Hospital readmission performance and patterns of readmission: retrospective cohort study of Medicare admissions

A closer look at acute heart failure: Putting Portuguese and European data into perspective

INTRODUCTION AND OBJECTIVES

Acute heart failure (AHF) is a heterogeneous clinical syndrome requiring urgent therapy. The prognosis is poor after the index hospitalization, with a high risk for rehospitalization and early death. The costs of managing AHF are thus increasing rapidly. A literature review was performed to gather and compare data on prevalence and treatment and to identify gaps in AHF management, based on European and Portuguese studies.

METHODS

A literature search from 1995 to 2014 was conducted in selected databases (BIOSIS Previews, EMBASE and Ovid MEDLINE).

RESULTS AND DISCUSSION

Seven Portuguese and nine European studies were analyzed. The mean age of AHF patients was ≥65 years and 30-50% were women. Coronary artery disease (42.3% vs. 61.9%) and hypertension (53.3% vs. 76.7%) were identified as primary etiologies in Europe and in Portugal. Similar proportions of heart failure with preserved ejection fraction were found in the Portuguese (19.9-44.7%) and European (32.8-39.1%) studies. Overall, all-cause mortality rates were comparable (six months: 9.3-25.5% vs. 13.5-27.4%; one year: 15.9-31% vs. 17.4-46.5%), as was in-hospital mortality (5.5-14% vs. 3.8-12%) in Portuguese and European studies, respectively. Length of stay was comparable. The studies were performed in very different hospital settings and data on treatment were scarce.

CONCLUSIONS

Gaps were identified in treatment and clinical pathways of patients with AHF. Based on the results of this review, collection and investigation of data on the disease and treatment solutions, training in disease management, and improved organization of healthcare should be the subject of further investment.

Post-Hospital Syndrome: A New Challenge in Cardiovascular Practice

The image of the hospital representing the modern medicine and its diagnostic and therapeutic advances becomes more evident in the face of an aging population and patients with multiple comorbidities requiring highly complex care. However, recent studies have shown a growing number of hospital readmissions within 30 days after discharge. The post-hospital syndrome is a new clinical entity associated with multiple vulnerabilities that contribute to hospital readmissions. During hospitalization, the patient is exposed to different stressors of physical, environmental, and psychosocial natures that trigger pathophysiological and multisystemic responses and increase the risk of complications after hospital discharge. Patients with a cardiac disease have high rates of readmission within 30 days. Therefore, it is important for cardiologists to recognize the post-hospital syndrome since it may impact their daily practice. This review aims at discussing the current scientific evidence regarding predictors and stressors involved in the post-hospital syndrome and the measures that are currently being taken to minimize their effects.

Multimorbidity in Older Adults with Heart Failure

Multimorbidity is common among older adults with heart failure and creates diagnostic and management challenges. Diagnosis of heart failure may be difficult, as many conditions commonly found in older persons produce dyspnea, exercise intolerance, fatigue, and weakness; no singular pathognomonic finding or diagnostic test differentiates them from one another. Treatment may also be complicated, as multimorbidity creates high potential for drug-disease and drug-drug interactions in settings of polypharmacy. The authors suggest that management of multimorbid older persons with heart failure be patient, rather than disease-focused, to best meet patients' unique health goals and minimize risk from excessive or poorly-coordinated treatments.

Characteristics and temporal pattern of the readmissions of patients with multiple hospital admissions in the medical departments of a general hospital

INTRODUCTION

Patients with multiple hospital admissions represent a small percentage of total hospitalizations but result in a considerable proportion of the healthcare expenditure. There are no studies that have analyzed their long-term clinical evolution.

OBJECTIVES

To study the characteristics, temporal patterns of readmissions and clinical evolution of patients with multiple hospital admission in the long term.

METHODS

A retrospective study was conducted of all hospital admissions in the medical area of the Hospital of Lugo (Spain) between January 1, 2000 and December 31, 2012, based on data from the center's minimum basic data set.

RESULTS

A total of 139,249 hospital admissions for 62,515 patients were studied. Six hospital admissions were recorded for 6.4% of the patients. The overall mortality rate was 16% (9982 patients). The readmissions rate steadily increased with each new admission, from 48% after the first event to 74.6% after the fifth. The rate of hospital readmission before 30days increased from 18.3% in the second admission to 36.3% in the sixth. The number of chronic diseases increased from 3.1 (SD, 2) in the first hospital admission up to 4.9 (2.8) in the sixth. The Department of Internal Medicine treated a third of all hospital admissions. In the sixth hospitalization, conditions associated with admission in Internal Medicine were CIRS score, age, heart failure, COPD, dementia, diabetes, atrial fibrillation and anemia.

