Integrated Health Care Delivery System Research Articles

Predictors of Postpartum Readmission Among Hypertensive Pregnancies in a Large Integrated Health Care Delivery System

To estimate the association between hypertensive disorders of pregnancy (HDP) and timing of all-cause postpartum (PP) hospital readmissions, and to examine patient-level and care-delivery characteristics associated with PP readmissions.

A mixed methods evaluation of medication reconciliation in the primary care setting.

To understand the extent to which behaviors consistent with high quality medication reconciliation occurred in primary care settings and explore barriers to high quality medication reconciliation. Fully mixed sequential equal status design including ethnographic observations, semi-structured interviews, and surveys. Primary care practices within an integrated healthcare delivery system in the United States. We conducted 170 observations of patient encounters across 15 primary care clinics, 48 semi-structured interviews with staff, and 10 semi-structured interviews with patients. We also sent out surveys to 2,541 eligible staff with 616 responses (24% response rate) and to 5,132 eligible patients with 577 responses (11% response rate). Inconsistency emerged as a major barrier to effective medication reconciliation. This inconsistency was present across a variety of factors such as the lack of standardized workflows for conducting medication reconciliation, a lack of knowledge about medication and the process of medication reconciliation, varying levels of importance ascribed to medication reconciliation, and inadequate integration of medication reconciliation into clinical workflows. Findings were generally consistent across all data collection methods. We have identified several barriers which impact the process of medication reconciliation in primary care settings. Our key finding is that the process of medication reconciliation is plagued by inconsistencies which contribute to inaccurate medication lists. These inconsistencies can be broken down into several categories (standardization, knowledge, importance, and inadequate integration) which can be targets for future studies and interventions.

Patient-Reported Outcomes Integrated Within an Electronic Medical Record in Patients With Head and Neck Cancer.

Patient-reported outcome (PRO) tools lead to clinical benefits, including improved overall survival for patients with cancer. However, routine implementation of PROs in clinical practice within the electronic medical record (EMR) by integrated health care delivery systems remains limited. We studied the use of a PRO tool for patients with head and neck cancer (HNC) integrated in an EMR at Kaiser Permanente in Northern California. Between August 2017 and December 2019, patients with newly diagnosed HNC were surveyed at baseline, then every 3 months using the Functional Assessment of Cancer Therapy-General 7 and Functional Assessment of Cancer Therapy-Head and Neck (version 4). A medical assistant performed a baseline survey on diagnosis and then notified patients electronically per surveillance protocol. Patients who did not respond to online PRO surveys could complete them via telephone or in-person appointments with medical assistants. Abnormal findings on PRO surveys were referred to appropriate members of the care team or the treating Otolaryngology-Head and Neck Surgery physicians. Two hundred ninety patients received baseline surveys. Patients received up to a maximum of eight subsequent surveys. Of a total of 597 electronic surveys, 585 (97.9%) were completed. The percentage of patients completing each interval survey ranged from 92% to 100%. Multivariate Poisson regression analysis showed patients with English as their primary language and an online secure account were the most likely to complete surveys compared with those patients with non-English as a primary language and without an online account. PRO tools can be effectively used within the EMR for patients with HNC with a high response rate provided there is strong engagement from a dedicated member of the care team. This has important implications for designing clinical trials and symptom monitoring in clinical practices that incorporate EMRs.

