and ED visit (55.6% vs 41.9%); were more likely to have an Elixhauser Comorbidity Index score greater than 15 (49.5% vs 34.2%); and were more likely to be discharged to locations other than home (56.5% vs 51.1%). Other differences for those with readmissions included greater likelihood of length of stay greater than 5 days (56.9% vs 43.8%), having more than 1 surgical procedure (78.5% vs 73.5%), having more than 3 physician consultations (49.4% vs 38.5%), and having seen more than 11 physicians (56.5% vs 44.2%) and more than 11 nonphysicians (54.7% vs 42.3%).

Risk Factors Associated With 30-Day Hospital Readmissions

Table 2 shows ORs of readmission risks based on patient demographic, medical, and encounter characteristics at baseline. Patients had higher odds of readmissions if they were male, were a non-English speaker, were Medicare insured, had more than 15 Elixhauser comorbidities, stayed in the hospital for more than 5 days, had more than 1 surgical procedure, had more than 3 physician consultations in the prior 6 months, had 3 or more inpatient admissions in the 6 months prior to index admission, and were discharged to other settings compared with each respective counterpart.

DISCUSSION

This study aimed to identify EHR-derived data to predict readmission risks for individuals with ADRD. In our cohort, we found a 19% readmission rate for patients with ADRD, which is within the 7% to 35% range reported in previous studies.22-24 Previous medical encounters, including more than 5 days in the hospital and more than 3 ED visits within the 6 months prior to index admission, were significant risks for readmission.

Our study found that certain nonclinical and clinical patient characteristics were associated with higher risk of readmission in the ADRD population. Consistent with findings from other studies in general and in patients with cognitive impairments or dementia, more comorbidities, chronic conditions, and inpatient admissions in the preceding 6 months from the index admission were associated with higher readmissions.24-26 The literature has shown mixed findings about sex-associated readmission risk. Globally, findings from some older studies in the US and Europe suggest that women had higher readmission risks,27,28 but this is not consistent; a recent study found that female sex and living in less deprived areas reduced the odds of readmission.29 Our study found that men with ADRD had higher readmissions. Historically, sex has been considered a binary construction (male or female) and defined by differences in chromosomes, sex organs, endogenous hormones, and other characteristics encoded by DNA.30 More recently, scientific advances in identification of biomarkers challenge the belief that sex is only binary.31 Understanding sex and gender in the population with ADRD has opportunities to improve health outcomes and health care for individuals with ADRD. To fully understand the influences of sex and gender in ADRD, further development of standardized concepts and measures for characterizing sex and gender is warranted. Other considerations include understanding that older adults may experience increased risk of frailty and vulnerability, ultimately leading to unfavorable health consequences.32 Further, poor care transitions increase the risk of readmissions in people of advanced age with dementia.33 Early detection of these factors provides an opportunity for risk mitigation to reduce unnecessary costs and the burdens of avoidable readmissions. Additionally, acute care hospital providers can partner with community-based organizations and services to improve quality of care and optimize outcomes for this vulnerable population.

This study adds to the growing literature that non-English speakers are at higher risk of hospital readmissions. Language barriers may delay treatment or affect the overall quality of care. Adults with nondominant language preference have more hospital readmissions and ED visits. Further, studies from English-speaking countries that cover populations across the life span have estimated that individuals who speak other languages have a 15% to 25% risk of readmission, even with the use of interpreter services.5,34,35 Our study found that the impact of language may potentially override the impact of race/ethnicity on health outcomes for the population with ADRD. Consistent with current literature findings, we also found that patients covered by a non-Medicare payer had a relatively lower readmission risk, indicating patients with private insurance might have better access to resources and care they need, which influences their ability to manage their health post discharge.36

Although we found equal and lower but not statistically significant odds of readmission risk for patients discharged to IRFs and SNFs vs home, we also found that patients discharged to other settings (such as critical access hospitals, federal health care facilities, acute hospitals, inpatient psychiatric hospitals, and long-term care) experienced higher readmission risks compared with those discharged to home. This may be due to other settings having variations in the level of support or disadvantages in quality of care and patient outcomes.

Although 30-day readmission rates are improving nationally,37 research on chronic and progressive conditions continues to show the need to prioritize preventing readmissions and reducing health care costs and care burdens for patients and caregivers. Utilizing the EHR to link care episodes allows better representation of real-life care scenarios. We suggest that future research include data from hospitals, postacute care, and the community to provide unique perspectives identifying factors impeding the continuity of care for patients, caregivers, health care providers, and administrators. This study serves as a foundation for using the EHR to identify a broader spectrum of risks associated with ADRD for hospitalized patients. Despite significant advances in the field of EHRs, challenges in cross-setting implementation38,39 and adoption40 of data collection processes persist. To prevent unnecessary readmissions, it is crucial to implement strategies that enhance linking care process data from multiple providers, providing patient and caregiver education, coordinating care with community programs and services, and addressing social determinants of health.

