Background

Heart failure (HF) is a major and growing health problem worldwide;1,2 it is a significant burden for patients and the health system in both high-, middle- and low-income countries.3 Approximately 1-2% of the adult population in developed countries suffer from HF and prevalence increases with age.4 The overall prevalence of HF is expected to increase in the future due to several factors including an aging population, and the high success levels in postponing coronary events and prolonging survival in patients suffering from coronary events.5

Heart Failure is a complex clinical syndrome in which patients have typical symptoms (e.g. shortness of breath, fatigue/limited exercise tolerance and ankle swelling) and signs (e.g. elevated jugular venous pressure, pulmonary crackles and displaced apex beat) resulting from any abnormality of the structure or function of the heart.1

Patients with HF are categorized according to their left ventricle ejection fraction (EF) (i.e. the ratio of the volume of blood the heart empties during systole to the volume of blood in the heart at the end of diastole, expressed in percentages).2,4 The majority of patients with HF have reduced EF (<=40%), but some have a midrange EF (41-49%) or preserved EF >=50%.2,4 The categorization of HF is important due to different underlying aetiologies, patient demographics, comorbidity statuses, prognoses, and responses to therapies.1,2 Where coronary artery disease is the main cause of HF with reduced EF, hypertension and atrial fibrillation are the main causes of HF with preserved EF.1,4 Furthermore, patients with HF with preserved EF are older and more often female and obese than those with HF with reduced EF.4

Heart failure is a serious condition associated with frequent hospitalizations, significant reduction in life expectancy and quality of life, and is one of the leading causes of hospital admission and readmission in elderly people.6 Despite declining mortality from HF within the last two decades, readmission rates after hospitalization for HF remain high - and even appear to have increased in the same period.7,8 The reported readmission rates vary across HF populations, settings and follow-up periods.9-15 Studies have reported 30-day, 90-day and one-year readmission rates after discharge for HF between 18% to 32%,10-12,16 between 21% to 47% within 90 days17-19 and 50% to 67% with one year,9,11 respectively. About 50% to 60% of the 30-day all-cause readmission after hospitalization with HF has been found to occur within seven and 15 days after discharge, respectively.15,16 It is important to be aware that the risk of readmission at different times may be related to patient, provider, system as well as post-discharge factors6,20 measured at specific time points during the illness trajectory.

The risk for readmission after HF hospitalization is highest immediately following discharge and just before death.21 The first two to three months after hospital discharge for HF is a particular vulnerable phase for patients.9,11,22 Despite the marked improvement in patients' HF symptoms during hospitalization, they are discharged with substantial sub-clinical haemodynamic abnormalities.11 Further, management of the health care situation, medical treatment and self-care activities after discharge are challenging for many patients with HF. At the same time, patients are recovering from an acute illness or facing a new life situation.22,23 For patients with newly diagnosed HF, becoming a person with HF is a process involving searching for meaning, taking on a new identity and appropriate role behaviors.24

Early readmission after hospital discharge is associated with poorer long-term outcomes. In addition, each HF readmission is associated with incrementally poorer outcomes related to mortality and quality of life.25 However, frequent hospital admissions also put a great strain on patients and their relatives.10 Readmissions are costly6,11,26 and readmission rates after HF hospitalization are used by health service systems as an indicator of the quality and efficiency of care.16

Although various transitional care interventions and disease management programs have been shown to be effective in reducing rates of readmission in patients with HF, it is important for health professionals to identify which patients may have the highest risk of readmission19 in order to provide individual care in accordance with the patient's risk. Previously, systematic reviews6,19,27-30 on risk factors for readmission in patients with HF have identified and synthesized available literature on individual risk predictors6,19,27-30 or risk prediction models27,31,32 for hospital readmission. The systematic reviews often describe or synthesize risk factors in heterogeneous study populations for all-cause readmission and/or HF-specific readmission with different follow-up times, without any distinction between patients with reduced EF or preserved EF.2,4 For example, a recent meta-analysis of all-cause hospital readmission or all-cause mortality within 90 days from discharge19 included 69 articles, of which only 39 articles reported EF data. Meta-analyses were conducted on 32 factors such as age, gender, single or living alone, cognitive impairment, depression, diabetes, EF<40%, beta-blocker, multidisciplinary intervention, etc. Heterogeneity was moderate or substantial high for 25 out of 32 metaanalyses.

