Roles of Nonclinical and Clinical Data in Prediction of 30-Day Rehospitalization or Death Among Heart Failure Patients

doi:10.1016/j.cardfail.2015.02.002

Journal of Cardiac Failure

Volume 21, Issue 5, May 2015, Pages 374-381

https://doi.org/10.1016/j.cardfail.2015.02.002 Get rights and content

Highlights

•
Both clinical and nonclinical data may predict HF readmissions or death.
•
Poor clinical features and socioeconomic disadvantages predict higher risk.
•
Clinical data are stronger predictors than nonclinical data.
•
Combining both data sources substantially increases predictive power.

Abstract

Background

Selecting heart failure (HF) patients for intensive management to reduce readmissions requires effective targeting. However, available prediction scores are only modestly effective. We sought to develop a prediction score for 30-day all-cause rehospitalization or death in HF with the use of nonclinical and clinical data.

Methods and Results

This statewide data linkage included all patients who survived their 1st HF admission (with either reduced or preserved ejection fraction) to a Tasmanian public hospital during 2009–2012. Nonclinical data (n = 1,537; 49.5% men, median age 80 y) included administrative, socioeconomic, and geomapping data. Clinical data before discharge were available from 977 patients. Prediction models were developed and internally and externally validated. Within 30 days of discharge, 390 patients (25.4%) died or were rehospitalized. The nonclinical model (length of hospital stay, age, living alone, discharge during winter, remoteness index, comorbidities, and sex) had fair discrimination (C-statistic 0.66 [95% confidence interval (CI) 0.63–0.69]). Clinical data (blood urea nitrogen, New York Heart Association functional class, albumin, heart rate, respiratory rate, diuretic use, angiotensin-converting enzyme inhibitor use, arrhythmia, and troponin) provided better discrimination (C-statistic 0.72 [95% CI 0.68–0.76]). Combining both data sources best predicted 30-day rehospitalization or death (C-statistic 0.76 [95% CI 0.72–0.80]).

Conclusions

Clinical data are stronger predictors than nonclinical data, but combining both best predicts 30-day rehospitalization or death among HF patients.

Section snippets

Study Population

This statewide data linkage included all 1,727 HF patients (with either reduced or preserved ejection fraction) who had their 1st admission to a public hospital in Tasmania with HF from July 2009 to June 2012. These patients were identified by their coded diagnoses (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD]: 402.x1, 404.x1, 404.x3, 428.x, and 428.xx). Because this study aimed to evaluate the post-discharge risk of readmission or death, we excluded 190

Baseline Characteristics

There were 1,537 HF patients (89% of all 1st-time HF patients during 2009–2012 in Tasmania) who survived their 1st admission. Nonclinical data were available for all patients, and clinical data were available for 977 patients.

Table 1 presents the nonclinical characteristics of patients at baseline. Their median age was 80 years and 49.5% were men. Nearly one-half were living alone. Although nearly one-half were classified as living outside of major population centers, the median distance to the

Discussion

Short-term outcomes for HF are important targets for quality improvement and are a central focus for patients, health care providers, taxpayers, and policy makers. The risk of readmission or death after a hospitalization with HF is high even in the short term. In the present study, ∼1 in 4 patients who survived their 1st admission with HF was readmitted and/or dead within 30 days of discharge. This finding was similar to those from other studies.2, 5, 26, 27 Targeting more intensive disease

Study Limitations

Our participants were mostly white, so the generalizability of our findings to other ethnic groups is uncertain. The restriction of analysis to patients admitted to public hospitals may have underestimated the roles of socioeconomic and geomapping factors in this study because patients without health insurance are more likely to be socioeconomically disadvantaged. By using log-binomial regression analysis rather than time-to-event analysis, we could not identify HF patients with high risk of

Conclusion

The present findings suggest that clinical data are stronger than nonclinical data in predicting 30-day readmission or death among HF patients. However, combining the 2 sources of data best predicts short-term adverse outcomes among HF patients.

Disclosures

None.