Limitations

This study has limited generalizability due to the inclusion of only 1 urban academic medical center; it did not include rehospitalization outside the system. Although the EHR presents an extensive array of variables, it remains uncertain whether the model encompasses all crucial variables. Using EHR data retrospectively is limited to the existing information systematically collected and coded in the health care system. EHR data may not capture the full range of patients with dementia. Some individuals who are considered to have dementia may have started off-label on FDA-approved dementia medication for other indications (eg, memantine for migraine or an acetylcholinesterase inhibitor for mild cognitive impairment), so our approach may have the potential to overestimate the dementia sample. Continued research is needed to better identify patients with dementia to ensure that inferences made reflect the true patient population. Further, patients’ level of education and other variables indicating cognitive impairment and socioeconomic status, such as family support and household income as a measure of wealth, were not available, and these specific variables may impact risk of readmissions. Additionally, the length of the previous admission and cause of admission were not available in the database and could affect the rate of return to the hospital. Although we controlled for the admission source and discharge disposition in the outcome model, we recognize that our model may not adequately capture cognitive impairment for the study population. Further, the proportion of patients who opted out of research is unknown.

CONCLUSIONS

The study results suggest that clinicians may use certain demographic, medical, and encounter characteristics as proxies or indicators to flag patients at high risk of 30-day readmissions when patients with ADRD are admitted to an acute care hospital. Risk factors for readmissions included but were not limited to being male, being a non-English speaker, having significant comorbidities, staying in the hospital for more than 5 days, having more than 3 inpatient admissions in the prior 6 months, having more than 3 physician consultations and more than 1 surgical procedure in the prior 6 months, and being discharged to settings other than home, IRFs, or SNFs. Although the adoption of EHRs has improved documentation and recordkeeping for clinicians, further work is needed to focus on the identification of complex disease risks and to understand and link data in the EHR from the demographic, medical, and encounter perspectives throughout the continuum of care to enhance patient care effectively.

Author Affiliations: Department of Physical Medicine and Rehabilitation (PSR, MN), Department of Medical Affairs (PSR), Department of Neurology (NQ, ES, NLS, ZST), and Department of Medicine (NQ), Cedars-Sinai, Los Angeles, CA; Department of Occupational Therapy, School of Health Professions, and Department of Population Health and Health Disparities, School of Public and Population Health, The University of Texas Medical Branch (CYL), Galveston, TX; Casa Colina Hospital and Centers for Healthcare (DSO), Pomona, CA.

Source of Funding: National Institutes of Health (NIH) National Center for Advancing Translational Science UCLA Clinical and Translational Science Institute grant No. UL1TR001881; NIH K01HD101589.

Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (PSR, CYL, ZST); acquisition of data (PSR); analysis and interpretation of data (PSR, CYL, DSO, ZST); drafting of the manuscript (PSR, CYL, DSO, ES, NQ, MN, NLS, ZST); critical revision of the manuscript for important intellectual content (CYL, DSO, NQ, MN, NLS, ZST); statistical analysis (PSR, CYL); provision of patients or study materials (ES); obtaining funding (PSR); administrative, technical, or logistic support (ES, ZST, NQ, MN, NLS); and supervision (PSR, ZST).

Address Correspondence to: Pamela S. Roberts, PhD, OTR/L, Cedars-Sinai, 8700 Beverly Blvd, Ste 2416, Los Angeles, CA 90048. Email: pamela.roberts@cshs.org.

REFERENCES

1. Conway SJ, Kuye IO, Yeatts J, Jaffery J, Berkowitz SA. Transforming health care from volume to value: moving the needle through population health. Am J Med. 2023;136(9):874-877. doi:10.1016/j.amjmed.2023.04.031

2. 2023 Alzheimer’s disease facts and figures. Alzheimers Dement. 2023;19(4):1598-1695. doi:10.1002/alz.13016

3. LaMantia MA, Stump TE, Messina FC, Miller DK, Callahan CM. Emergency department use among older adults with dementia. Alzheimer Dis Assoc Disord. 2016;30(1):35-40. doi:10.1097/WAD.0000000000000118

