Chamberlain, Alanna M.

Background: Multimorbidity is common in atrial fibrillation (AF); however, the impact of the number and type of comorbid conditions on death and hospitalizations is uncertain. Methods: The prevalence of 17 chronic conditions was obtained in 1430 patients with incident AF from 2000-2010 from Olmsted County, MN. Chronic conditions were ascertained electronically prior to AF and were classified into 3 groups: cardiovascular (CV), other physical, and mental. All hospitalizations and deaths were obtained through 12/31/2014. Associations of the number and type of condition and each individual condition with hospitalizations and death were determined using Andersen-Gill and Cox regression, respectively. Results: Among the 1430 patients with AF (median age 76 years, 48.6% men), the mean (SD) number of chronic conditions was 4.7 (2.7). A total of 4496 hospitalizations and 812 deaths occurred over a mean (SD) follow-up of 5.4 (4.1) years. An increase in the number of conditions of each type was associated with increased risks of death and hospitalizations (Table). Mental conditions (which included depression, anxiety, dementia, schizophrenia, and substance abuse) exhibited the strongest association with death, whereas other physical conditions exhibited the strongest association with hospitalizations. Individual predictors of death included heart failure (HF), hypertension, chronic obstructive pulmonary disease (COPD), chronic kidney disease (CKD), depression, dementia, and substance abuse. Predictors of hospitalizations included HF, coronary artery disease, diabetes, COPD, CKD, arthritis, osteoporosis, and substance abuse. Conclusions: Patients with AF have many comorbid conditions which are associated with death and hospitalizations. Although conditions of each type were associated with outcomes, mental conditions exhibited strongest association with death. These data highlight the importance of management of not only CV comorbidities but other physical and mental comorbidities in patients with AF.

Background: Accurate identification of stroke cases from electronic health record (EHR) is needed for efficient and valid clinical and epidemiology research. A number of studies have evaluated the validity of using ICD-9 and ICD-10 codes for stroke identification. Previous research results show that sensitivity and positive predictive value (PPV) for stroke identification is the lowest among all other cerebrovascular diseases, with notable differences observed by stroke pathological subtypes. While most prevalent cerebrovascular disease cases can be detected using 430-438/I60-I69 collectively, more accurate and comprehensive stroke phenotyping algorithm is needed to identify incident stroke cases from EHR. In our work, we compared case identification results using ICD-codes exclusively with ICD plus a natural language processing (NLP) algorithm. Methods: We developed the NLP part of our stroke algorithm using a list of expert provided stroke-related keywords, which covers transient ischemic attack (TIA), ischemic stroke, and hemorrhagic stroke. We validated the hybrid (ICD+NLP) algorithm and compared it with ICD-exclusive algorithm in a previously established atrial fibrillation cohort (n=5,062). Clinical notes and ICD codes of all patients after the incident AF event were reviewed by two nurse abstracters to confirm new stroke incidences. All past clinical notes and ICD codes of a subset of patients (n=402) were reviewed to confirm lifetime (prior and current) stroke cases. Manual abstraction results were considered the gold standard for evaluation of ICD only and ICD+NLP automatic extraction. Sensitivity and positive predictive values (PPV) of both algorithms were calculated. Results: Among 5,062 patients, 593 patients were confirmed to have suffered a stroke after atrial fibrillation while 4,469 patients were confirmed with no EHR evidence of stroke after AF. The ICD-exclusive algorithm had a sensitivity of 47.2% and a PPV of 47.8% for detecting new stroke incidences. The hybrid stroke algorithm achieved a sensitivity of 92.4% and a PPV of 60.6% for extraction of incident stroke. For extraction of lifetime stroke cases, the hybrid approach achieved a sensitivity of 93.9% and a PPV of 88.7%. Performance of new stroke incidence extraction is limited because past/current stroke mentions in clinical notes are difficult to distinguish by our NLP algorithm (may be a future direction). Conclusions: We developed and validated a stroke algorithm that performed well for identifying incident and lifetime stroke cases. The addition of NLP into the stroke algorithm improved the sensitivity and PPV of compared to an ICD-exclusive algorithm.

