Diagnosis of behavioral symptoms as a predictor of institutionalization among Medicaid patients with dementia.

OBJECTIVES: Behavioral symptoms are commonly observed in the course of dementia. This study aimed to assess the association of the diagnosis of a cluster of behavioral symptoms (e.g., agitation, aggression, psychotic symptoms, and delirium/wandering) with the likelihood of subsequent institutionalization. METHODS: A retrospective cohort study of adults aged 65 and above diagnosed with dementia identified in the IBM® MarketScan® Multistate Medicaid database between October 01, 2015, and September 30, 2019, was conducted. The index date was defined as the first diagnosis date of dementia. The presence or absence of behavioral symptoms was identified in the 6 months prior to the index date (baseline). Institutionalization was evaluated 12 months (follow-up) post the index date. The association between diagnosed behavioral symptoms during the baseline period and institutionalization in the follow-up period was assessed using a multivariable logistic regression, adjusting for baseline sociodemographic and clinical characteristics. RESULTS: The study cohort included 40,714 patients with dementia. A diagnosis of behavioral symptoms was found among 2,067 (5.1%) patients during the baseline period. An increased likelihood of institutionalization was found during the follow-up among patients with agitation and aggression in baseline (OR = 1.51 (95% CI: 1.18-1.92)) compared to patients without these symptoms at baseline. Patients with psychotic symptoms in baseline had significantly higher odds of getting institutionalized during the follow-up compared to patients without psychotic symptoms in baseline (OR = 1.36 (95% CI: 1.20-1.54)). Similarly, patients with symptoms of delirium and wandering in baseline had a higher likelihood of institutionalization than patients without these symptoms at baseline (OR = 1.61 (95% CI: 1.30-1.99)). CONCLUSION: Several diagnosed behavioral symptoms were associated with a higher risk of institutionalization among older adults with dementia and should be considered when planning treatment strategies for the effective management of the condition.

Development and Temporal Validation of an Electronic Medical Record-Based Insomnia Prediction Model Using Data from a Statewide Health Information Exchange.

This study aimed to develop and temporally validate an electronic medical record (EMR)-based insomnia prediction model. In this nested case-control study, we analyzed EMR data from 2011-2018 obtained from a statewide health information exchange. The study sample included 19,843 insomnia cases and 19,843 controls matched by age, sex, and race. Models using different ML techniques were trained to predict insomnia using demographics, diagnosis, and medication order data from two surveillance periods: -1 to -365 days and -180 to -365 days before the first documentation of insomnia. Separate models were also trained with patient data from three time periods (2011-2013, 2011-2015, and 2011-2017). After selecting the best model, predictive performance was evaluated on holdout patients as well as patients from subsequent years to assess the temporal validity of the models. An extreme gradient boosting (XGBoost) model outperformed all other classifiers. XGboost models trained on 2011-2017 data from -1 to -365 and -180 to -365 days before index had AUCs of 0.80 (SD 0.005) and 0.70 (SD 0.006), respectively, on the holdout set. On patients with data from subsequent years, a drop of at most 4% in AUC is observed for all models, even when there is a five-year difference between the collection period of the training and the temporal validation data. The proposed EMR-based prediction models can be used to identify insomnia up to six months before clinical detection. These models may provide an inexpensive, scalable, and longitudinally viable method to screen for individuals at high risk of insomnia.

Economic outcomes associated with diagnosed behavioral symptoms among patients with dementia in the United States: a health care claims database analysis.

