Figure 1 - uploaded by Farrokh Alemi
Content may be subject to copyright.
Transition among 25 common combinations of ADL impairments. Nodes show combination of disabilities, using following abbreviations: B = bathing; D = dressing; F = feeding; G = grooming; L = bowel continence; S = transferring; T = toileting; U = urinary continence; W = walking disabilities. Numbers show days before a resident changes in the direction of the arrow. Numbers inside the nodes show days of no change. Dashed lines show days to partial or complete recovery. Colored nodes show the most likely path to next additional disability (i.e., assuming no recovery and no remaining in the current state).
Source publication
This study provides benchmarks for likelihood, number of days until, and sequence of functional decline and recovery.
We analyzed activities of daily living (ADLs) of 296,051 residents in Veteran Affairs nursing homes between January 1, 2000 and October 9, 2012. ADLs were extracted from standard minimum data set assessments. Because of significant...
Contexts in source publication
Context 1
... each combination, we created a pathway that showed transition to another state, where there was no change, a decline, or partial/complete recovery in ADL measures. These path- ways described a network model of transitions among the 25 combinations of ADL activities (Figure 1). In this model, we show recovery with a dashed line and decline with a straight line. ...
Context 2
... the five learned networks, we chose the one with best fit to the test data (contingency table fit = 99% with 40% of the data set left out as the test data). Figure 1 shows the pattern of functional impairments observed among 25 combinations of ADL deficits. Each combination is shown as a node with letters indicating the combination of ADL impairments. ...
Context 3
... blanket statement about sequence of losing walking and grooming functions will not be accurate in one of the two situations. These data point to the complex pattern of loss and recov- ery that are depicted in Figure 1. ...
Context 4
... single path in Figure 1 describes the experience of a majority of residents. To describe how the majority of residents (57.4%) lose their function, we included the four paths of greatest likelihood (Figure 2). ...
Context 5
... 3 shows a Bayesian network for ADL loss. This figure simplifies the observed relationships in Figure 1 by ignoring recovery (no cycles are allowed in Bayesian net- works as previously noted) and by replacing combination of disabilities with the presence of any one of the nine ADL impairments. For example, B and BG in Figure 1 both indi- cate a bathing disability in Figure 3. ...
Context 6
... figure simplifies the observed relationships in Figure 1 by ignoring recovery (no cycles are allowed in Bayesian net- works as previously noted) and by replacing combination of disabilities with the presence of any one of the nine ADL impairments. For example, B and BG in Figure 1 both indi- cate a bathing disability in Figure 3. The most likely path is shown in dashed red color. ...
Context 7
... body of research suggests that loss of the ability to eat may represent a state in which the accumulation of func- tional losses is too great to benefit from some interventions, such as tube feeding (Hwang, Teno, Gozalo, & Mitchell, 2014;Teno et al., 2010). In Figure 1, loss of feeding ability occurred very late in the sequence of functional loss, fewer than one third recovered feeding ability and, among those who did recover feeding ability, the recovery times ranged from 115 to 272 days. This is an example of how data may be piloted in care planning to reflect expected likelihood and time to experience recovery of individual deficits. ...
Citations
... Finally, these analyses are consistent with work examining the sequence of functional losses and highlight the need for sophisticated analytic strategies. Levy et al. (33), using semi-Markov chains, calculated the transition probabilities associated with the addition and recovery from ADL impairments. These network analyses identified multiple pathways and increased understanding about how illness and disability unfold. ...
This study sought to prospectively examine the association between the onset of five chronic health conditions (arthritis, diabetes, heart disease, hypertension, and pulmonary disease) and depressive symptoms in a community dwelling sample of older adults. Longitudinal multi-level modeling was used to examine the effects of illness onset on depressive symptoms. We found that, from 2006 and 2011, onset of each of the 5 conditions had independent effects that increased depressive symptoms for people with zero or one chronic condition at baseline, while controlling for a lifetime diagnosis of depression, age, gender, income, and race. The transition from no or a single chronic health condition to multimorbidity regardless of the type of illness of onset impacted older adults’ depressive symptoms. Developing ways to prevent or delay the onset of additional chronic illnesses will impact the well-being of older adults.
... Finally, these analyses are consistent with work examining the sequence of functional losses and highlight the need for sophisticated analytic strategies. Levy et al. (33), using semi-Markov chains, calculated the transition probabilities associated with the addition and recovery from ADL impairments. These network analyses identified multiple pathways and increased understanding about how illness and disability unfold. ...
Background:
The U.S. Department of Health and Human Services recently called for a paradigm shift from the study of individual chronic conditions to multiple chronic conditions (MCCs). We identify the most common combinations of chronic diseases experienced by a sample of community-dwelling older people and assess whether depression is differentially associated with combinations of illnesses.