Caesarean Delivery and the Risk of Atopic Dermatitis in Children

To investigate whether cesarean delivery increases a child’s risk of atopic dermatitis (AD) by the age of 4 by using data from a large integrated health care delivery system. It has been hypothesized that cesarean delivery may disrupt maternal–infant microbial transfer, thereby affecting immune system development and increasing the risk of atopic dermatitis in children.Mothers who had a singleton live birth at a Kaiser Permanente North California or contracting hospital between January 1, 2005, to January 1, 2014 and whose child had continuous enrollment in the Kaiser Permanente North California system for at least 4 years were included in the study (n = 173 105). Among this birth cohort, a subsample of births (n = 102 327) occurring mostly after 2008 because of the adoption of electronic health records included more detailed information of cesarean delivery conditions (start of labor, timing of membrane rupture, and cesarean delivery indication).In this observational study, the researchers collected maternal and child information using these data sources: electronic health record, pharmacy databases, state birth records, and a prospectively collected survey regarding breastfeeding behavior at 2 months of age. The authors used modified Poisson regression models with robust variance estimation to assess the association between cesarean delivery and atopic dermatitis overall and when stratified by demographic and labor and delivery characteristics.There were 173 105 children born to 135 102 mothers during the study time period. A total of 12% of the children (n = 20 819) met the case definition for AD, and 26.6% of the births were born via cesarean delivery. The prevalence of AD was higher among children who were male, born term, were first born and did not have an NICU admission. Mothers of children with AD were more likely to be people of color (African-American, Asian, or Pacific Islander) and have underweight prepregnancy BMI. The prevalence of cesarean delivery was higher among mothers who were older, had higher levels of education, had a higher prepregnancy BMI, and were African American race. Infants born by cesarean delivery were more likely to be born at an earlier gestational age, had a high or low birth weight, were in the NICU, and were first born. In the primary multivariate model controlling for all covariates, there was little evidence that children born via cesarean delivery were more likely to develop AD by age 4, compared with children born vaginally (risk ratio [RR]: 1.02; confidence interval [CI]: 0.99 to 1.05). Among all stratified models (AD diagnosis before or after 6 months, maternal atopy, gestational age, birth order, prepregnancy BMI, and child’s sex), the RR for cesarean delivery and AD was minimal. Cesarean delivery conditions indicative of the least exposure to maternal microbiome (ie, no labor and short interval between membrane rupture and delivery) revealed no evidence of association with AD. In addition, intrapartum antibiotics, breastfeeding, or familial factors did not strongly influence an association. Only term infants born via cesarean delivery because of a breech or other malpresentation with labor (RR: 1.07; CI 0.94 to 1.21) or because of failure to progress (RR: 1.09; CI: 1.01 to 1.17) showed modest evidence of higher risk of AD by the age of 4 years, compared with term infants born vaginally.Delivery by cesarean delivery was not associated with an increased risk of atopic dermatitis by the age of 4 years.The authors point out that this is the largest study-to-date using observational data for a US-based birth cohort regarding mode of delivery and atopic dermatitis development. The results of this study are consistent with previous research findings. The strengths of this study are the large amount of data analyzed and the prospectively collected information on cesarean delivery conditions, intrapartum antibiotic administration, breastfeeding information, and within-family characteristics (via a matched sibling analysis).

Is depression associated with cognitive decline in a diverse cohort of the oldest‐old? Findings from the LifeAfter90 Study

AbstractBackgroundDepression impacts 50% of the elderly population and is a risk factor for dementia but its prevalence and role in cognitive decline in Non‐Whites and the oldest‐old is unclear.MethodLifeAfter90 is a racially/ethnically diverse cohort of individuals aged 90+ that launched in 2018 with the aim of characterizing cognitive and brain aging after the 9th decade of life. Participants are members of an integrated healthcare delivery system without a dementia diagnosis in their medical record at recruitment. Depression at baseline was evaluated with the Global Depression Scale (GDS) and defined as GDS>4. Cognitive testing every 6 months uses the Spanish and English Neuropsychological Assessment Scales (SENAS), to assess cognitive domains of episodic memory, semantic memory, and executive function. Linear mixed models with random intercepts, evaluated the baseline and longitudinal association of depression with each SENAS domain. The models used years since baseline as the time variable, and adjusted for baseline age, sex, race/ethnicity, education, and practice effects. An interaction term (depression*time) tested for differences in decline in cognitive domains by baseline depression status.ResultOf 655 participants 32% are White, 22% Asian, 22% Black, 15% Latino, and 8% Other/Multiracial. 35% had <high school education, 38% were male, and mean age was 93 years. At baseline, 32% were classified as having depression with the highest prevalence among Latinos and those with <high school education (Table 1). The average follow‐up was 0.9 years (range: 0‐1.9). Depression was associated with poorer semantic memory (ß=‐0.17) and executive function (ß=‐0.18) but not with episodic memory (Table 2). Semantic memory (ß=‐0.28) and executive function (ß=‐0.11) decreased over time, however episodic memory did not. The semantic memory and executive function decline did not differ by depression (Figure 1 & Table 2).ConclusionDepression impacts one third of this ethnically diverse cohort of oldest‐old individuals and ais more prevalent in Latino participants. While depression is associated with poorer baseline semantic memory and executive function; it is not associated with episodic memory, nor is it associated with greater cognitive decline. These results suggest that while depression may impact cognitive performance it may not impact cognitive change in those aged 90+.