Background and Purpose— Despite modest predictive ability for ischemic stroke (IS), the CHA 2 DS 2 -VASc score is widely used for stroke prediction in atrial fibrillation. Among patients with atrial fibrillation, we aimed to (1) compare the IS or transient ischemic attack (TIA) incidence by CHA 2 DS 2 -VASc in blacks and Hispanics versus whites; (2) compare predictive ability of CHA 2 DS 2 -VASc score for IS or TIA in blacks and Hispanics versus whites; and (3) determine improvement in predictive ability of CHA 2 DS 2 -VASc score from addition of race/ethnicity. Methods— Using data from Optum Clinformatics, a large administrative claims database, we analyzed patients with atrial fibrillation enrolled in commercial and Medicare Advantage health plans from 2009 to 2015. We computed IS or TIA incidence rates, improvement in C statistic, continuous and categorical net reclassification improvement, and relative integrated discrimination improvement from addition of race/ethnicity to CHA 2 DS 2 -VASc. Results— A total of 267 419 patients (mean age, 73.1 [SD, 12.3] years; 46.6% women; 84.2% white, 8.5% black, 7.3% Hispanic) were studied. After a mean follow-up of 22 months, there were 6202 IS or TIA events. IS or TIA incidence rates were higher in blacks than Hispanics or whites (1.65, 1.40, and 1.22 cases per 100 person-years, respectively) and increased with higher CHA 2 DS 2 -VASc, with no race/ethnicity-based differences ( P for interaction=0.17). The CHA 2 DS 2 -VASc and CHA 2 DS 2 -VASc+race/ethnicity C statistic (95% CI) were 0.679 (0.670–0.686) and 0.679 (0.671–0.688). The CHA 2 DS 2 -VASc C statistic in the 3 groups were comparable. With addition of race/ethnicity, the categorical net reclassification improvement, continuous net reclassification improvement, and relative integrated discrimination improvement were −0.045 (95% CI, −0.067 to −0.025), 0.045 (95% CI, 0.025–0.068), and 0.016 (95% CI, 0.014–0.018). Conclusions— The predictive ability of CHA 2 DS 2 -VASc for IS or TIA in atrial fibrillation is comparable among whites, blacks, and Hispanics; hence, it can be used in the latter 2 groups. Addition of race/ethnicity to the CHA 2 DS 2 -VASc does not improve its predictive ability.

Background Oral anticoagulants ( OACs ) in patients with atrial fibrillation ( AF ), in addition to reducing stroke risk, could also prevent adverse cognitive outcomes. The purpose of this study was to compare the risk of dementia incidence across patients with AF initiating different OAC s. Methods and Results We identified patients with nonvalvular AF initiating OAC s in 2 US healthcare claim databases, MarketScan (2007–2015) and Optum Clinformatics (2009–2015). Dementia, comorbidities, and use of medications were defined on the basis of inpatient and outpatient claims. We performed head‐to‐head comparisons of warfarin, dabigatran, rivaroxaban, and apixaban in propensity score–matched cohorts. We calculated hazard ratios ( HR s) and 95% confidence intervals ( CI s) of incident dementia for each propensity score–matched cohort and meta‐analyzed database‐specific results. We analyzed 307 099 patients with AF from the MarketScan database and 161 346 from the Optum database, of which 6572 and 4391, respectively, had a diagnosis of incident dementia. The mean follow‐up of each cohort ranged between 0.7 and 2.2 years. Patients initiating direct OACs experienced lower rates of dementia than those initiating warfarin (dabigatran: HR , 0.85; 95% CI , 0.71–1.01; rivaroxaban: HR , 0.85; 95% CI , 0.76–0.94; apixaban: HR , 0.80; 95% CI , 0.65–0.97). There were no differences in rates of dementia comparing direct OAC user groups (dabigatran versus rivaroxaban: HR , 1.02; 95% CI , 0.79–1.32; dabigatran versus apixaban: HR , 0.92; 95% CI , 0.63–1.36; apixaban versus rivaroxaban: HR , 1.01; 95% CI , 0.86–1.19). Conclusions Patients with AF initiating direct OACs experienced lower rates of incident dementia than warfarin users. No obvious benefit was observed for any particular direct OAC in relation to dementia rates.

