Dementia is underdiagnosed in both the general population and among Veterans. This underdiagnosis decreases quality of life, reduces opportunities for interventions, and increases health-care costs. New approaches are therefore necessary to facilitate the timely detection of dementia. This study seeks to identify cases of undiagnosed dementia by developing and validating a weakly supervised machine-learning approach that incorporates the analysis of both structured and unstructured electronic health record (EHR) data. A topic modeling approach that included latent Dirichlet allocation, stable topic extraction, and random sampling was applied to VHA EHRs. Topic features from unstructured data and features from structured data were compared between Veterans with (n = 1861) and without (n = 9305) ICD-9 dementia codes. A logistic regression model was used to develop dementia prediction scores, and manual reviews were conducted to validate the machine-learning results. A total of 853 features were identified (290 topics, 174 non-dementia ICD codes, 159 CPT codes, 59 medications, and 171 note types) for the development of logistic regression prediction scores. These scores were validated in a subset of Veterans without ICD-9 dementia codes (n = 120) by experts in dementia who performed manual record reviews and achieved a high level of inter-rater agreement. The manual reviews were used to develop a receiver of characteristic (ROC) curve with different thresholds for case detection, including a threshold of 0.061, which produced an optimal sensitivity (0.825) and specificity (0.832). Dementia is underdiagnosed, and thus, ICD codes alone cannot serve as a gold standard for diagnosis. However, this study suggests that imperfect data (e.g., ICD codes in combination with other EHR features) can serve as a silver standard to develop a risk model, apply that model to patients without dementia codes, and then select a case-detection threshold. The study is one of the first to utilize both structured and unstructured EHRs to develop risk scores for the diagnosis of dementia.

The number of older adults on parole and probation is growing at an unprecedented rate, yet little is known about the mental health needs and treatment utilization patterns among this group. The objective of this study is to compare the prevalence of serious or moderate mental illness (SMMI), and the proportion of those with SMMI who receive mental health treatment, among community-dwelling older adults on correctional supervision (parole or probation) vs. not on correctional supervision. Design: Cross-sectional analysis of data from the 2008–2014 National Surveys for Drug Use and Health (NSDUH). Setting: Population-based national survey data. Participants: Older adults (age ≥ 50) who participated in the NSDUH between 2008 and 2014 (n = 44,624). Participants were categorized according to whether they were on parole or probation during the 12 months prior to survey completion (n = 379) vs. not (n = 44,245). Measurements: Probable SMMI was defined using a validated measure in the NSDUH. Mental health treatment included any outpatient mental health services or prescriptions over the 12 months prior to survey completion. We compared the prevalence of SMMI, and the proportion of those with SMMI who received any treatment, by correctional status. Overall, 7% (N = 3266) of participants had SMMI; the prevalence was disproportionately higher among those on parole or probation (21% vs. 7%, p < 0.001). Sixty-two percent of those with SMMI received any mental health treatment, including 81% of those on parole or probation and 61% of those who were not (p < 0.001). This result remained statistically significant after logistic regression accounted for differences in sociodemographics and health. SMMI is disproportionally prevalent among older adults on parole or probation, and community correctional supervision programs may be facilitating linkages to needed community-based mental health treatment.

In observational studies concerning drug use and misuse, persons misusing drugs may be less likely to respond to surveys. However, little is known about differences in drug use and drug misuse risk factors between survey respondents and nonrespondents. Using electronic health record (EHR) data, we compared respondents and non-respondents in a telephone survey of middle-aged and older chronic opioid therapy patients to assess predictors of interview nonresponse. We compared general patient characteristics, specific opioid misuse risk factors, and patterns of opioid use associated with increased risk of opioid misuse. Inverse probability weights were calculated to account for nonresponse bias by EHR-measured covariates. EHR-measured covariate distributions for the full sample (nonrespondents and respondents), the unweighted respondent sample, and the inverse probability weighted respondent sample are reported. We present weighted and unweighted prevalence of self-reported opioid misuse risk factors. Among 2489 potentially eligible patients, 1477 (59.3%) completed interviews. Response rates differed with age (45–54 years, 51.8%; 55–64 years, 58.7%; 65–74 years, 67.9%; and 75 years or older, 59.9%). Tobacco users had lower response rates than did nonusers (53.5% versus 60.9%). Charlson comorbidity score was also related to response rates. Individuals with a Charlson score of 2 had the highest response rate at 65.6%; response rates were lower amoung patients with the lowest (the patients with the fewest health conditions had response rates of 56.7–60.0%) and the highest Charlson scores (patients with the most health conditions had response rates of 52.2–56.0%). These bivariate relationships persisted in adjusted multivariable logistic regression models predicting survey response. Response rates of persons with and without specific opioid misuse risk factors were similar (e.g., 58.7% for persons with substance abuse diagnoses, 59.4% for those without). Opioid use patterns associated with opioid misuse did not predict response rates (e.g., 60.6% versus 59.2% for those receiving versus not receiving opioids from 3 or more physicians outside their primary care clinic). Very few patient characteristics predicted non-response; thus, inverse probability weights accounting for nonresponse had little impact on the distributions of EHR-measured covariates or self-reported measures related to opioid use and misuse. Response rates differed by characteristics that predict nonresponse in general health surveys (age, tobacco use), but did not appear to differ by specific patient or drug use risk factors for prescription opioid misuse among middle- and older-aged chronic opioid therapy patients. When observational studies are conducted in health plan populations, electronic health records may be used to evaluate nonresponse bias and to adjust for variables predicting interview nonresponse, complementing other research uses of EHR data in observational studies.

Background: Past studies identified a link between weight loss and dementia, but lacked consistent conclusions. We sought to establish this link by examining the weight change profiles before and after dementia diagnosis. Methods: Using data from the Health and Retirement Study (1996–2020), we examined 13,123 participants. We conducted a nested case–control analysis to assess differences in biennial weight change profile while controlling for BMI, longevity polygenic risk scores, and APOE gene variants. Results: Participants with a memory disorder lost weight (−0.63%) biennially, whereas those without a diagnosis did not (+0.013%, p-value < 0.0001). Our case–control study shows a significant difference (p-value < 0.01) in pre-dementia % weight changes between the cases (−0.29%) and controls (0.19%), but not in post-dementia weight changes. The weight loss group have the highest risk (OR = 2.01; p-value < 0.0001) of developing a memory disorder compared to the stable weight and weight gain groups. The observations hold true after adjusting for BMI, longevity polygenic risk scores, and APOE variant in a multivariable model. Conclusions: We observe that weight loss in dementia is a physiological process independent of genetic factors associated with BMI and longevity. Pre-dementia weight loss may be an important prognostic criterion to assess a person’s risk of developing a memory disorder.

Palliative care discussions offer a unique opportunity for helping patients choose end-of-life (EOL) treatments. These are among the most difficult decisions in later life, and protecting patients’ ability to make these choices is one of healthcare’s strongest ethical mandates. Yet, traditional approaches to advance care planning (ACP) have only been moderately successful in helping patients make decisions that lead to treatments concordant with their values. In particular, neglect of attention to the emotions that occur during consideration of the end of one’s life contributes to patients’ difficulty with engaging in the process and following through on decisions. To improve ACP outcomes, providers can address the patient’s emotional experiences, and can use motivational interviewing as a way attend to elicit them and incorporate them into care planning. Applying personalizing emotion-attuned protocols like Conditional Medical Orders (CMO) also promotes this end.