Abstract MP75: Recalibration and Additional Data Domains Leads to Modestly Improved Performance of Risk Calculators for Heart Failure Readmission

Circulation ◽  
2020 ◽  
Vol 141 (Suppl_1) ◽  
Author(s):  
Samuel T Savitz ◽  
Keane Lee ◽  
Jamal S Rana ◽  
Thomas K Leong ◽  
Grace Tabada ◽  
...  

Introduction: Heart failure (HF)-related hospitalizations are a growing public health burden. We evaluated two published risk calculators for predicting 30-day readmission after HF hospitalizations: 1) using the original coefficients, 2) updating the coefficients 3) developing a new model with additional variables and updated coefficients. Hypothesis: Recalibrating model coefficients and adding variables would improve the performance of existing 30-day readmission risk calculators. Methods: We identified 45,059 adults hospitalized for HF between 2012-2017 within Kaiser Permanente Northern California, an integrated healthcare delivery system. We used split sampling for development and validation testing. The risk calculators tested included: LACE+ Index and Yale CORE. We used logistic regression on our population to derive the recalibrated coefficients. For the model with additional variables, we included all variables used in the original models plus additional variables, including cardiovascular medication use and socioeconomic status. We used gradient boosting with k-fold cross validation to avoid overfitting. We assessed model performance using area under the curve (AUC) and calibration plots. Results: Discrimination (AUC) was poor using original models: LACE+ [0.56 (0.54-0.58)] and Yale CORE [0.55 (0.54-0.57)]. Recalibrating coefficients resulted in small improvements for LACE+ [0.58 (0.57, 0.60)] and Yale CORE [0.58 (0.57, 0.60)]. Adding variables resulted in a modest improvement for the gradient boosting model [0.61 (0.59, 0.62)]. Calibration plots (Figure 1) showed good calibration except for the Yale CORE model with the original coefficients. Conclusions: Recalibrating coefficients and incorporating prior medication and socioeconomic status led to modest, significant improvements in discrimination while maintaining good calibration. However, overall performance improvements are needed to increase the utility of these published risk calculators to predict readmission.

Author(s):  
Keane K. Lee ◽  
Rachel C. Thomas ◽  
Thida C. Tan ◽  
Thomas K. Leong ◽  
Anthony Steimle ◽  
...  

Background: In-person clinic follow-up within 7 days after discharge from a heart failure hospitalization is associated with lower 30-day readmission. However, health systems and patients may find it difficult to complete an early postdischarge clinic visit, especially during the current pandemic. We evaluated the effect on 30-day readmission and death of follow-up within 7 days postdischarge guided by an initial structured nonphysician telephone visit compared with follow-up guided by an initial clinic visit with a physician. Methods and Results: We conducted a pragmatic randomized trial in a large integrated healthcare delivery system. Adults being discharged home after hospitalization for heart failure were randomly assigned to either an initial telephone visit with a nurse or pharmacist to guide follow-up or an initial in-person clinic appointment with primary care physicians providing usual care within the first 7 days postdischarge. Telephone appointments included a structured protocol enabling medication titration, laboratory ordering, and booking urgent clinic visits as needed under physician supervision. Outcomes included 30-day readmissions and death and frequency and type of completed follow-up within 7 days of discharge. Among 2091 participants (mean age 78 years, 44% women), there were no significant differences in 30-day heart failure readmission (8.6% telephone, 10.6% clinic, P =0.11), all-cause readmission (18.8% telephone, 20.6% clinic, P =0.30), and all-cause death (4.0% telephone, 4.6% clinic, P =0.49). Completed 7-day follow-up was higher in 1027 patients randomized to telephone follow-up (92%) compared with 1064 patients assigned to physician clinic follow-up (79%, P <0.001). Overall frequency of clinic visits during the first 7 days postdischarge was lower in participants assigned to nonphysician telephone guided follow-up (48%) compared with physician clinic-guided follow-up (77%, P <0.001). Conclusions: Early, structured telephone follow-up after hospitalization for heart failure can increase 7-day follow-up and reduce in-person visits with comparable 30-day clinical outcomes within an integrated care delivery framework. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT03524534.


Author(s):  
Teresa Dalla Zuanna ◽  
Laura Cacciani ◽  
Giulia Barbieri ◽  
Erich Batzella ◽  
Francesco Tona ◽  
...  