Implementation of a Multidisciplinary Expert Testicular Cancer Tumor Board Across a Large Integrated Healthcare Delivery System Via Early Case Ascertainment

In 2016, Kaiser Permanente Northern California began regionalizing testicular cancer care using population-based tumor board review. This mixed methods evaluation describes implementation outcomes and learnings. We conducted in-depth interviews with key stakeholders, administered surveys to local oncologists and urologists, and used clinical data to evaluate changes in care delivery during 2015-2018. An average of 135 patients with testicular cancer were diagnosed each year. Interviews with 16 key stakeholders provided several insights. Implementation resulted in high levels of satisfaction, was dependent on leadership and staff at various levels, and required technology and consulting solutions aligned to user agreements and clinical workflows. Of 123 local oncologists and urologists who completed surveys, 97% understood why care was regionalized and 89% agreed that tumor board review improved treatment decisions. Among 177 patients with stage I seminoma, the percentage appropriately observed rather than treated with adjuvant chemotherapy or radiation therapy increased from 48% (95% CI, 35 to 62) in 2015 to 87% (75 to 99) in 2018. Review altered care based on pathology and radiology re-review in 14.5 % of cases. Regionalization was feasible and effective.

Examining possible gender differences in the association between education and late‐life cognitive function in the LifeAfter90 Study, a large multiethnic cohort of the oldest‐old

AbstractBackgroundEducational attainment predicts late‐life cognitive ability and dementia. Historically, women have had fewer educational opportunities than men. We examine if the association between educational attainment and cognitive level and decline differ by gender.MethodLifeAfter90 is an ongoing cohort of long‐time members of an integrated healthcare delivery system ≥90 years old without a dementia diagnosis at recruitment. Interviews and cognitive assessment occurred every six months and included self‐reported education (≤high school, some college, and ≥college graduate (reference)) and Spanish and English Neuropsychological Assessment Scales (SENAS). SENAS encompasses the domains of verbal episodic memory, semantic memory, and executive function. Mixed models with random intercepts estimated possible differences in the association between education and late‐life cognition adjusting for time (years since baseline), baseline age, race/ethnicity, and practice effects. Practice effects were accounted for by using an indicator variable for the first visit. Interactions terms allowed for possible gender differences in the association of education with baseline cognitive scores (i.e. education*male) and cognitive change (i.e., education*male*time).ResultThe 669 participants were on average 92.8 years of age, 63% were women, 34% were college graduates, and 69% were racial/ethnic minorities (Table 1 and Table 2). Lower levels of education were associated with worse performance on baseline cognitive assessment but not performance over time. Women with ≤high school or some college averaged worse baseline cognition in each domain than women who had completed college (Table 3). There was no significant gender by education interaction for verbal episodic memory (p‐value=0.22) or semantic memory (p‐value=0.26), but the disadvantage in executive function associated with lower education was significantly attenuated in men (some college*male β=0.32, 95% CI=0.08, 0.56; p‐value=0.01). There were no differences across genders in the association between education and any of the domain specific cognitive scores over time.ConclusionThe association between education and late‐life cognitive function appears to be similar across genders with the exception of college completion having a larger effect on baseline executive function scores for women than for men. We will continue to explore these differences as additional visits occur in this ongoing cohort of ethnically diverse oldest‐old individuals.

Human immunodeficiency virus infection and risks of morbidity and death in adults with incident heart failure.

AimsHuman immunodeficiency virus (HIV) increases the risk of heart failure (HF), but whether it influences subsequent morbidity and mortality remains unclear.Methods and resultsWe investigated the risks of hospitalization for HF, HF-related emergency department (ED) visits, and all-cause death in an observational cohort of incident HF patients with and without HIV using data from three large US integrated healthcare delivery systems. We estimated incidence rates and adjusted hazard ratios (aHRs) by HIV status at the time of HF diagnosis for subsequent outcomes. We identified 448 persons living with HIV (PLWH) and 3429 without HIV who developed HF from a frequency-matched source cohort of 38 868 PLWH and 386 586 without HIV. Mean age was 59.5 ± 11.3 years with 9.8% women and 31.8% Black, 13.1% Hispanic, and 2.2% Asian/Pacific Islander. Compared with persons without HIV, PLWH had similar adjusted rates of HF hospitalization [aHR 1.01, 95% confidence interval (CI): 0.81–1.26] and of HF-related ED visits [aHR 1.22 (95% CI: 0.99–1.50)], but higher adjusted rates of all-cause death [aHR 1.31 (95% CI: 1.08–1.58)]. Adjusted rates of HF-related morbidity and all-cause death were directionally consistent across a wide range of CD4 counts but most pronounced in the subset with a baseline CD4 count <200 or 200–499 cells/μL.ConclusionIn a large, diverse cohort of adults with incident HF receiving care within integrated healthcare delivery systems, PLWH were at an independently higher risk of all-cause death but not HF hospitalizations or HF-related ED visits. Future studies investigating modifiable HIV-specific risk factors may facilitate more personalized care to optimize outcomes for PLWH and HF.