Methods:
Self-reports of diagnosed chronic conditions and depressive symptoms were provided by 5,688 people participating in the ORANJ BOWLSM research panel. Each respondent was categorized as belonging to one of 32 groups. ANOVA examined the association between depressive symptoms and combinations of illnesses.
Results:
People with more health conditions experienced higher levels of depression than people with fewer health conditions. People with some illness combinations had higher levels of depressive symptoms than people with other illness combinations.
Conclusions:
Findings confirm extensive variability in the combinations of illnesses experienced by older adults and demonstrate the complex associations of specific illness combinations with depressive symptoms. Results highlight the need to expand our conceptualization of research and treatment around MCCs and call for a person-centered approach that addresses the unique needs of individuals with MCCs.
Independently performing activities of daily living (ADLs) is vital for maintaining one’s quality of life. Losing this ability can significantly impact an individual’s overall health status, including their mental health and social well-being. Aging is an important factor contributing to the loss of ADL abilities, and our study focuses on investigating the trajectories of functional decline and recovery in older adults. Employing trajectory analytics methodologies, this research delves into the intricate dynamics of ADL pathways, unveiling their complexity, diversity, and inherent characteristics. The study leverages a substantial dataset encompassing ADL assessments of nursing home residents with diverse disability profiles in the United States. The investigation begins by transforming these assessments into sequences of disability combinations, followed by applying various statistical measures, indicators, and visual analytics. Valuable insights are gained into the typical disability states, transitions, and patterns over time. The results also indicate that while predicting the progression of ADL disabilities presents manageable challenges, the duration of these states proves more complicated. Our findings hold significant potential for improving healthcare decision-making by enabling clinicians to anticipate possible patterns, develop targeted and effective interventions that support older patients in preserving their independence, and enhance overall care quality.
Purpose of the article:
This article describes a conceptual and methodological approach to integrating functional information into an ontology to categorize mental functioning, which to date is an under-developed area of classification, and supports our work with the United States (U.S.) Social Security Administration (SSA).
Design and methodological procedures:
Conceptualizing and defining mental functioning was paramount to develop natural language processing (NLP) tools to support our use case. The International Classification of Functioning, Disability, and Health (ICF) was the framework used to conceptualize mental functioning at the activities and participation level in clinical records. To address challenges that arose when applying the ICF as to what should or should not be classified as mental functioning, a mental functioning domain ontology was developed that rearranged, reclassified and incorporated all ICF key components, concepts, classifications, and their definitions.
Conclusions:
Challenges emerged in the extent to which we could directly align components in the ICF into an applied ontology of mental functioning. These conceptual challenges required rearrangement of ICF components to adequately support our use case within the social security disability determination process. Findings also have implications to support future NLP efforts for behavioral health outcomes and policy research.
Discovering causal relationships among symptoms is a topical issue in the analysis of observational patient datasets. A Causal Bayesian Network (CBN) is a popular analytical framework for causal inference. While there are many methods and algorithms capable of learning a Bayesian network, they are reliant on the complexity and thoroughness of the algorithm and do not consider prior expertise from authoritative sources. This paper proposes a novel method of extracting prior causal knowledge contained in Authoritative Medical Ontologies (AMOs) and using this prior knowledge to orient arcs in a CBN learned from observational patient data. Since AMOs are robust biomedical ontologies containing the collective knowledge of the experts who created them, utilizing the ordering information contained within them produces improved CBNs which provide additional insight into the disease domain.
To demonstrate our method, we obtained prior causal ordering information among symptoms from three AMOs: 1) the Medical Dictionary for Regulatory Activities Terminology (MedDRA), 2) the International Classification of Diseases Version 10 Clinical Modification (ICD-10-CM), and 3) Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT). The prior ontological knowledge from these three AMOs is then used to orient arcs in a series of CBNs learned from the National Institutes of Mental Health study on Sequenced Treatment Alternatives to Relieve Depression (STAR*D) patient dataset using the Max-Min Hill-Climbing (MMHC) algorithm. Six distinct CBNs are generated using MMHC: an unmodified baseline model using only the algorithm, three CBNs oriented with ordered-variable pairs from MedDRA, ICD-10-CM, and SNOMED CT, and two more with ordered pairs from a combination of these AMOs. The resulting CBNs modified using ordered-variable pairs significantly change the structure of the network. The agreement between the Modified networks and the Baseline ranges from 50% to 90%. A modified network using ordering information from all ontologies obtained an agreement of 50% (10 out of 20 arcs exist in both the Baseline and Modified models) while maintaining comparable predictive accuracy. This indicates that the Modified CBN reflects the causal claims in the AMOs and agrees with both the AMOs and the observational STAR*D dataset. Furthermore, the Modified models discovered new potentially causal relationships among symptoms in the model, while eliminating weaker edges in a qualitative analysis of the significance of these relationships in existing epidemiological research.