Background: Heart failure (HF) represents a severe public health burden. In Europe, differences in hospitalizations for HF have been found between immigrants and native individuals, with inconsistent results. Immigrants face many barriers in their access to health services, and their needs may be poorly met. We aimed to compare the rates of avoidable hospitalization for HF among immigrants and native individuals in Italy. Methods: All 18- to 64-year-old residents of Turin, Venice, Reggio Emilia, Modena, Bologna, and Rome between January 1, 2001 and December 31, 2013 were included in this multicenter open-cohort study. Immigrants from high migratory pressure countries (divided by area of origin) were compared with Italian citizens. Age-, sex-, and calendar year-adjusted hospitalization rate ratios and the 95% CIs of avoidable hospitalization for HF by citizenship were estimated using negative binomial regression models. The hospitalization rate ratios were summarized using a random effects meta-analysis. Additionally, we tested the contribution of socioeconomic status to these disparities. Results: Of the 4 470 702 subjects included, 15.8% were immigrants from high migratory pressure countries. Overall, immigrants showed a nonsignificant increased risk of avoidable hospitalization for HF (hospitalization rate ratio, 1.26 [95% CI, 0.97–1.68]). Risks were higher for immigrants from Sub-Saharan Africa and for males from Northern Africa and Central-Eastern Europe than for their Italian citizen counterparts. Risks were attenuated adjusting for socioeconomic status, although they remained consistent with nonadjusted results. Conclusions: Adult immigrants from different geographic macroareas had higher risks of avoidable hospitalization for HF than Italian citizens. Possible explanations might be higher risk factors among immigrants and reduced access to primary health care services.


2014 ◽  
Vol 63 (12) ◽  
pp. A538
Author(s):  
Ali R. Rahimi ◽  
Elizabeth Neeley ◽  
Sherry Bowen ◽  
Carla Leto ◽  
Binwei Song

2022 ◽  
Vol 8 ◽  
Author(s):  
Chien-Liang Liu ◽  
You-Lin Tain ◽  
Yun-Chun Lin ◽  
Chien-Ning Hsu

ObjectiveThis study aimed to identify phenotypic clinical features associated with acute kidney injury (AKI) to predict non-recovery from AKI at hospital discharge using electronic health record data.MethodsData for hospitalized patients in the AKI Recovery Evaluation Study were derived from a large healthcare delivery system in Taiwan between January 2011 and December 2017. Living patients with AKI non-recovery were used to derive and validate multiple predictive models. In total, 64 candidates variables, such as demographic characteristics, comorbidities, healthcare services utilization, laboratory values, and nephrotoxic medication use, were measured within 1 year before the index admission and during hospitalization for AKI.ResultsAmong the top 20 important features in the predictive model, 8 features had a positive effect on AKI non-recovery prediction: AKI during hospitalization, serum creatinine (SCr) level at admission, receipt of dialysis during hospitalization, baseline comorbidity of cancer, AKI at admission, baseline lymphocyte count, baseline potassium, and low-density lipoprotein cholesterol levels. The predicted AKI non-recovery risk model using the eXtreme Gradient Boosting (XGBoost) algorithm achieved an area under the receiver operating characteristic (AUROC) curve statistic of 0.807, discrimination with a sensitivity of 0.724, and a specificity of 0.738 in the temporal validation cohort.ConclusionThe machine learning model approach can accurately predict AKI non-recovery using routinely collected health data in clinical practice. These results suggest that multifactorial risk factors are involved in AKI non-recovery, requiring patient-centered risk assessments and promotion of post-discharge AKI care to prevent AKI complications.


2020 ◽  
Vol 13 (Suppl_1) ◽  
Author(s):  
Matthew Mefford ◽  
Zimin Zhuang ◽  
Zhi Liang ◽  
Wansu Chen ◽  
Heather Watson ◽  
...  