CONCLUSIONS

Patients with multiple hospital admissions are complex patients whose temporal pattern of readmissions changes with time, such that each hospital admission constitutes a factor facilitating the next.

The Risk Factors of Readmission in Postoperative Gynecologic Oncology Patients at a Single Institution

Introduction

Hospital readmission rates are an important measure of quality care and have recently been tied to reimbursement. This study seeks to identify the risk factors for postoperative readmission in patients treated by a gynecologic oncology service.

Methods

A 7-year retrospective review (2007–2013) of all patients operated on by the University of Virginia gynecologic oncology service who were readmitted within 30 days of discharge was performed. Abstracted data included demographics, dates of surgery, operative details, cancer history, and relevant medical history. The readmitted patients (n = 166) were compared with randomly selected controls (n = 168) from the same service in a matching time frame and analyzed using univariate and multivariate models.

Results

In the study period, 2993 operations were performed. One hundred sixty-six unique patients (5.5%) were readmitted within 30 days of discharge from their operative procedure. On multivariate analysis, the factors that were associated with a higher risk of readmission were a history of psychiatric disease, postoperative complication, type of insurance, surgical modality, and lysis of adhesions at the time of surgery. The most common readmission diagnoses were infection (44%), nausea/vomiting (28%), thrombosis (6%), bowel leak (4%), and bleeding (4%).

Conclusions

Postoperative readmissions are a common problem and are increasingly important as a measure of quality. Although patients were generally admitted for infections or gastrointestinal complaints, we also found that individual factors such as mental health and socioeconomic status also contributed. Our data suggest that we can preoperatively identify high-risk individuals for whom extra resources can be directed postoperatively to avoid unnecessary readmissions.

Contribution of recurrent admissions in children and young people to emergency hospital admissions: retrospective cohort analysis of hospital episode statistics

OBJECTIVE

To examine the contribution of recurrent admissions to the high rate of emergency admissions among children and young people (CYP) in England, and to what extent readmissions are accounted for by patients with chronic conditions.

DESIGN

All hospital admissions to the National Health Service (NHS) in England using hospital episode statistics (HES) from 2009 to 2011 for CYP aged 0-24 years. We followed CYP for 2 years from discharge of their first emergency admission in 2009. We determined the number of subsequent emergency admissions, time to next admission, length of stay and the proportion of injury and chronic condition admissions measured by diagnostic codes in all following admissions.

RESULTS

869 895 children had an index emergency admission in 2009, resulting in a further 939 710 admissions (of which 600 322, or 64%, were emergency admissions) over the next 2 years. After discharge from the index admission, 32% of 274,986 (32%) children were readmitted within 2 years, 26% of these readmissions occurring within 30 days of discharge. Recurrent emergency admission accounted for 41% of all emergency admissions in the 2-year cohort and 66% of inpatient days. 41% of index admissions, but 76% of the recurrent emergency admissions, were in children with a chronic condition.

CONCLUSIONS

Recurrent admissions contribute substantially to total emergency admissions. They often occur soon after discharge, and disproportionately affect CYP with chronic conditions. Policies aiming to discourage readmissions should consider whether they could undermine necessary inpatient care for children with chronic conditions.

Unplanned Readmissions After Hospitalization for Severe Sepsis at Academic Medical Center-Affiliated Hospitals

OBJECTIVE

In the United States, national efforts to reduce hospital readmissions have been enacted, including the application of substantial insurance reimbursement penalties for hospitals with elevated rates. Readmissions after severe sepsis remain understudied and could possibly signify lapses in care and missed opportunities for intervention. We sought to characterize 7- and 30-day readmission rates following hospital admission for severe sepsis as well as institutional variations in readmission.

DESIGN

Retrospective analysis of 345,657 severe sepsis discharges from University HealthSystem Consortium hospitals in 2012.

SETTING

United States.