It is, however, important to understand the risk factors and their temporal impact on the rate of readmission in more homogeneous study populations with well-defined HF in order to develop effective interventions to prevent readmission.6 It is therefore crucial to understand risk factors for readmission in HF populations with HF and reduced EF, midrange EF and preserved EF, respectively, due to different underlying aetiologies, demographics, comorbidity statuses and responses to therapy.1 To cover all patient types in one systematic review is not realistic. Therefore, this review will be limited to patients with reduced EF because the definition of HF with reduced EF is clear, and treatment of patients with HF with reduced EF is evidence-based. No treatments for HF with preserved EF have yet been shown to reduce mortality or morbidity in this patient group.1 Further, compared with patients with HF and reduced EF, the literature in patients with HF and preserved EF is less comprehensive.

A systematic literature review identifying and synthesizing available research literature on risk factors for hospital readmission in patients with HF and reduced EF assessed at various times after discharge could contribute important information and evidence for clinicians. A search in October 2016 in the JBI Database of Systematic Reviews and Implementation Reports, Cochrane Library, MEDLINE, PROSPERO and DARE databases did not reveal a recent or forthcoming systemic review on the same topic.

Better understanding of the risk predictors for readmission and the trajectory of risk predictors during the first year after hospitalization for HF with reduced EF would be valuable in communication, shared decision-making and goal setting for recovery and treatment between healthcare professionals and patients and their relatives. Furthermore, better understanding of the risks for readmission (all-cause and HF) at various time points related to HF hospitalization can help health professionals to differentiate and personalize care interventions to prevent hospitalization due to worsening of HF and thereby the burden on the patient and the healthcare system.

Inclusion criteria

Types of participants

This review will consider studies that include patients with HF with reduced EF i.e. EF<=40%,2,4 aged 18 years or older, discharged after a hospitalization related to HF. The diagnosis of HF has to be confirmed by echocardiography.

Types of exposure

The exposure of interest is risk factors for prediction of hospital readmission. This review will consider studies that measure/estimate the association between any risk factor (e.g. age, gender, cohabiting status, smoking, depression, beta-blocker, multidisciplinary intervention, single or living alone, comorbidities, etc.) and risk of readmission.

Outcomes

In order to understand risk factors and their temporal impact on rate of readmission, the outcomes of interest in this review are: i) all-cause hospital readmission, and ii) HF hospital readmission within seven, 15, 30, 60, 90, 180 and 365 days of discharge from an index HF hospitalization. Only studies that report readmission for all-causes and/or for HF as a primary outcome, secondary outcome or part of a composite outcome within the above time periods will be included. Where relevant, study authors will be contacted for further details or information of the study. Readmission is defined as an acute and unplanned admission after hospitalization for HF. Heart failure hospitalization is defined as a hospital admission due to new onset or worsening of symptoms and/or signs of HF.1

Studies reporting results based on measures of risk such as frequencies, rates, medians, percentiles or relative measures such as relative risk (RR), odds ratios (OR) or hazard ratio (HR) will be included.

Types of studies

This review will consider studies with experimental and observational study designs, including prospective and retrospective cohort studies, case-control studies, cross-sectional studies, cohort studies nested in experimental studies evaluating risk/risk factors for seven-, 15-, 30-, 60-, 90-, 180- and/or 365-day readmission for inclusion. Case series, individual case reports, pediatric studies, abstracts, conference presentations/abstracts and publications without primary data or quantitative outcomes will not be included.

Search strategy

The search strategy aims to identify both published and unpublished studies.

A three-step search strategy will be applied. An initial limited search in PubMed and CINAHL has been undertaken followed by analysis of the text words contained in the title and abstract, and of the index terms used to describe the article. A second search using all identified keywords and index terms will be undertaken across the included databases and in the order described below. References from the search in each database will be imported directly into RefWorks/Endnote (citation software). Thirdly, the reference lists of studies included and systematic reviews identified in the search will be hand-searched for further eligible studies.