A Natural Language Processing–Based Approach for Identifying Hospitalizations for Worsening Heart Failure Within an Integrated Health Care Delivery System

The current understanding of epidemiological mechanisms and temporal trends in hospitalizations for worsening heart failure (WHF) is based on claims and national reporting databases. However, these data sources are inherently limited by the accuracy and completeness of diagnostic coding and/or voluntary reporting. To assess the overall burden of and temporal trends in the rate of hospitalizations for WHF. This cohort study, performed from January 1, 2010, to December 31, 2019, used electronic health record (EHR) data from a large integrated health care delivery system. Calendar year trends. Hospitalizations for WHF (ie, excluding observation stays) were defined as 1 symptom or more, 2 objective findings or more including 1 sign or more, and 2 doses or more of intravenous loop diuretics and/or new hemodialysis or continuous kidney replacement therapy. Symptoms and signs were identified using natural language processing (NLP) algorithms applied to EHR data. The study population was composed of 118 002 eligible patients experiencing 287 992 unique hospitalizations (mean [SD] age, 75.6 [13.1] years; 147 203 [51.1%] male; 1655 [0.6%] American Indian or Alaska Native, 28 451 [9.9%] Asian or Pacific Islander, 34 903 [12.1%] Black, 23 452 [8.1%] multiracial, 175 840 [61.1%] White, and 23 691 [8.2%] unknown), including 65 357 with a principal discharge diagnosis and 222 635 with a secondary discharge diagnosis of HF. The study population included 59 868 patients (20.8%) with HF with a reduced ejection fraction (HFrEF) (<40%), 33 361 (11.6%) with HF with a midrange EF (HFmrEF) (40%-49%), 142 347 (49.4%) with HF with a preserved EF (HFpEF) (≥50%), and 52 416 (18.2%) with unknown EF. A total of 58 042 admissions (88.8%) with a primary discharge diagnosis of HF and 62 764 admissions (28.2%) with a secondary discharge diagnosis of HF met the prespecified diagnostic criteria for WHF. Overall, hospitalizations for WHF identified on NLP-based algorithms increased from 5.2 to 7.6 per 100 hospitalizations per year during the study period. Subgroup analyses found an increase in hospitalizations for WHF based on NLP from 1.5 to 1.9 per 100 hospitalizations for HFrEF, from 0.6 to 1.0 per 100 hospitalizations for HFmrEF, and from 2.6 to 3.9 per 100 hospitalizations for HFpEF. The findings of this cohort study suggest that the burden of hospitalizations for WHF may be more than double that previously estimated using only principal discharge diagnosis. There has been a gradual increase in the rate of hospitalizations for WHF with a more noticeable increase observed for HFpEF.

Incidence of SARS-CoV-2 Infection, Emergency Department Visits, and Hospitalizations Because of COVID-19 Among Persons Aged ≥12 Years, by COVID-19 Vaccination Status - Oregon and Washington, July 4-September 25, 2021.

Population-based rates of infection with SARS-CoV-2 (the virus that causes COVID-19) and related health care utilization help determine estimates of COVID-19 vaccine effectiveness and averted illnesses, especially since the SARS-CoV-2 B.1.617.2 (Delta) variant began circulating in June 2021. Among members aged ≥12 years of a large integrated health care delivery system in Oregon and Washington, incidence of laboratory-confirmed SARS-CoV-2 infection, emergency department (ED) visits, and hospitalizations were calculated by COVID-19 vaccination status, vaccine product, age, race, and ethnicity. Infection after full vaccination was defined as a positive SARS-CoV-2 molecular test result ≥14 days after completion of an authorized COVID-19 vaccination series.* During the July-September 2021 surveillance period, SARS-CoV-2 infection occurred among 4,146 of 137,616 unvaccinated persons (30.1 per 1,000 persons) and 3,009 of 344,848 fully vaccinated persons (8.7 per 1,000). Incidence was higher among unvaccinated persons than among vaccinated persons across all demographic strata. Unvaccinated persons with SARS-CoV-2 infection were more than twice as likely to receive ED care (18.5%) or to be hospitalized (9.0%) than were vaccinated persons with COVID-19 (8.1% and 3.9%, respectively). The crude mortality rate was also higher among unvaccinated patients (0.43 per 1,000) than in fully vaccinated patients (0.06 per 1,000). These data support CDC recommendations for COVID-19 vaccination, including additional and booster doses, to protect individual persons and communities against COVID-19, including illness and hospitalization caused by the Delta variant (1).