PATIENTS

We applied the commonly cited method described by Angus et al for identification of severe sepsis, including only discharges with sepsis present at admission.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

We identified unplanned, all-cause readmissions within 7 and 30 days of discharge using claims-based algorithms. Using mixed-effects logistic regression, we determined factors associated with 30-day readmission. We used risk-standardized readmission rates to assess institutional variations. Among 216,328 eligible severe sepsis discharges, there were 14,932 readmissions within 7 days (6.9%; 95% CI, 6.8-7.0) and 43,092 within 30 days (19.9%; 95% CI, 19.8-20.1). Among those readmitted within 30 days, 66.9% had an infection and 40.3% had severe sepsis at readmission. Patient severity, length of stay, and specific diagnoses were associated with increased odds of 30-day readmission. Observed institutional 7-day readmission rates ranged from 0% to 12.3%, 30-day rates from 3.6% to 29.1%, and 30-day risk-standardized readmission rates from 14.1% to 31.1%. Greater institutional volume, teaching status, trauma services, location in the Northeast, and lower ICU rates were associated with poor risk-standardized readmission rate performance.

CONCLUSIONS

Severe sepsis readmission places a substantial burden on the healthcare system, with one in 15 and one in five severe sepsis discharges readmitted within 7 and 30 days, respectively. Hospitals and clinicians should be aware of this important sequela of severe sepsis.

The Impact of Inpatient Palliative Care Consultations on 30-Day Hospital Readmissions

BACKGROUND

Inpatient palliative care consultations have been shown to reduce acute care utilization by reducing length of stay, but less is known about their impact on subsequent costs including hospital readmissions.

OBJECTIVE

The study's objective was to examine the impact of inpatient palliative care consultations on 30-day hospital readmissions to a large urban academic medical center.

METHODS

The hospital's electronic medical record system was used to identify all live discharges between August 2013 and November 2014. After adjusting for a propensity score, readmission rates were compared between palliative care and usual care groups.

RESULTS

Of the 34,541 hospitalizations included in the study, 1430 (4.1%) involved a palliative care consult. After adjusting for the propensity score, patients seen by palliative care had a lower 30-day readmission rate-adjusted odds ratio (AOR) 0.66, 0.55-0.78; p<0.001. Adjusted rates were 10.3% (95% confidence interval [CI] 8.9%-12.0%) for palliative care and 15.0% (95% CI 14.4%-15.4%) for usual care. Among all palliative care patients, consultations that involved goals of care discussions were associated with a lower readmission rate (AOR 0.36, 0.27-0.48; p<0.001), but consultations involving symptom management were not (AOR 1.05, 0.82-1.35; p=0.684).

CONCLUSIONS

Palliative care palliative care consultations facilitate goals discussions, which in turn are associated with reduced rates of 30-day readmissions.

One-year survival and resource use after critical illness: impact of organ failure and residual organ dysfunction in a cohort study in Brazil

Introduction

In this study, we evaluated the impacts of organ failure and residual dysfunction on 1-year survival and health care resource use using Intensive Care Unit (ICU) discharge as the starting point.

Methods

We conducted a historical cohort study, including all adult patients discharged alive after at least 72 h of ICU stay in a tertiary teaching hospital in Brazil. The starting point of follow-up was ICU discharge. Organ failure was defined as a value of 3 or 4 in its corresponding component of the Sequential Organ Failure Assessment score, and residual organ dysfunction was defined as a score of 1 or 2. We fit a multivariate flexible Cox model to predict 1-year survival.

Results

We analyzed 690 patients. Mortality at 1 year after discharge was 27 %. Using multivariate modeling, age, chronic obstructive pulmonary disease, cancer, organ dysfunctions and albumin at ICU discharge were the main determinants of 1-year survival. Age and organ failure were non-linearly associated with survival, and the impact of organ failure diminished over time. We conducted a subset analysis with 561 patients (81 %) discharged without organ failure within the previous 24 h of discharge, and the number of residual organs in dysfunction remained strongly associated with reduced 1-year survival. The use of health care resources among hospital survivors was substantial within 1 year: 40 % of the patients were rehospitalized, 52 % visited the emergency department, 90 % were seen at the outpatient clinic, 14 % attended rehabilitation outpatient services, 11 % were followed by the psychological or psychiatric service and 7 % used the day hospital facility. Use of health care resources up to 30 days after hospital discharge was associated with the number of organs in dysfunction at ICU discharge.

Conclusions

Organ failure was an important determinant of 1-year outcome of critically ill survivors. Nevertheless, the impact of organ failure tended to diminish over time. Resource use after critical illness was elevated among ICU survivors, and a targeted action is needed to deliver appropriate care and to reduce the late critical illness burden.