Background: In recent years declines in the rate of mortality attributable to cardiovascular diseases have slowed and mortality attributable to heart failure (HF) has increased. Objective: To examine secular trends in mortality with HF as the underlying cause in Kaiser Permanente Southern California (KPSC), California, and the US among adults 45 years of age and older from 2001 and 2017. Methods: KPSC mortality rates with HF as an underlying cause from 2001 to 2017 were derived through linkage with California State death files and were compared with rates in California and the US. Rates were age-standardized to the 2000 US Census population. Trends were examined overall and among men and women, separately, using best-fit Joinpoint regression models. Average annual percent change (AAPC) and 95% confidence intervals (CI) were calculated for the overall study period, and within earlier (2001-2011) and later (2011-2017) time periods. Results: Between 2001-2017, age-adjusted mortality rates with HF as the underlying cause were lower comparing KPSC to California and the US. In KPSC, rates increased from 23.9 to 44.7 per 100,000 person-years (PY) in KPSC, representing an AAPC of 1.3% (95% CI 0.0%, 2.6%). (Table) During the same time period, HF mortality rates in California also increased from 33.9 to 46.5 per 100,000 PY (AAPC 1.5%, 95% CI 0.3%, 2.7%), while remaining unchanged in the US at 57.9 per 100,000 PY in 2001 and 2017 (AAPC 0.0%, 95% CI -0.5%, 0.5%). AAPCs were not statistically different comparing KPSC to both California and the US (all p > 0.05). Between 2001-2011, rates of HF mortality increased in KPSC (AAPC 1.3%, 95% CI 0.0, 2.6), non-significantly increased in California (AAPC 0.2%, 95% CI -0.8%, 1.2%) and decreased in the US (AAPC -2.1%, 95% CI -2.7%, -1.5%). Between 2011-2017, rates of HF mortality increased in KPSC (AAPC 1.3%, 95% CI 0.0%, 2.6%), California (AAPC 3.7%, 95% CI 1.0%, 6.5%), and the US (AAPC 3.6%, 95% CI 2.4%, 4.8%) except among KPSC women (AAPC 0.3% [95% CI -1.6%, 2.2%]). Conclusion: Despite increases in HF mortality after 2011, rates of HF mortality were lower among KPSC compared to California and the US. Given the mortality burden of HF at older age, there is a need to improve HF prevention, treatment and management efforts earlier in life.


2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Matthew T. Mefford ◽  
Zimin Zhuang ◽  
Zhi Liang ◽  
Wansu Chen ◽  
Sandra Y. Koyama ◽  
...  

Abstract Background In recent years, decreases in mortality rates attributable to cardiovascular diseases have slowed but mortality attributable to heart failure (HF) has increased. Methods Between 2001–2017, trends in age-adjusted mortality with HF as an underlying cause for Kaiser Permanente Southern California (KPSC) members were derived through linkage with state death files and compared with trends among California residents and the US. Average annual percent change (AAPC) and 95% confidence intervals (CI) were calculated using Joinpoint regression. Analyses were repeated examining HF as a contributing cause of death. Results In KPSC, the age-adjusted HF mortality rates were comparable to California but lower than the US, increasing from 23.9 per 100,000 person-years (PY) in 2001 to 44.7 per 100,000 PY in 2017, representing an AAPC of 1.3% (95% CI 0.0%, 2.6%). HF mortality also increased in California from 33.9 to 46.5 per 100,000 PY (AAPC 1.5%, 95% CI 0.3%, 2.7%), while remaining unchanged in the US at 57.9 per 100,000 PY in 2001 and 2017 (AAPC 0.0%, 95% CI − 0.5%, 0.5%). Trends among KPSC members ≥ 65 years old were similar to the overall population, while trends among members 45–64 years old were flat between 2001–2017. Small changes in mortality with HF as a contributing cause were observed in KPSC members between 2001 and 2017, which differed from California and the US. Conclusion Lower rates of HF mortality were observed in KPSC compared to the US. Given the aging of the US population and increasing prevalence of HF, it will be important to examine individual and care-related factors driving susceptibility to HF mortality.


Cells ◽  
2021 ◽  
Vol 10 (9) ◽  
pp. 2430
Author(s):  
Ning-I Yang ◽  
Chi-Hsiao Yeh ◽  
Tsung-Hsien Tsai ◽  
Yi-Ju Chou ◽  
Paul Wei-Che Hsu ◽  
...  

Heart failure (HF) is a global pandemic public health burden affecting one in five of the general population in their lifetime. For high-risk individuals, early detection and prediction of HF progression reduces hospitalizations, reduces mortality, improves the individual’s quality of life, and reduces associated medical costs. In using an artificial intelligence (AI)-assisted genome-wide association study of a single nucleotide polymorphism (SNP) database from 117 asymptomatic high-risk individuals, we identified a SNP signature composed of 13 SNPs. These were annotated and mapped into six protein-coding genes (GAD2, APP, RASGEF1C, MACROD2, DMD, and DOCK1), a pseudogene (PGAM1P5), and various non-coding RNA genes (LINC01968, LINC00687, LOC105372209, LOC101928047, LOC105372208, and LOC105371356). The SNP signature was found to have a good performance when predicting HF progression, namely with an accuracy rate of 0.857 and an area under the curve of 0.912. Intriguingly, analysis of the protein connectivity map revealed that DMD, RASGEF1C, MACROD2, DOCK1, and PGAM1P5 appear to form a protein interaction network in the heart. This suggests that, together, they may contribute to the pathogenesis of HF. Our findings demonstrate that a combination of AI-assisted identifications of SNP signatures and clinical parameters are able to effectively identify asymptomatic high-risk subjects that are predisposed to HF.


Sign in / Sign up

Export Citation Format

Share Document