BACKGROUND: Behavioral symptoms are common in patients with dementia. However, there is limited evidence of their economic burden. Among commercially insured patients with dementia in the United States, this study assessed the prevalence of diagnosed behavioral symptoms and whether healthcare resources utilization and costs were associated with these symptoms. METHODS: This retrospective observational study was conducted using the IBM® MarketScan® Commercial Claims and Encounters and Medicare Supplemental database from October 1, 2015, to September 30, 2019. Diagnoses of dementia and behavioral symptoms were identified using the International Classification of Diseases, 10th Modification codes. To test differences in patient characteristics among those with and without diagnosed behavioral symptoms, t-tests were used for continuous variables, and chi-square tests were used for categories. Generalized linear models were used to compare healthcare resource utilization and costs between patients with and without diagnosed behavioral symptoms, adjusted for baseline characteristics. RESULTS: Of the 62,901 patients with dementia included in the analysis, 16.5% had diagnosed behavioral symptoms 12 months post dementia diagnosis. Patients with diagnosed behavioral symptoms used more health care resources (mean annual pharmacy visits per patient: 39.83 vs. 33.08, mean annual outpatient visits per patient: 24.20 vs. 16.94, mean annual inpatient visits per patient: 0.98 vs. 0.47, mean annual ER visits per patient: 2.45 vs. 1.21) and incurred higher cost of care than those without diagnosed behavioral symptoms (mean annual total health care costs per patients: $63,268 versus $33,383). Inpatient care was the most significant contributor to total costs (adjusted annual mean cost per patient: $28,195 versus $12,275). CONCLUSION: Behavioral symptoms were significantly associated with higher healthcare resource utilization and costs among patients with dementia. Further research is warranted to address the unmet medical needs of this patient population.

Modeling acute care utilization: practical implications for insomnia patients.

Machine learning models can help improve health care services. However, they need to be practical to gain wide-adoption. In this study, we investigate the practical utility of different data modalities and cohort segmentation strategies when designing models for emergency department (ED) and inpatient hospital (IH) visits. The data modalities include socio-demographics, diagnosis and medications. Segmentation compares a cohort of insomnia patients to a cohort of general non-insomnia patients under varying age and disease severity criteria. Transfer testing between the two cohorts is introduced to demonstrate that an insomnia-specific model is not necessary when predicting future ED visits, but may have merit when predicting IH visits especially for patients with an insomnia diagnosis. The results also indicate that using both diagnosis and medications as a source of data does not generally improve model performance and may increase its overhead. Based on these findings, the proposed evaluation methodologies are recommended to ascertain the utility of disease-specific models in addition to the traditional intra-cohort testing.

Medications in Patients with Dementia and Behavioral Disturbance.

BACKGROUND: Behavioral disturbance (BD) is common in dementia patients, with no FDA approved medications for this condition. Little data exists on the real-world medication use in this population. OBJECTIVE: To describe real-world medications use in this population. METHODS: A cross-sectional study was conducted using the MarketScan database for outpatient medications and the Cerner database for inpatient medications. The study period was Oct 2015-Jun 2018. Patients with dementia and BD were identified through ICD-10-CM. We examined outpatient medications prescribed during 6-month before or after BD event date, and inpatient medications during inpatient visits, especially on central nervous systems (CNS) drugs including antidementia drugs, antidepressants, antipsychotics, anxiolytics, and anticonvulsants. RESULTS: A total of 56,544 outpatients and 34,245 patient hospitalizations were assessed separately. Among outpatients, patients filled more medications after a BD event. The use of the five CNS drug classes generally increased after a BD event, and the largest increase was seen in antipsychotics (23%to 33%). Among inpatients, the median number of medications used in each hospitalization was 14. The use of antipsychotics was particularly high (64%), followed by anxiolytics (51%). A list of 60 unique medications were suggested to be the commonly used drugs in dementia patients with BD. CONCLUSION: In dementia patients with BD, anti-dementia medications, antidepressants, anticonvulsants, hypnotics and antipsychotics were the most used drug classes. Antidepressants and antipsychotics use were more frequent after a BD event, which suggests a need for safe drugs targeting BD in dementia patients.

Predicting dementia with routine care EMR data.

Our aim is to develop a machine learning (ML) model that can predict dementia in a general patient population from multiple health care institutions one year and three years prior to the onset of the disease without any additional monitoring or screening. The purpose of the model is to automate the cost-effective, non-invasive, digital pre-screening of patients at risk for dementia. Towards this purpose, routine care data, which is widely available through Electronic Medical Record (EMR) systems is used as a data source. These data embody a rich knowledge and make related medical applications easy to deploy at scale in a cost-effective manner. Specifically, the model is trained by using structured and unstructured data from three EMR data sets: diagnosis, prescriptions, and medical notes. Each of these three data sets is used to construct an individual model along with a combined model which is derived by using all three data sets. Human-interpretable data processing and ML techniques are selected in order to facilitate adoption of the proposed model by health care providers from multiple institutions. The results show that the combined model is generalizable across multiple institutions and is able to predict dementia within one year of its onset with an accuracy of nearly 80% despite the fact that it was trained using routine care data. Moreover, the analysis of the models identified important predictors for dementia. Some of these predictors (e.g., age and hypertensive disorders) are already confirmed by the literature while others, especially the ones derived from the unstructured medical notes, require further clinical analysis.