Electronic supplementary material

The online version of this article (doi:10.1186/s13054-015-0986-6) contains supplementary material, which is available to authorized users.

Comorbidity-polypharmacy score predicts readmission in older trauma patients

BACKGROUND

Hospital readmissions are considered to be a measure of quality of care, correlate with worse outcomes, and may soon lead to decreased reimbursement. The comorbidity-polypharmacy score (CPS) is the sum of the number of preinjury medications and comorbidities, and may estimate patient frailty more effectively than patient age. This study evaluates the association between CPS and readmission.

METHODS

Medical records for trauma patients ≥45 y evaluated between January 1 and December 31, 2008, at our American College of Surgeons-verified level 1 trauma center were reviewed to obtain information on demographics, injuries, preinjury comorbidities, and medications, and occurrences of readmission to our facility within 30 d of discharge. Chi-square and Kruskal-Wallis testing was used to evaluate differences between readmitted and nonreadmitted patients, with multiple logistic regression used to evaluate the contribution of independent risk factors for readmission.

RESULTS

A total of 879 patients were included; their ages ranged from 45-103 y (median 58), injury severity scores from 0-50 y (median 5), and CPS from 0-39 y (median 7). A total of 76 patients (8.6%) were readmitted to our facility within 30 d of discharge. The readmitted cohort had higher CPS (median, 9.5, range 0-32, P = 0.031) and injury severity score (median, 9, range 1-38, P = 0.045), but no difference in age (median, 59.5, range 47-99, P = 0.646). Logistic regression demonstrated independent association of higher CPS with increased risk of readmission, with each CPS point increasing readmission likelihood by 3.5% (P = 0.03).

CONCLUSIONS

CPS appears to correlate well with readmissions within 30 d. Frailty defined by CPS was a significantly stronger predictor of readmission than was patient age. Early recognition of elevated CPS may improve discharge planning and help guide interventions to decrease readmission rates in older trauma patients.

Impact of Clinical Pharmacist Interventions on 30-Day Readmission Rate in Hospitalized Patients With Acute Myocardial Infarction

Background: Limited literature exists on the positive impact of pharmacists specifically on hospital readmission of patients with acute myocardial infarction (AMI). Objective: To evaluate the overall effect of clinical pharmacist interventions on preventing hospital readmissions and improving the health of patients with AMI. Secondary objectives include identifying trends in the demographic characteristics of AMI patients, identifying potential barriers to adherence, and assessing the average time spent by a pharmacist counseling AMI patients. Methods: This prospective, nonrandomized, single-center study was approved by the institutional review board. The hospital's 30-day AMI readmission rate prior to study initiation was used as the control group. An AMI report was generated daily to identify patients admitted to the hospital diagnosed with either non-ST or ST segment elevation myocardial infarction. The clinical pharmacist then counseled the included patient prior to discharge and provided a follow-up phone call 48 hours after discharge. The primary outcome was the all-cause 30-day readmission rate for AMI patients. Results: Out of 71 patients screened, 50 patients were included in the study. Only 3 of the 50 patients included were readmitted (6.0%). The prestudy rate from October 2012 to October 2013 was 11.6%, or 58 readmissions out of 498 AMI admissions. Although the study group was much smaller in size, a 6% readmission rate is encouraging and offers potential for a future intervention. Conclusion: Clinical pharmacist services for AMI patients, including counseling, interventions, and a follow-up phone call after discharge, may benefit decreasing the 30-day AMI readmission rate; however, further studies are needed.

Quality of care factors associated with unplanned readmissions of older medical patients: a case-control study

BACKGROUND

Unplanned readmissions befall up to 25% of acutely hospitalised older patients, and many may be potentially preventable.

AIM

To assess the type and prevalence of quality of care factors associated with potentially preventable readmissions to a tertiary hospital general medicine service.

METHODS

A retrospective case-control study was undertaken of hospital records of patients 65 years or older admitted acutely between 1 January 2005 and 31 December 2010. Readmissions up to 30 days postdischarge (cases) were purposively sampled according to frequencies of primary discharge diagnoses coded during the study period. Non-readmitted patients (controls), matched according to age, sex and primary discharge diagnosis on index admission, were selected in a 1.7:1 ratio.