Studies published in English, Swedish, Norwegian and Danish languages will be considered for inclusion.

There has been a substantial development in HF treatment and care in the last 25 years.33 Angiotensin-converting enzyme (ACE) inhibitors/angiotensin receptor blockers, beta-blockers and aldosterone antagonists have been the cornerstone of the pharmacological treatment for HF with reduced EF since 2001 with the revision of the European Society of Cardiology (ESC) and American College of Cardiology/American Heart Association (ACC/AHA) guidelines for evaluation and management of chronic heart failure.34,35 Further, most of the development of heart failure management programs including heart failure clinics and outpatient management of HF patients took place in the second half of the 1990 s.36 Therefore to include data applicable to the present time of HF care, databases will be searched from 1 January 2000 to the present.

The databases to be searched for published studies include:

PubMed (contains MEDLINE)

Embase

CINAHL

Cochrane CENTRAL

PsycINFO.

The search for unpublished studies (grey literature, i.e. publications such as theses, papers and reports produced by agencies (such as government, academic, non-profit organizations, business and industry)37 that are not published by commercial publishers) will include:

OpenGrey

Mednar

DART-Europe

ProQuest Dissertations and Theses Database (PQDT)

The Grey-literature Report in Public Health

Unpublished studies such as theses, dissertations, reports and papers will be considered in this review. Conference proceedings/abstracts, presentations and government documents will be excluded.

Critical search terms are identified as:

Population: Heart failure

Exposure: Risk*, factor*, predict* and characteristic*

Outcome: Patient readmission, readmis*, re-admis*, re-admit*, readmit* rehospitali*and re-hospitali*

The search strategies have been developed in cooperation with research librarians from Aarhus University, aiming to identify the largest portion of relevant studies and to ensure replication of the search process.

Assessment of methodological quality

Initially, identified studies will be assessed independently by the primary and secondary reviewer for relevance based on the title or abstract. Thereafter, studies selected for retrieval will be assessed independently by the primary and secondary reviewer for methodological validity prior to inclusion into the review using standardized critical appraisal instruments from the Joanna Briggs Institute System for the Unified Management, Assessment and Review of Information (JBI-SUMARI); SUMARI critical appraisal tools for case control studies, cohort studies, analytical cross sectional studies, quasi-experimental studies, prognostic studies (QUIPS - Quality in Prognostic Studies) and randomized control trials (Appendix I). Any disagreements between the reviewers at each stage will be resolved through discussion, or by involving a third reviewer.

The number of papers included and excluded at each stage and the main reason for exclusion will be recorded (flowchart). All searches, decisions and steps will be documented and archived by the primary reviewer.

Data extraction

Quantitative data will be extracted independently by the primary and secondary reviewer from study papers included in the review using the standardized data extraction tool from JBI-SUMARI (Appendix II). The data extracted will include specific details about study methods, populations, risk factors/exposures, outcomes and results of significance to the review question and specific objectives. Authors will be contacted regarding any missing information. Disagreements between reviewers regarding data extraction will be resolved through discussion among all authors or by involving a third reviewer.

Data synthesis

Quantitative data will, where possible, be pooled in statistical meta-analysis using JBI-SUMARI. All results will be subject to double data entry. Effect sizes expressed as relative risk (RR) or odds ratios (OR) for categorical data and weighted mean differences for continuous data and their 95% confidence intervals will be calculated for analysis. A random effects model will be used and heterogeneity will be assessed by forest plot and statistically using the standard Chi-square. When appropriate heterogeneity will also be explored using sensitivity analysis, subgroup analysis (e.g. based on the different study designs), meta-regression and funnel plots. When statistical pooling is not possible the findings will be presented in a narrative form including tables and figures to aid in data presentation where appropriate.

Acknowledgements

The literature search strategy has been developed in cooperation with research librarians, Aarhus University Library, Health Sciences.

Appendix I: Critical appraisal tools

SUMARI critical appraisal tools

Quality in Prognostic Studies (QUIPS) tool

Appendix II: Data extraction instruments

SUMARI data extraction instrument

References