References (40)

Z.J. Eapen et al.
Validated, electronic health record deployable prediction models for assessing patient risk of 30-day rehospitalization and mortality in older heart failure patients
JACC Heart Fail
(2013)
J.E. Blair et al.
Continental differences in clinical characteristics, management, and outcomes in patients hospitalized with worsening heart failure results from the EVEREST (Efficacy of Vasopressin Antagonism in Heart Failure: Outcome Study With Tolvaptan) program
J Am Coll Cardiol
(2008)
S. Setoguchi et al.
Repeated hospitalizations predict mortality in the community population with heart failure
Am Heart J
(2007)
M.E. Charlson et al.
A new method of classifying prognostic comorbidity in longitudinal studies: development and validation
J Chronic Dis
(1987)
H.M. Krumholz et al.
Predictors of readmission among elderly survivors of admission with heart failure
Am Heart J
(2000)
S. Stewart et al.
Heart failure in a cold climate. Seasonal variation in heart failure-related morbidity and mortality
J Am Coll Cardiol
(2002)
A.G. Barnett et al.
The seasonality in heart failure deaths and total cardiovascular deaths
Aust N Z J Public Health
(2008)
P. Thavendiranathan et al.
Prediction of 30-day heart failure-specific readmission risk by echocardiographic parameters
Am J Cardiol
(2014)
W. Rosamond et al.
Heart disease and stroke statistics—2008 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee
Circulation
(2008)
S.F. Jencks et al.
Rehospitalizations among patients in the Medicare fee-for-service program
N Engl J Med
(2009)

J. Chen et al.

National and regional trends in heart failure hospitalization and mortality rates for Medicare beneficiaries, 1998–2008

JAMA

(2011)

H. Bueno et al.

Trends in length of stay and short-term outcomes among Medicare patients hospitalized for heart failure, 1993–2006

JAMA

(2010)

R.P. Vivo et al.

Short- and long-term rehospitalization and mortality for heart failure in 4 racial/ethnic populations

J Am Heart Assoc

(2014)

A.M. Ryan et al.

Medicare's public reporting initiative on hospital quality had modest or no impact on mortality from three key conditions

Health Aff (Millwood)

(2012)

An Act Entitled The Patient Protection and Affordable Care Act. Public Law 111–48. The United States...

K. Page et al.

A systematic approach to chronic heart failure care: a consensus statement

Med J Aust

(2014)

R.D. Kociol et al.

National survey of hospital strategies to reduce heart failure readmissions: findings from the Get With the Guidelines–Heart Failure registry

Circ Heart Fail

(2012)

H. Seow et al.

Isolation of health services research from practice and policy: the example of chronic heart failure management

J Am Geriatr Soc

(2006)

J.S. Ross et al.

Statistical models and patient predictors of readmission for heart failure: a systematic review

Arch Intern Med

(2008)

N. Allaudeen et al.

Inability of providers to predict unplanned readmissions

J Gen Intern Med

(2011)

Cited by (48)