RESULTS

One hundred and thirteen cases and 198 controls were analysed, the former demonstrating a significantly higher comorbidity burden (mean (±standard deviation) comorbidity score 6.6 (±2.2) vs 5.6 (±2.4), P = 0.003) and a higher proportion of individuals with one or more hospitalisations over the preceding 6 months (55.7% vs 8.1%, P < 0.001). Among readmitted patients, 50 (44.3%) were associated with one or more quality factors versus 23 (11.6%) controls (P < 0.001). The most common were: failure to develop/activate an advance care plan (18, 15.9% vs 2, 1.0%; P < 0.001); suboptimal management of presenting illness (13, 11.4% vs 0, 0%; P < 0.001); inadequate assessment of functional limitations (11, 9.7% vs 0, 0%; P < 0.001); and potentially preventable complication of therapy (8, 7.1% vs 1, 0.5%, P = 0.002).

CONCLUSIONS

Quality of care factors are more common among readmitted than among non-readmitted older patients suggesting potential for remedial strategies. Such strategies may still have limited effects as older, frail patients with advanced diseases and multimorbidity will likely retain a high propensity for readmission despite optimal care.

The crucial factor of hospital readmissions: a retrospective cohort study of patients evaluated in the emergency department and admitted to the department of medicine of a general hospital in Italy

Background

Early hospital readmissions, defined as rehospitalization within 30 days from a previous discharge, represent an economic and social burden for public health management. As data about early readmission in Italy are scarce, we aimed to relate the phenomenon of 30-day readmission to factors identified at the time of emergency department (ED) visits in subjects admitted to medical wards of a general hospital in Italy.

Methods

We performed a retrospective 30-month observational study, evaluating all patients admitted to the Department of Medicine of the Hospital of Ferrara, Italy. Our study compared early and late readmission: patients were evaluated on the basis of the ED admission diagnosis and classified differently on the basis of a concordant or discordant readmission diagnosis in respect to the diagnosis of a first hospitalization.

Results

Out of 13,237 patients admitted during the study period, 3,631 (27.4%) were readmitted; of those, 656 were 30-day rehospitalizations (5% of total admissions). Early rehospitalization occurred 12 days (median) later than previous discharge. The most frequent causes of rehospitalization were cardiovascular disease (CVD) in 29.3% and pulmonary disease (PD) in 29.7% of cases. Patients admitted with the same diagnosis were younger, had lower length of stay (LOS) and higher prevalence of CVD, PD and cancer. Age, CVD and PD were independently associated with 30-day readmission with concordant diagnosis and kidney disease with 30-day rehospitalization with a discordant diagnosis.

Conclusions

Comorbid patients are at higher risk for 30-day readmission. Reduction of LOS, especially in elderly subjects, could increase early rehospitalization rates.

Summary of STARNet: Seamless Transitions and (Re)admissions Network

The Seamless Transitions and (Re)admissions Network (STARNet) met in December 2012 to synthesize ongoing hospital-to-home transition work, discuss goals, and develop a plan to centralize transition information in the future. STARNet participants consisted of experts in the field of pediatric hospital medicine quality improvement and research, and included physicians and key stakeholders from hospital groups, private payers, as well as representatives from current transition collaboratives. In this report, we (1) review the current knowledge regarding hospital-to-home transitions; (2) outline the challenges of measuring and reducing readmissions; and (3) highlight research gaps and list potential measures for transition quality. STARNet met with the support of the American Academy of Pediatrics' Quality Improvement Innovation Networks and the Section on Hospital Medicine.

Heart failure is a major contributor to hospital readmission penalties

BACKGROUND

The Hospital Readmissions Reduction Program provides incentives to hospitals to reduce early readmissions for heart failure (HF), acute myocardial infarction (AMI), and pneumonia (PNE).

METHODS AND RESULTS

To examine the contribution of each diagnosis to readmissions penalty size, data were obtained from the Center for Medicare and Medicaid Services, American Hospital Association, and United States Census Bureau including number of cases; readmissions payment adjustment factor (values <1 indicate a penalty for excess readmissions), excess readmission ratio (ERR, or ratio of adjusted predicted readmission based on comorbidities, frailty, and individual patient demographics to expected probability of readmission at an average hospital) for each diagnosis, hospital teaching status, bed number, and zip code socioeconomic status. Of 2,228 hospitals with ≥25 cases per diagnosis, 1,636 received a penalty. Univariate correlation coefficients between penalty and ERR were -0.66, -0.61, and -0.43 for HF, PNE, and AMI, respectively (all P < .001). Correlation between ERRs was greatest for PNE and HF (0.30; P < .001) and weakest for PNE and AMI (0.12; P < .001). In regression analyses, the HF ERR explained the most variance in the penalty (R(2) range 0.21-0.44).