Background:
Medically tailored meals (MTM) may be beneficial to patients after hospital discharge.
Objective:
To determine if 2 versus 4 weeks of MTM posthospitalization will improve patient outcomes.
Design:
Randomized unblinded trial.
Settings and participants:
Six hundred and fifty patients pending hospital discharge with at least one chronic condition.
Intervention:
One MTM a day for 2 versus 4 weeks.
Main outcome and measures:
The primary outcome was a change from baseline to 60 days in the Hospital Anxiety Depression Scale (HADS). Secondary outcomes measured change in the Katz activities of daily living (ADLs), DETERMINE nutritional risk, and all-cause emergency department (ED) visits and rehospitalizations.
Results:
From baseline to 60 days the HADS anxiety subscale changed 5.4-4.9 in the 2-week group (p = .03) and 5.4-5.3 in the 4-week group (p = .49); the difference in change between groups 0.4 (p = .25). HADS changed 5.4-4.8 in the 2-week group (p = .005) and 5.3-5.1 in the 4-week group (p = .34); the difference in change between groups 0.4 (p = .18). ADL score changed from 5.3 to 5.6 in the 2-week group (p ≤ .0001) and 5.2-5.5 in the 4-week group (p ≤ .0001); the difference in change between groups -0.01 (p = .90). The DETERMINE changed in the 2-week group from 7.2 to 6.4 (p = .0006) and from 7 to 6.7 in the 4-week group (p = .19); the difference in change between groups 0.5 (p = .13). There was no difference in ED visits and rehospitalizations between groups or time to rehospitalization.
Conclusions:
Different durations of short-term MTM did not affect patient-centered or utilization outcomes.
Objective:
The Veterans Health Administration (VHA) Medical Foster Home (MFH) program was created to give veterans a community-based alternative to institutional long-term care (LTC). This study describes demographic, clinical, and functional characteristics of veterans in MFHs.
Methods:
Findings from in-home assessments of veterans in MFHs tied to 4 VHA medical centers for ≥ 90 days between April 2014 and December 2015 were collected. Trained nurses completed Minimum Data Set (MDS) 3.0 assessments for 92 veterans in MFHs. The assessment included demographic characteristics, cognition, behaviors, depression, pain, functional status, mobility, and morbidity.
Results:
MFH veterans were primarily male (85%), aged > 65 years (83%), cognitively impaired (55%), and had a diagnosis of depression (52%). Overall, 22% had caregiverreported aggressive behaviors and 45% self-reported pain. More than half used a wheelchair (56%). Of the 11 activities of daily living (ADLs) assessed, MFH residents were most likely to require assistance with bathing and least likely to require assistance with bed mobility and eating, although more than half required eating assistance.
Conclusions:
Veterans residing in MFHs have a wide range of care needs, including some veterans with high needs for help with ADLs and others who are completely independent in performing ADLs. These results provide insights about which veterans are staying in MFH care. Future studies should explore how VHA care providers refer veterans to LTC settings.
Background
Assessment of functional ability, including activities of daily living (ADLs), is a manual process completed by skilled health professionals. In the presented research, an automated decision support tool, the Computational Barthel Index Tool (CBIT), was constructed that can automatically assess and predict probabilities of current and future ADLs based on patients’ medical history.
Methods
The data used to construct the tool include the demographic information, inpatient and outpatient diagnosis codes, and reported disabilities of 181,213 residents of the Department of Veterans Affairs’ (VA) Community Living Centers. Supervised machine learning methods were applied to construct the CBIT. Temporal information about times from the first and the most recent occurrence of diagnoses was encoded. Ten-fold cross-validation was used to tune hyperparameters, and independent test sets were used to evaluate models using AUC, accuracy, recall and precision. Random forest achieved the best model quality. Models were calibrated using isotonic regression.
Results
The unabridged version of CBIT uses 578 patient characteristics and achieved average AUC of 0.94 (0.93–0.95), accuracy of 0.90 (0.89–0.91), precision of 0.91 (0.89–0.92), and recall of 0.90 (0.84–0.95) when re-evaluating patients. CBIT is also capable of predicting ADLs up to one year ahead, with accuracy decreasing over time, giving average AUC of 0.77 (0.73–0.79), accuracy of 0.73 (0.69–0.80), precision of 0.74 (0.66–0.81), and recall of 0.69 (0.34–0.96). A simplified version of CBIT with 50 top patient characteristics reached performance that does not significantly differ from full CBIT.
Conclusion
Discharge planners, disability application reviewers and clinicians evaluating comparative effectiveness of treatments can use CBIT to assess and predict information on functional status of patients.