Clinical outcomes of dabigatran use in patients with non-valvular atrial fibrillation and weight >120 kg

BackgroundPatients with obesity were underrepresented in studies evaluating the safety and effectiveness of direct oral anticoagulants (DOAC) in patients with non-valvular atrial fibrillation (NVAF). This study compared clinical outcomes in patients with NVAF and weighing >120 kg and ≤120 kg who were receiving dabigatran. Materials and methodsThis retrospective, matched, longitudinal cohort study included patients from three integrated healthcare delivery systems. Patients ≥18 years of age with NVAF were included if between September 1, 2016 and June 30, 2019 they received dabigatran. Patients >120 kg and ≤120 kg were matched up to 1:6 on age, sex, and CHA2DS2-VASc score. Data were extracted from administrative databases. The primary outcome was a composite of ischemic stroke, clinically-relevant bleeding, systemic embolism, and all-cause mortality. Multivariable regression analyses were performed. Results777 and 3522 patients >120 kg and ≤120 kg, respectively, were matched. The >120 kg group tended to be younger with a higher burden of chronic disease. There was no difference between groups in the composite outcome (adjusted hazard ratio [AHR] 1.10, 95% confidence interval 0.89–1.37) or individual components of the composite. A subanalysis of clinically-relevant bleeding identified that patients >120 kg were at a greater risk of gastrointestinal bleeding (AHR 1.44, 95% CI 1.01–2.05). ConclusionsIn patients with NVAF and >120 kg, dabigatran use was associated with a small increased risk of gastrointestinal bleeding but no differences in stroke, mortality or clinically-relevant bleeding. These findings suggest that dabigatran use is reasonable in patients with NVAF and weight >120 kg.

Treatment and Follow-up Care Associated With Patient-Scheduled Primary Care Telemedicine and In-Person Visits in a Large Integrated Health System

Telemedicine visits can offer patients convenient access to a clinician, but it is unclear whether treatment differs from that with in-person visits or how often patients require in-person follow-up. To examine whether physician prescribing and orders differ between telemedicine and office visits, whether physicians conducting telemedicine visits are more likely to require in-person follow-up, and whether telemedicine visits are associated with more health events. This cohort study included all patients who scheduled primary care appointments through the patient portal of a large integrated health care delivery system newly implementing patient-scheduled video telemedicine visits from January 2016 to May 2018. Adjusted rates of any medication prescribed or laboratory tests or imaging ordered and rates of follow-up health care utilization (in-person visits, emergency department visits, and hospitalizations) within 7 days after the index visit, stratified by index primary care visit type, were generated using multivariable adjustment for patient, access, and clinical characteristics. This study included 1 131 722 patients (611 821 [54%] female; mean [SD] age, 43 [22] years) with 2 178 440 total appointments (307 888 [14%] telemedicine), of which 13.5% were for patients younger than 18 years, 22.2% were for patients 65 years or older, and 54.9% were for female patients. After adjustment, 38.6% (95% CI, 38.0%-39.3%) of video visits, 34.7% (95% CI, 34.5%-34.9%) of telephone visits, and 51.9% (95% CI, 51.8%-52.0%) of office visits had any medication prescribed; laboratory tests or imaging were ordered for 29.2% (95% CI, 28.5%-29.8%) of video visits, 27.3% (95% CI, 27.1%-27.5%) of telephone visits, and 59.3% (95% CI, 59.3%-59.4%) of clinic visits. After adjustment, follow-up visits within 7 days occurred after 25.4% (95% CI, 24.7%-26.0%) of video visits, 26.0% (95% CI, 25.9%-26.2%) of telephone visits, and 24.5% (95% CI, 24.5%-24.6%) of office visits. Adjusted emergency department visits and rates of hospitalizations were not statistically significantly different by primary care index visit type. In this cohort study of patient self-scheduled primary care telemedicine visits within ongoing patient-physician relationships, prescribing and orders were significantly lower for telemedicine visits than for clinic visits, with slightly higher follow-up office visits for telemedicine but no difference in health events (emergency department visits or hospitalizations). Video or telephone visits may be a convenient and efficient way to access primary care and address patient needs.

Challenges with implementing the Diabetes Prevention Program for Medicare beneficiaries in an integrated health system.