Global Comparison of Readmission Rates for Patients With Heart Failure
2023, Journal of the American College of Cardiology
Heart failure (HF) readmission rates are low in some jurisdictions. However, international comparisons are lacking and could serve as a foundation for identifying regional patient management strategies that could be shared to improve outcomes.
This study sought to summarize 30-day and 1-year all-cause readmission and mortality rates of hospitalized HF patients across countries and to explore potential differences in rates globally.
We performed a systematic review and meta-analysis using MEDLINE, Embase, and CENTRAL for observational reports on hospitalized adult HF patients at risk for readmission or mortality published between January 2010 and March 2021. We conducted a meta-analysis of proportions using a random-effects model, and sources of heterogeneity were evaluated with meta-regression.
In total, 24 papers reporting on 30-day and 23 papers on 1-year readmission were included. Of the 1.5 million individuals at risk, 13.2% (95% CI: 10.5%-16.1%) were readmitted within 30 days and 35.7% (95% CI: 27.1%-44.9%) within 1 year. A total of 33 papers reported on 30-day and 45 papers on 1-year mortality. Of the 1.5 million individuals hospitalized for HF, 7.6% (95% CI: 6.1%-9.3%) died within 30 days and 23.3% (95% CI: 20.8%-25.9%) died within 1 year. Substantial variation in risk across countries was unexplained by countries’ gross domestic product, proportion of gross domestic product spent on health care, and Gini coefficient.
Globally, hospitalized HF patients exhibit high rates of readmission and mortality, and the variability in readmission rates was not explained by health care expenditure, risk of mortality, or comorbidities.
Risk Prediction in Transition: MAGGIC Score Performance at Discharge and Incremental Utility of Natriuretic Peptides
2020, Journal of Cardiac Failure
Risk stratification for hospitalized patients with heart failure (HF) remains a critical need. The Meta-Analysis Global Group in Chronic Heart Failure (MAGGIC) score is a robust model derived from patients with ambulatory HF. Its validity at the time of discharge and the incremental value of natriuretic peptides (NPs) in this setting is unclear.
This was a single-center study examining a total of 4138 patients with HF from 2 groups; hospital discharge patients from administrative data (n = 2503, 60.5%) and a prospective registry of patients with ambulatory HF (n = 1635, 39.5%). The ambulatory registry patients underwent N-terminal pro–B-type NP (BNP) measurement at enrollment, and in the hospitalize discharge cohort clinical BNP levels were abstracted. The primary endpoint was all-cause mortality within 1 year. MAGGIC score performance was compared between cohorts utilizing Cox regression and calibration plots. The incremental value of NPs was assessed using calculated area under the curve and net reclassification improvement (NRI).
The hospitalized and ambulatory cohorts differed with respect to primary outcome (777 and 100 deaths, respectively), sex (52.1% vs 41.7% female) and race (35% vs 49.5% African American). The MAGGIC score showed poor discrimination of mortality risk in the hospital discharge (C statistic: 0.668, hazard ratio [HR]: 1.1 per point, 95% confidence interval [CI]: 0.652, 0.684) but fair discrimination in the ambulatory cohorts (C statistic: 0.784, HR: 1.16 per point, 95% CI: 0.74, 0.83), respectively, a difference that was statistically significant (P = .001 for C statistic, 0.002 for HR). Calibration assessment indicated that the slope and intercept (of MAGGIC-predicted to observed mortality) did not statistically differ from ideal in either cohort and did not differ between the cohorts (all P > .1). NP levels did not significantly improve prediction in the hospitalized cohort (P = .127) but did in the ambulatory cohort (C statistic: 0.784 [95% CI: 0.74, 0.83] vs 0.82 [95% CI: 0.78, 0.85]; P = .018) with a favorable NRI of 0.354 (95% CI: 0.202–0.469; P = .002).
The MAGGIC score showed poor discrimination when used in patients with HF at hospital discharge, which was inferior to its performance in patients with ambulatory HF. Discrimination within the hospital discharge group was not improved by including hospital NP levels.
High-Sensitivity Troponin T and Incident Heart Failure in Older Men: British Regional Heart Study
2019, Journal of Cardiac Failure
Citation Excerpt :
For example, in chronic HF patients in the RED-HF (Reduction of Events by Darbepoetin Alfa in Heart Failure) trial, among a range of cardiac biomarkers only hsTnT added to NT-proBNP in the prediction of adverse outcome.26 In the MARATHON (Multicentre Australian Risk Algorithm to Predict Heart Failure Readmission) study, comprising patients hospitalized with preserved or reduced ejection fraction HF, elevated troponin I predicted 30-day readmission to hospital or death, both individually and as part of a panel of variables.27 Therefore, elevated troponin appears to be a predictor of adverse outcomes in the presymptomatic, chronic, and acute phases of HF.
The aim of this work was to study the association of high-sensitivity troponin T (hsTnT) with incident heart failure (HF), and implications for its use in prediction models.
In the British Regional Heart Study, 3852 men aged 60–79years without baseline HF (3165 without baseline chronic heart disease) were followed for a median of 12.6years, during which 295 incident cases of HF occurred (7.7%). A 1-SD increase in log-transformed hsTnT was associated with a higher risk of incident HF after adjusting for classic risk factors (hazard ratio [HR] 1.58, 95% confidence interval [CI] 1.42–1.77) and after additional adjustment for N-terminal pro–B-type natriuretic peptide (NT-proBNP; HR 1.34, 95% CI 1.19–1.52). The strength of the association between hsTnT and incident HF did not differ by strata of other risk factors. An hsTnT concentration of <5ng/L had a sensitivity of 99.7% (95% CI 98.1%–99.9%) and a specificity of 3.4% (95% CI 2.8%–4.0%). A risk-prediction model including classic risk factors and NT-proBNP yielded a C-index of 0.791, but addition of hsTnT did not further improve prediction (P = .28).
Elevated hsTnT is consistently associated with risk of HF in older men. HF occurred rarely over 12years when baseline hsTnT was below the limit of detection. hsTnT measurement, however, does not improve HF prediction in a model already containing NT-proBNP.
Mortality and Readmission Following Hospitalisation for Heart Failure in Australia: A Systematic Review and Meta-Analysis
2018, Heart Lung and Circulation
Citation Excerpt :
Again, there was significant heterogeneity between studies (I2 = 98%). Five studies reported all-cause readmission [16,27,33,35,36], while three studies reported unplanned readmission only [28,30,34]. There was an increase of 1-year readmission rate with time (based on year of publication) by 2.3% (95% CI is 1.1–3.6%, p = 0.014).
Heart failure (HF) is a common, costly condition with an increasing burden on Australian health care system resources. Knowledge of the burden of HF on patients and on the health system is important for resource allocation. This study is the first systematic review to estimate the mortality and readmission rates after hospitalisation for HF in the Australian population.
We searched for studies of HF hospitalisation in Australia published between January 1990 and May 2016, using a systematic search of PubMed, Medline, Scopus, Web of Science, EMBASE and Cochrane Library databases. Studies reporting 30-day and/or 1-year outcomes for mortality or readmission following hospitalisation were eligible and included in this study.
Out of 2889 articles matching the initial search criteria, a total of 13 studies representing 67,255 patients were included in the final analysis. The pooled mean age of heart failure patients was 76.3 years and 51% were male (n = 34,271). The pooled estimated 30-day and 1-year all-cause mortality were 8% and 25% respectively. The pooled estimated 30-day and 1-year all-cause readmission rates were 20% and 56% respectively. There is a high prevalence of comorbidities in heart failure patients. There were limited data on readmission and mortality in rural patients and Indigenous people.
Heart failure hospitalisations in Australia are followed by substantial readmission and mortality rates.
The impact of perceived and objective social isolation on hospital readmission in patients with heart failure: A systematic review and meta-analysis of observational studies
2019, General Hospital Psychiatry
Several psychosocial risk factors have been identified that increase the rate of readmission in heart failure (HF) patients. However, the impact of social isolation (SI) on the rate of readmission is unclear. Therefore, the current review focused on the impact of SI on readmission rates of patients with HF.
A Medline-based strategy was applied to search PubMed, SCOPUS, Cochrane library, ProQuest, and Embase from inception until November 15, 2018. We performed a meta-analysis and pooled results using random effects model. The primary outcome was the odds ratio of readmission in HF patients suffering from SI. We examined the impact of both perceived and objective SI on readmission rates. We also examined the differences in readmission rates between these concepts. The secondary outcomes were the incidence of readmission and the prevalence of SI.
From 3326 titles, 13 studies (n = 6468 participants) were eligible. The mean follow-up period was 13 months. The cumulative incidence for HF-related hospital readmission was 35.47% (95% CI: 34.29–36.67). The pooled prevalence ratio (PR, (95% CI)) was 37.31% (36.14–38.49), 31.51% (30.36–32.68), 32.82% (29.90–35.88), and 39.57% (37.73–41.45) for SI, living alone, lack of social support, and poor social network, respectively. SI was associated with a 55% greater risk of hospital readmission in patients with HF (OR = 1.55; 95% CI = 1.39–1.73; p < .001). Our analysis did not show a significant difference in the rate of hospital readmission between perceived and objective SI.
SI is prevalent in patients with HF and seems to be consistently linked to hospital readmission in HF patients, regardless of how it is measured. Therefore, it is necessary to develop interventions to reduce the burden of SI in patients with HF.
Impact of mental disorders on unplanned readmissions for congestive heart failure patients: a population-level study
2024, ESC Heart Failure