4. Shepherd H, Livingston G, Chan J, Sommerlad A. Hospitalization rates and predictors in people with dementia: a systematic review and meta-analysis. BMC Med. 2019;17(1):130. doi:10.1186/s12916-019-1369-7

5. Biswas S, Dinh D, Lucas M, et al. Impact of limited English proficiency on presentation and clinical outcomes of patients undergoing primary percutaneous coronary intervention. Eur Heart J Qual Care Clin Outcomes. 2020;6(4):254-262. doi:10.1093/ehjqcco/qcz061

6. Ibrahim AM, Dimick JB, Sinha SS, Hollingsworth JM, Nuliyalu U, Ryan AM. Association of coded severity with readmission reduction after the Hospital Readmissions Reduction Program. JAMA Intern Med. 2018;178(2):290-292. doi:10.1001/jamainternmed.2017.6148

7. Karliner LS, Pérez-Stable EJ, Gregorich SE. Convenient access to professional interpreters in the hospital decreases readmission rates and estimated hospital expenditures for patient with limited English proficiency. Med Care. 2017;55(3):199-206. doi:10.1097/MLR.0000000000000643

8. Anderson TS, Marcantonio ER, McCarthy EP, Herzig SJ. National trends in potentially preventable hospitalizations of older adults with dementia. J Am Geriatr Soc. 2020;68(10):2240-2248. doi:10.1111/jgs.16636

9. Anderson TS, Marcantonio ER, McCarthy EP, Ngo L, Schonberg MA, Herzig SJ. Association of diagnosed dementia with post-discharge mortality and readmission among hospitalized Medicare beneficiaries. J Gen Intern Med. 2022;37(16):4062-4070. doi:10.1007/s11606-022-07549-7

10. Lindeza P, Rodrigues M, Costa J, Guerreiro M, Rosa MM. Impact of dementia on informal care: a systematic review of family caregivers’ perceptions. BMJ Support Palliat Care. 2024;14:e38-e49. doi:10.1136/bmjspcare-2020-002242

11. Foster NL. Barriers to treatment: the unique challenges for physicians providing dementia care. J Geriatr Psychiatry Neurol. 2001;14(4):188-198. doi:10.1177/089198870101400404

12. Boustani M, Schubert C, Sennour Y. The challenge of supporting care for dementia in primary care. Clin Interv Aging. 2007;2(4):631-636. doi:10.2147/cia.s1802

13. Hospital Readmissions Reduction Program. CMS. Updated August 11, 2025. Accessed November 10, 2025. https://www.cms.gov/medicare/payment/prospective-payment-systems/acute-inpatient-pps/hospital-readmissions-reduction-program-hrrp

14. Advantages of electronic health records. HealthIT.gov. Updated March 8, 2022. Accessed January 30, 2024. https://web.archive.org/web/20240118084157/https://www.healthit.gov/faq/what-are-advantages-electronic-health-records

15. Taksler GB, Dalton JE, Perzynski AT, et al. Opportunities, pitfalls, and alternatives in adapting electronic health records for health services research. Med Decis Making. 2021;41(2):133-142. doi:10.1177/0272989X20954403

16. Our software. Epic. Accessed March 10, 2024. https://www.epic.com/software

17. Tan ZS, Qureshi N, Roberts P, et al. Alerting providers to hospitalized persons with dementia using the electronic health record. J Am Geriatr Soc. 2024;72(3):822-827. doi:10.1111/jgs.18673

18. Shao Y, Zeng QT, Chen KK, Shutes-David A, Thielke SM, Tsuang DW. Detection of probable dementia cases in undiagnosed patients using structured and unstructured electronic health records. BMC Med Inform Decis Mak. 2019;19(1):128. doi:10.1186/s12911-019-0846-4

19. Estiri H, Azhir A, Blacker DL, Ritchie CS, Patel CJ, Murphy SN. Temporal characterization of Alzheimer’s disease with sequences of clinical records. EBioMedicine. 2023;92:104629. doi:10.1016/j.ebiom.2023.104629

20. Harwood DM, Hope T, Jacoby R. Cognitive impairment in medical inpatients. I: screening for dementia—is history better than mental state? Age Ageing. 1997;26(1):31-35. doi:10.1093/ageing/26.1.31

21. Joray S, Wietlisbach V, Büla CJ. Cognitive impairment in elderly medical inpatients: detection and associated six-month outcomes. Am J Geriatr Psychiatry. 2004;12(6):639-647. doi:10.1176/appi.ajgp.12.6.639

22. Kandor N, Syriamaki J, Aikens J, Mahmoudi E. Readmission rates and episode costs for Alzheimer disease and related dementias across hospitals in a statewide collaborative. JAMA Netw Open. 2023;6(3):e232109. doi:10.1001/jamanetworkopen.2023.2109