In April 2018, CMS began reimbursing both clinical and community settings for providing the CDC-recognized Diabetes Prevention Program (DPP) to eligible Medicare beneficiaries. To better understand the process of offering the program to Medicare beneficiaries, we interviewed relevant stakeholders in a large, integrated health care delivery system. Qualitative interview study. We conducted semistructured interviews with 12 delivery system stakeholders. Data were analyzed following a thematic analysis approach. Stakeholders described systemic challenges to the implementation of Medicare DPP (MDPP), including inadequate reimbursement for the health care system, low awareness of MDPP among patients and providers, and challenges with utilizing third-party vendors to connect patients to CDC-recognized MDPPs. Although the reimbursement of DPP for Medicare beneficiaries was a landmark decision, the current structure and requirements make it difficult for health systems and community-based providers to implement and promote this benefit. This study highlights the challenges that even integrated health systems are facing to implement MDPP, as well as potential strategies to overcome these barriers and expand the reach of the program. Medicare should seek ways to increase the financial incentives and decrease the barriers associated with implementing MDPP.

Epidemiology Unknown: Developing an Approach to Identify Potential Smoldering Multiple Myeloma Cases

BackgroundSmoldering Multiple Myeloma (SMM) is an asymptomatic clonal plasma cell disorder that identifies patients at risk for progression to Multiple Myeloma (MM). The standard of care for SMM has traditionally been observation, but some cancer centers are now treating high-risk SMM before progression to MM. The diagnostic criteria for SMM have also changed in recent years, and current estimates of SMM are derived from large MM databases and observations from tertiary centers. The goal of this study was to develop an approach for identifying SMM cases in a large integrated healthcare delivery system to better characterize the epidemiology of SMM in community-based populations.MethodsThis retrospective, observational study was conducted in Kaiser Permanente Northern California (KPNC) using KPNC SEER-based Cancer Registry data and information from the electronic health record (EHR). Potential SMM cases from 1/1/2010 to 12/31/2018 were identified using three approaches: Group 1 - identified via the KPNC Cancer Registry based on indicators of ‘asymptomatic myeloma’, ‘evolving myeloma’, and ‘smoldering myeloma’; Group 2 - identified via the KPNC Cancer Registry as MM cases who had a physician visit note containing the word ‘smoldering’ but did not begin treatment within 1 year of diagnosis; Group 3 - identified via the KPNC Cancer Registry as MM cases who had a physician visit note containing the word ‘smoldering’ but did begin treatment within 1 year of diagnosis. Chart review was performed for these potential SMM cases (Groups 1-3) to document initial bone marrow biopsy results (bone marrow plasma cell percentage, BMPC) and skeletal findings (presence or absence of lytic bone lesions) around the time of biopsy. When BMPC was reported as a range, the highest value was captured. Patient demographics (age, sex and race/ethnicity) were obtained from the EHR. Bivariate analyses were performed using the chi-squared test and the Wilcoxon-Mann-Whitney nonparametric test. For binomial comparisons by mode of potential SMM case identification, Groups 1 and 2 were combined and compared to Group 3.ResultsA total of 471 potential SMM cases were identified, including 178 (37.8%) via Group 1, 35 (7.4%) via Group 2, and 258 (54.8%) via Group 3 (Figure). The median age was 71 years (interquartile range, IQR 62-78) and 40.0% were female. The racial/ethnic distribution included 57.1% White, 17.6% Black, 10.8% Hispanic, 13.6% Asian, and 0.9% other/unknown race. There were no significant differences across groups (Group 1+2 vs Group 3) with respect to age (p=0.07), sex (p=0.85), or race/ethnicity (p=0.81).There were 442 (93.8%) who underwent bone marrow biopsy. Among those with BMPC data, the median BMPC for Group 1 was 20.0% (IQR 10.0%-28.0%); for Group 2 was 25.0% (IQR 12.5%-50.0%), for Group 3 was 28.0% (IQR 15.0%-50.0%) (p<0.001 comparing Groups 1+2, combined median 20.0%, IQR 10.0%-30.0%, vs Group 3, 28.0%, IQR 15.0%-50.0%). The proportion with BMPC ≥60% was 4%, 13%, and 22% for Groups 1, 2, and 3, respectively (Figure).There were 413 (87.7%) who had skeletal imaging (n=405, 86.0% with available results) within 6 months of diagnosis. Of those with imaging results available to view, n=68 were found to have lytic lesions; 7.6% among Group 1, 6.9% among Group 2, 24.9% among Group 3 (p<0.001 comparing Groups 1+2, 7.5%, vs Group 3, 24.9%).DiscussionThis study used a multifaceted approach to identify potential SMM cases from a large real-world clinical population in an integrated health system. We used an approach similar to prior SMM epidemiological studies and also included those with physician visit notes specifically containing the word ‘smoldering’ within the text. The vast majority of our cohort had BMPC between 10 and 60%, but those who received treatment within 1 year had greater BMPC and a higher proportion of lytic lesions. This suggests those who received treatment, Group 3, may have actually had a MM diagnosis, and physician visit notes containing the word ‘smoldering’ may have been intended to communicate something other than a SMM diagnosis. Further analyses will determine the effectiveness of each approach by confirming SMM cases according to the International Myeloma Working Group diagnostic criteria with incorporation of laboratory data and additional clinical findings. Among confirmed cases of SMM, the rate of progression to MM and the severity of end-organ damage at time of progression will be assessed. [Display omitted] DisclosuresLo: Novartis: Research Funding; Bristol-Myers-Squibb: Research Funding; CSL-Bering: Research Funding.