Background Aortic syndromes (ASs), including aortic dissection, intramural hematoma, and penetrating aortic ulcer, carry significant acute and long-term morbidity and mortality. However, the contemporary incidence and outcomes of AS are unknown. Methods and Results We used the Rochester Epidemiology Project record linkage system to identify all Olmsted County, MN, residents with AS (1995–2015). Diagnostic imaging, medical records, and death certificates were reviewed to confirm the diagnosis and AS subtype. Age- and sex-adjusted incidence rates were estimated using annual county-level census data. Survival for patients with AS was compared with age- and sex-matched controls using Cox regression to adjust for comorbid conditions. We identified 133 patients with AS (77, aortic dissection; 21, intramural hematoma; and 35, penetrating aortic ulcer). Average age was 71.8 years (SD=14.1), and 57% were men. The age- and sex-adjusted incidence was 7.7 per 100 000 person-years, was higher for men than women (10.2 versus 5.7 per 100 000 person-years), and increased with age. Among subtypes, the incidence of aortic dissection was highest (4.4 per 100 000 person-years), whereas the incidence of penetrating aortic ulcer and intramural hematoma was lower (2.1 and 1.2 per 100 000 person-years). Overall, the incidence of AS was stable over time ( P trend=0.33), although the incidence of penetrating aortic ulcer seemed to increase from 0.6 to 2.6 per 100 000 person-years ( P =0.008) with variability over the study interval. Patients with AS had more than twice the mortality rate at 5, 10, and 20 years when compared with population-based controls (5-, 10-, and 20-year mortality 39%, 57%, and 91% versus 18%, 41%, and 66%; overall adjusted mortality hazards ratio=2.1; P <0.001). Survival was lower than expected up to 90 days after AS diagnosis and did not differ significantly by subtype or by 5-year strata of diagnosis. Conclusions Overall, the incidence of aortic dissection and intramural hematoma has remained stable since 1995, despite the decline noted for other cardiovascular disease. AS confers increased early and long-term mortality that has not changed. These data highlight the need to improve long-term care to impact the prognosis of this patient group.

Background As the population with cardiovascular disease ages, geriatric conditions are of increasing relevance. A possible geriatric prognostic indicator may be a fall risk score, which is mandated by The Joint Commission to be measured on all hospitalized patients. The prognostic value of a fall risk score on outcomes after dismissal is not well known. Thus, we aimed to determine whether a fall risk score is associated with death and hospital readmissions in patients with a recent incident cardiovascular disease event. Methods and Results In this retrospective cohort study, Olmsted County, MN patients with incident heart failure, myocardial infarction, or atrial fibrillation between August 1, 2005, and December 31, 2011, who were hospitalized within 180 days after the event were studied. Fall risk was measured by the Hendrich II fall risk model. Patients were followed for death or readmission within 30 days or 1 year. Among 2456 hospitalized patients with recent incident cardiovascular disease (549 heart failure, 784 myocardial infarction, 1123 atrial fibrillation; mean [SD] age, 71 [15] years; 55% men), the fall risk score was high in 22% of patients and moderate in 38%. The risk of death was increased if the fall risk score was increased, independent of age and comorbidities (moderate hazard ratio, 1.51; 95% CI, 1.09–2.08; high hazard ratio, 3.49; 95% CI, 2.52–4.85). Similarly, the risk of 30-day readmissions was substantially increased with a greater fall risk score (moderate hazard ratio, 1.29; 95% CI, 1.03–1.62; high hazard ratio, 1.63; 95% CI, 1.23–2.15). Results were similar for readmissions within 1 year. Conclusions More than half of hospitalized patients with recent incident cardiovascular disease have an elevated fall risk score, which is associated with an increased risk in readmissions and death. These results delineate an approach for risk stratification and management that may prevent readmissions and improve survival.

Background Ideal cardiovascular health metrics (defined by the American Heart Association Life's Simple 7 [LS7]) are suboptimal among blacks, which results in high risk of cardiovascular disease. We examined the association of multiple stressors with LS7 components among blacks. Methods and Results Using a community‐based cohort of blacks (N=4383), we examined associations of chronic stress, minor stressors, major life events, and a cumulative stress score with LS7 components (smoking, diet, physical activity, body mass index, blood pressure, total cholesterol, and fasting plasma glucose) and an LS7 composite score. Multivariable logistic regression assessed the odds of achieving intermediate/ideal levels of cardiovascular health adjusted for demographic, socioeconomic, behavioral, and biomedical factors. The LS7 components with the lowest percentages of intermediate/ideal cardiovascular health levels were diet (39%), body mass index (47%), and physical activity (51%). Higher chronic, minor, and cumulative stress scores were associated with decreased odds (odds ratio [OR]) of achieving intermediate/ideal levels for smoking (OR [95% confidence interval], 0.80 [0.73–0.88], 0.84 [0.75–0.94], and 0.81 [0.74–0.90], respectively). Participants with more major life events had decreased odds of achieving intermediate/ideal levels for smoking (OR, 0.84; 95% confidence interval, 0.76–0.92) and fasting plasma glucose (OR, 0.90; 95% confidence interval, 0.82–0.98). Those with higher scores for minor stressors and major life events were less likely to achieve intermediate or ideal LS7 composite scores (OR [95% confidence interval], 0.89 [0.81–0.97] and 0.91 [0.84–0.98], respectively). Conclusions Blacks with higher levels of multiple stress measures are less likely to achieve intermediate or ideal levels of overall cardiovascular health (LS7 composite score), specific behaviors (smoking), and biological factors (fasting plasma glucose).