View all citing articles on Scopus

: Funding: Supported in part by a partnership grant (GRT1059738) from the National Health and Medical Research Council, Canberra, the National Heart Foundation of Australia, Tasmania Medicare Local, and the Tasmanian Department of Health and Human Services, Hobart, Australia.

: See page 380 for disclosure information.

View full text

Clinical InvestigationRoles of Nonclinical and Clinical Data in Prediction of 30-Day Rehospitalization or Death Among Heart Failure Patients

Highlights

Abstract

Background

Methods and Results

Conclusions

Section snippets

Study Population

Baseline Characteristics

Discussion

Study Limitations

Conclusion

Disclosures

JACC Heart Fail

J Am Coll Cardiol

Am Heart J

J Chronic Dis

Am Heart J

J Am Coll Cardiol

Aust N Z J Public Health

Am J Cardiol

Heart disease and stroke statistics—2008 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee

Circulation

Rehospitalizations among patients in the Medicare fee-for-service program

N Engl J Med

National and regional trends in heart failure hospitalization and mortality rates for Medicare beneficiaries, 1998–2008

JAMA

Trends in length of stay and short-term outcomes among Medicare patients hospitalized for heart failure, 1993–2006

JAMA

Short- and long-term rehospitalization and mortality for heart failure in 4 racial/ethnic populations

J Am Heart Assoc

Medicare's public reporting initiative on hospital quality had modest or no impact on mortality from three key conditions

Health Aff (Millwood)

A systematic approach to chronic heart failure care: a consensus statement

Med J Aust

National survey of hospital strategies to reduce heart failure readmissions: findings from the Get With the Guidelines–Heart Failure registry

Circ Heart Fail

Isolation of health services research from practice and policy: the example of chronic heart failure management

J Am Geriatr Soc

Statistical models and patient predictors of readmission for heart failure: a systematic review

Arch Intern Med

Inability of providers to predict unplanned readmissions

J Gen Intern Med

Clinical Investigation
Roles of Nonclinical and Clinical Data in Prediction of 30-Day Rehospitalization or Death Among Heart Failure Patients