TCT-413 Outcomes of Adults With Severe Aortic Stenosis Undergoing Elective Versus Urgent or Emergent Transcatheter Aortic Valve Replacement Within an Integrated Health Care Delivery System

TCT-413 Outcomes of Adults With Severe Aortic Stenosis Undergoing Elective Versus Urgent or Emergent Transcatheter Aortic Valve Replacement Within an Integrated Health Care Delivery System

Population-based identification and temporal trend of children with primary nephrotic syndrome: The Kaiser Permanente nephrotic syndrome study.

Limited population-based data exist about children with primary nephrotic syndrome (NS). We identified a cohort of children with primary NS receiving care in Kaiser Permanente Northern California, an integrated healthcare delivery system caring for >750,000 children. We identified all children <18 years between 1996 and 2012 who had nephrotic range proteinuria (urine ACR>3500 mg/g, urine PCR>3.5 mg/mg, 24-hour urine protein>3500 mg or urine dipstick>300 mg/dL) in laboratory databases or a diagnosis of NS in electronic health records. Nephrologists reviewed health records for clinical presentation and laboratory and biopsy results to confirm primary NS. Among 365 cases of confirmed NS, 179 had confirmed primary NS attributed to presumed minimal change disease (MCD) (72%), focal segmental glomerulosclerosis (FSGS) (23%) or membranous nephropathy (MN) (5%). The overall incidence of primary NS was 1.47 (95% Confidence Interval:1.27-1.70) per 100,000 person-years. Biopsy data were available in 40% of cases. Median age for patients with primary NS was 6.9 (interquartile range:3.7 to 12.9) years, 43% were female and 26% were white, 13% black, 17% Asian/Pacific Islander, and 32% Hispanic. This population-based identification of children with primary NS leveraging electronic health records can provide a unique approach and platform for describing the natural history of NS and identifying determinants of outcomes in children with primary NS.

Hospitalization and mortality in heart failure with preserved ejection fraction: real-world data from a US integrated healthcare delivery system

Abstract Introduction Heart Failure with preserved ejection fraction (HFpEF) is a major public health and economic burden, but large real-world incident rates of clinical outcomes are scarce. Purpose Calculate adjusted incidence rates of hospitalization and mortality in patients with HFpEF compared with heart failure with reduced ejection fraction (HFrEF) using an integrated healthcare delivery System in the United States. Methods Non-interventional longitudinal cohort analysis between 2005 and 2017 of existing data of adult patients with a HF diagnosis (ICD9/ICD10) in the electronic medical record of Kaiser Permanent Northwest. Clinical outcomes of interest were all-cause hospitalizations, heart failure hospitalization (HHF, defined as a discharge diagnosis of HF in the primary position) and all-cause mortality. Age/sex adjusted incidence rates were assessed generalized estimating equations over 15 years of follow-up (2005–2019). Results Of the 37,773 patients with HF diagnosis, 36% had no EF available. 46.4% were categorized as HFpEF, 7.2% HFmrEF and 10.4% HFrEF. Compared with patients with HFrEF, those with HFpEF were older (72.1 vs. 68.3 years), more likely to be female (55% vs. 33.2%), had higher BMI (31.9 vs. 29.3 kg/m2), higher SBP (130 vs. 121mmHg) and were less likely to have a history of myocardial infarction (26.1% vs. 43.8%). Patients with HFpEF were less likely to receive RAAS blockade compared with HFrEF (66.0% vs. 86.7%) but use of any HF-related medication (ACE/ARB, diuretics, β-blockers, or aldosterone agonists) was similarly high (89.6% vs 88.8%). As shown in Figure 1, adjusted mortality rates per 1,000 person-years, [95% CI] were lower in HFpEF (107.9 [105.5, 110.3]) vs. HFrEF (143.0 [136.5, 149.7]). Also, HHF were lower in HFpEF (41.6 [40.1, 43.2]) compared to HFrEF (72.6 [67.7, 77.9]). All-cause hospitalization rates were about 12% lower in HFpEF (197.3 [193.9, 200.7]) compared with HFrEF (222.8 [214.1, 231.7]). Nonetheless, due to the much larger size of the HFpEF group, these patients generated 1,982 more HHF, 10,427 more hospitalizations for any reason, and 6,723 more deaths during follow-up. Conclusions In a large real-world dataset of HF patients, incidence rates for hospitalization and mortality were relatively lower in HFpEF vs. HFrEF. However, given the greater prevalence of HFpEF (46.4% vs. 10.4% for HFrEF), this diagnosis posed a much higher public health burden. Funding Acknowledgement Type of funding sources: Private company. Main funding source(s): Boehringer Ingelheim