Background Perceived social isolation has been shown to have a negative impact on health outcomes, particularly among older adults. However, these relationships have not been fully examined among patients with heart failure. Methods and Results Residents from 11 southeast Minnesota counties with a first‐ever International Classification of Diseases, Ninth Revision ( ICD‐9 ) code 428 for heart failure between January 1, 2013, and March 31, 2015 (N=3867), were prospectively surveyed to measure perceived social isolation. A total of 2003 patients returned the survey (response rate, 52%); 1681 patients completed all questions and were retained for analysis. Among these patients (53% men; mean age, 73 years), ≈19% (n=312) had moderate perceived social isolation and 6% (n=108) had high perceived social isolation. After adjustment, patients reporting moderate perceived social isolation did not have an increased risk of death, hospitalizations, or emergency department visits compared with patients reporting low perceived social isolation; however, patients reporting high perceived social isolation had >3.5 times increased risk of death (hazard ratio, 3.74; 95% confidence interval [ CI ], 1.82–7.70), 68% increased risk of hospitalization (hazard ratio, 1.68; 95% CI , 1.18–2.39), and 57% increased risk of emergency department visits (hazard ratio, 1.57; 95% CI , 1.09–2.27). Compared with patients who self‐reported low perceived social isolation, patients reporting moderate perceived social isolation had a 16% increased risk of outpatient visits (rate ratio, 1.16; 95% CI , 1.03–1.31), whereas those reporting high perceived social isolation had a 26% increased risk (rate ratio, 1.26; 95% CI , 1.04–1.53). Conclusions In patients with heart failure, greater perceived social isolation is associated with an increased risk of death and healthcare use. Assessing perceived social isolation during the clinical encounter with a brief screening tool may help identify patients with heart failure at greater risk of poor outcomes.

Objectives To characterize the individual and combined effects of multimorbidity and functional limitation on healthcare use and mortality in a large, community cohort of individuals with heart failure (HF). Design Prospective cohort study. Setting Eleven southeastern Minnesota counties. Participants Individuals (mean age 74, 54% male) with a first‐ever HF code ( International Classification of Diseases, Ninth Revision code 428 or Tenth Revision code I50) between January 1, 2013 and March 31, 2016 (N=2,692). Measurements Eight activities of daily living measured using a survey on a Likert scale (1=without any difficulty, 5=unable to do; median=8). Participants with a score greater than 8 were categorized as having functional limitation. Multimorbidity was defined as having 2 or more noncardiac comorbidities. Results Twenty‐five percent of participants had neither multimorbidity nor functional limitation, 35% had multimorbidity, 9% had functional limitation, and 31% had both. After adjustment, participants with multimorbidity and functional limitation had greater risks of all outcomes (death: hazard ratio (HR)=4.92, 95% confidence interval (CI)=3.03–8.00; emergency department (ED) visit: HR=3.67, 95% CI=2.94–4.59; hospitalization: HR=3.66, 95% CI=2.85–4.70; outpatient visit: HR=1.73, 95% CI=1.52–1.96) than those with neither. Participants with functional limitation alone had greater risks of death (HR=4.84, 95% CI=2.78–8.43), ED visits (HR=2.35, 95% CI=1.75–3.16), and hospitalizations (HR=2.10, 95% CI=1.52–2.88) but not outpatient visits. Those with multimorbidity alone had similar risks of ED visits and hospitalizations as those with functional limitation alone but were more likely to have outpatient visits (HR=1.50, 95% CI=1.34–1.67). Conclusion Individuals with both multimorbidity and functional limitation have the highest risk of death and healthcare use. Individuals with only functional limitation have similar rates of hospitalizations and ED visits as those with only multimorbidity, underscoring the need to consider both when managing individuals with HF.