23. Ma C, Bao S, Dull P, Wu B, Yu F. Hospital readmission in persons with dementia: a systematic review. Int J Geriatr Psychiatry. 2019;34(8):1170-1184. doi:10.1002/gps.5140

24. Considine J, Fox K, Plunkett D, Mecner M, O’Reilly M, Darzins P. Factors associated with unplanned readmissions in a major Australian health service. Aust Health Rev. 2019;43(1):1-9. doi:10.1071/AH16287

25. MacNeil-Vroomen JL, Nagurney JM, Allore HG. Comorbid conditions and emergency department treat and release utilization in multimorbid persons with cognitive impairment. Am J Emerg Med. 2020;38(1):127-131. doi:10.1016/j.ajem.2019.07.023

26. Golüke NMS, van de Vorst IE, Vaartjes IH, et al. Risk factors for in-hospital mortality in patients with dementia. Maturitas. 2019;129:57-61. doi:10.1016/j.maturitas.2019.08.007

27. Pérès K, Helmer C, Amieva H, et al. Gender differences in the prodromal signs of dementia: memory complaint and IADL-restriction. a prospective population-based cohort. J Alzheimers Dis. 2011;27(1):39-47. doi:10.3233/JAD-2011-110428

28. Nye J. Patients with ADRD experience higher readmission rates and episode costs. Neurology Advisor. April 4, 2023. Accessed September 1, 2024. https://www.neurologyadvisor.com/news/patients-with-adrd-experience-higher-readmission-rates-and-episode-costs

29. Zafeiridi E, McMichael A, O’Hara L, Passmore P, McGuinness B. Hospital admissions and emergency department visits for people with dementia. QJM. 2024;117(2):119-124. doi:10.1093/qjmed/hcad232

30. Including women and minorities in clinical research background. Office of Research on Women’s Health. Updated November 6, 2025. Accessed September 1, 2024. https://orwh.od.nih.gov/including-women-and-minorities-in-clinical-research-background

31. Ainsworth C. Sex redefined. Nature. 2015;518(7539):288-291. doi:10.1038/518288a

32. Bag Soytas R, Levinoff EJ, Smith L, et al. Predictive strategies to reduce the risk of rehospitalization with a focus on frail older adults: a narrative review. Epidemiologia (Basel). 2023;4(4):382-407. doi:10.3390/epidemiologia4040035

33. Browne B, Ali K, Tabet N. Determinants of hospital readmissions in older people with dementia: a scoping review. Age Ageing. 2024;53(suppl 1):afad246.124. doi:10.1093/ageing/afad246.124

34. Inagaki E, Farber A, Kalish J, et al. Role of language discordance in complication and readmission rate after infrainguinal bypass. J Vasc Surg. 2017;66(5):1473-1478. doi:10.1016/j.jvs.2017.03.453

35. Chu JN, Wong J, Bardach NS, et al. Association between language discordance and unplanned hospital readmissions or emergency department revisits: a systematic review and meta-analysis. BMJ Qual Saf. 2024;33(7):456-469. doi:10.1136/bmjqs-2023-016295

36. Basu J. Multilevel risk factors for hospital readmission among patients with opioid use disorder in selected US states: role of socioeconomic characteristics of patients and their community. Health Serv Res Manag Epidemiol. 2020;7:2333392820904240. doi:10.1177/2333392820904240

37. Jiang HJ, Hensche M. Characteristics of 30-day all-cause hospital readmissions, 2016-2020. Healthcare Cost and Utilization Project statistical brief 304. Updated September 23, 2023. Accessed March 10, 2024. https://hcup-us.ahrq.gov/reports/statbriefs/sb304-readmissions-2016-2020.jsp

38. Ratwani R, Fairbanks T, Savage E, et al. Mind the gap: a systematic review to identify usability and safety challenges and practices during electronic health record implementation. Appl Clin Inform. 2016;7(4):1069-1087. doi:10.4338/ACI-2016-06-R-0105

39. Tsai CH, Eghdam A, Davoody N, Wright G, Flowerday S, Koch S. Effects of electronic health record implementation and barriers to adoption and use: a scoping review and qualitative analysis of the content. Life (Basel). 2020;10(12):327. doi:10.3390/life10120327

40. Kruse CS, Kristof C, Jones B, Mitchell E, Martinez A. Barriers to electronic health record adoption: a systematic literature review. J Med Syst. 2016;40(12):252. doi:10.1007/s10916-016-0628-9