Differential Cardiometabolic Risk Factor Clustering Across U.S. Asian Ethnic Groups

Ethnic differences in the prevalence of diabetes, obesity, and hypertension are well described,1 but the differential clustering of these risk conditions, which have multiplicative effects on cardiovascular disease risk,2 is less well characterized for Asian Americans. Such information may inform intervention strategies that target multiple cardiovascular risk factors simultaneously. To better understand their contemporary risk profile in clinical settings, this study examines clustered cardiometabolic risks (CCRs) among Filipino, Chinese, and South Asian adults compared with those among other races/ethnicities in a large integrated healthcare delivery system.

Health Care for Our Nation's Veterans: A Policy Paper From the American College of Physicians.

The Veterans Health Administration (VHA) is the United States' largest integrated health care delivery system, serving over 9 million enrollees at nearly 1300 health care facilities. In addition to providing health care to the nation's military veterans, the VHA has a research and development program, trains thousands of medical residents and other health care professionals, and conducts emergency preparedness and response activities. The VHA has been celebrated for delivering high-quality care to veterans, early adoption of electronic medical records, and high patient satisfaction. However, the system faces challenges, including implementation of an expanded community care program, modernization of its electronic medical records system, and providing care to a population with complex needs. The position paper offers policy recommendations on VHA funding, the community care program, medical and health care professions training, and research and development.

QT Interval Dynamics and Cardiovascular Outcomes: A Cohort Study in an Integrated Health Care Delivery System.

BackgroundLong QT has been associated with ventricular dysrhythmias, cardiovascular disease (CVD) mortality, and sudden cardiac death. However, no studies to date have investigated the dynamics of within‐person QT change over time in relation to risk of incident CVD and all‐cause mortality in a real‐world setting.Methods and ResultsA cohort study among members of an integrated health care delivery system in Northern California including 61 455 people (mean age, 62 years; 60% women, 42% non‐White) with 3 or more ECGs (baseline in 2005–2009; mean±SD follow‐up time, 7.6±2.6 years). In fully adjusted models, tertile 3 versus tertile 1 of average QT corrected (using the Fridericia correction) was associated with cardiac arrest (hazard ratio [HR], 1.66), heart failure (HR, 1.62), ventricular dysrhythmias (HR, 1.56), all CVD (HR, 1.31), ischemic heart disease (HR, 1.28), total stroke (HR, 1.18), and all‐cause mortality (HR, 1.24). Tertile 3 versus tertile 2 of the QT corrected linear slope was associated with cardiac arrest (HR, 1.22), ventricular dysrhythmias (HR, 1.12), and all‐cause mortality (HR, 1.09). Tertile 3 versus tertile 1 of the QT corrected root mean squared error was associated with ventricular dysrhythmias (HR, 1.34), heart failure (HR, 1.28), all‐cause mortality (HR, 1.20), all CVD (HR, 1.14), total stroke (HR, 1.08), and ischemic heart disease (HR, 1.07).ConclusionsOur results demonstrate improved predictive ability for CVD outcomes using longitudinal information from serial ECGs. Long‐term average QT corrected was more strongly associated with CVD outcomes than the linear slope or the root mean squared error. This new evidence is clinically relevant because ECGs are frequently used, noninvasive, and inexpensive.