Passive Digital Signature for Early Identification of Alzheimer's Disease and Related Dementia.

OBJECTIVES: Developing scalable strategies for the early identification of Alzheimer's disease and related dementia (ADRD) is important. We aimed to develop a passive digital signature for early identification of ADRD using electronic medical record (EMR) data. DESIGN: A case-control study. SETTING: The Indiana Network for Patient Care (INPC), a regional health information exchange in Indiana. PARTICIPANTS: Patients identified with ADRD and matched controls. MEASUREMENTS: We used data from the INPC that includes structured and unstructured (visit notes, progress notes, medication notes) EMR data. Cases and controls were matched on age, race, and sex. The derivation sample consisted of 10 504 cases and 39 510 controls; the validation sample included 4500 cases and 16 952 controls. We constructed models to identify early 1- to 10-year, 3- to 10-year, and 5- to 10-year ADRD signatures. The analyses included 14 diagnostic risk variables and 10 drug classes in addition to new variables produced from unstructured data (eg, disorientation, confusion, wandering, apraxia, etc). The area under the receiver operating characteristics (AUROC) curve was used to determine the best models. RESULTS: The AUROC curves for the validation samples for the 1- to 10-year, 3- to 10-year, and 5- to 10-year models that used only structured data were .689, .649, and .633, respectively. For the same samples and years, models that used both structured and unstructured data produced AUROC curves of .798, .748, and .704, respectively. Using a cutoff to maximize sensitivity and specificity, the 1- to 10-year, 3- to 10-year, and 5- to 10-year models had sensitivity that ranged from 51% to 62% and specificity that ranged from 80% to 89%. CONCLUSION: EMR-based data provide a targeted and scalable process for early identification of risk of ADRD as an alternative to traditional population screening. J Am Geriatr Soc 68:511-518, 2020.

Physician Practice Patterns Associated with Diagnostic Evaluation of Patients with Suspected Mild Cognitive Impairment and Alzheimer's Disease.

The diagnostic process for patients presenting with cognitive decline and suspected dementia is complex. Physicians face challenges distinguishing between normal aging, mild cognitive impairment, Alzheimer's disease, and other dementias. Although there is some evidence for improving attitudes towards the importance of prompt diagnosis, there is limited information describing how physicians approach this diagnostic challenge in practice. This was explored in the present study. Across-sectional survey of primary care and specialist physicians, in 5 European countries, Canada, and the United States, was conducted. Participants were asked about their use of cognitive screening tools and diagnostic technologies, as well as the rationales and barriers for use. In total, 1365 physicians participated in the survey, 63% of whom were specialists. Most physicians stated they use objective cognitive tools to aid the early detection of suspected mild cognitive impairment or Alzheimer's disease in patients. The Mini-Mental State Examination was the most common tool used for initial screening; respondents cited speed and ease of use but noted its lack of specificity. Cerebrospinal fluid biomarker and amyloid positron emission tomography tests, respectively, had been used by only 26% and 32% of physicians in the preceding 6 months, although patterns of use varied across countries. The most commonly cited reasons for not ordering such tests were invasiveness (for cerebrospinal fluid biomarker testing) and cost (for amyloid positron emission tomography imaging). Data reported by physicians reveal differences in the approaches to the diagnostics process in Alzheimer's. A higher proportion of primary care physicians in the United States are routinely incorporating cognitive assessment tools into annual visits, but this is due to country differences in clinical practice. The value of screening tools and regular use could be discussed further with physicians; however, lack of specificity associated with cognitive tools and the investment required from patients and the healthcare system are limiting factors.