CONCLUSION

HF ERR, not the number of cases, was related to penalty magnitude. These findings have implications for the design of hospital-based quality initiatives regarding readmissions.

Strategies to Reduce 30-Day Readmissions in Older Patients Hospitalized with Heart Failure and Acute Myocardial Infarction

Readmission within 30 days after hospital discharge for common cardiovascular conditions such as heart failure and acute myocardial infarction is extremely common among older persons. To incentivize investment in reducing preventable rehospitalizations, the United States federal government has directed increasing financial penalties to hospitals with higher-than-expected 30-day readmission rates. Uncertainty exists, however, regarding the best approaches to reducing these adverse outcomes. In this review, we summarize the literature on predictors of 30-day readmission, the utility of risk prediction models, and strategies to reduce short-term readmission after hospitalization for heart failure and acute myocardial infarction. We report that few variables have been found to consistently predict the occurrence of 30-day readmission and that risk prediction models lack strong discriminative ability. We additionally report that the literature on interventions to reduce 30-day rehospitalization has significant limitations due to heterogeneity, susceptibility to bias, and lack of reporting on important contextual factors and details of program implementation. New information is characterizing the period after hospitalization as a time of high generalized risk, which has been termed the post-hospital syndrome. This framework for characterizing inherent post-discharge instability suggests new approaches to reducing readmissions.

Is there a relationship between patient satisfaction and favorable outcomes?

OBJECTIVE

Patient satisfaction with the health care experience has become a top priority for Centers for Medicare and Medicaid Services. With resources and efforts directed at patient satisfaction, we evaluated whether high patient satisfaction measured by HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems) surveys correlates with favorable outcomes.

METHODS

Medical centers were identified from the University HealthSystem Consortium database from 2011 to 2012. Variables included hospital characteristics, process measure compliance, and surgical outcomes. Chi-squared analysis was used to evaluate for variables associated with high patient satisfaction (defined as hospitals that scored above the 50th percentile of top box scores).

RESULTS

We identified 171 hospitals with complete data. The following variables were significantly associated with high overall patient satisfaction: large hospitals, high surgical volume, and low mortality (P < 0.001). Compliance with process measures and patient safety indicators, as well as length of stay, did not correlate with overall satisfaction. The presence of complications (P = 0.491) or increased rate of readmission (P = 0.056) were not found to affect patient satisfaction. Low mortality index was consistently found to be associated with high satisfaction across 9 of 10 HCAHPS domains.

CONCLUSIONS

We found that hospital size, surgical volume, and low mortality were associated with high overall patient satisfaction. However, with the exception of low mortality, favorable surgical outcomes were not consistently associated with high HCAHPS scores. With existing satisfaction surveys, we conclude that factors outside of surgical outcomes appear to influence patients' perceptions of their care.

Can valid and practical risk-prediction or casemix adjustment models, including adjustment for comorbidity, be generated from English hospital administrative data (Hospital Episode Statistics)? A national observational study

Background

NHS hospitals collect a wealth of administrative data covering accident and emergency (A&E) department attendances, inpatient and day case activity, and outpatient appointments. Such data are increasingly being used to compare units and services, but adjusting for risk is difficult.

Objectives

To derive robust risk-adjustment models for various patient groups, including those admitted for heart failure (HF), acute myocardial infarction, colorectal and orthopaedic surgery, and outcomes adjusting for available patient factors such as comorbidity, using England’s Hospital Episode Statistics (HES) data. To assess if more sophisticated statistical methods based on machine learning such as artificial neural networks (ANNs) outperform traditional logistic regression (LR) for risk prediction. To update and assess for the NHS the Charlson index for comorbidity. To assess the usefulness of outpatient data for these models.

Main outcome measures

Mortality, readmission, return to theatre, outpatient non-attendance. For HF patients we considered various readmission measures such as diagnosis-specific and total within a year.

Methods

We systematically reviewed studies comparing two or more comorbidity indices. Logistic regression, ANNs, support vector machines and random forests were compared for mortality and readmission. Models were assessed using discrimination and calibration statistics. Competing risks proportional hazards regression and various count models were used for future admissions and bed-days.

Results

Our systematic review and empirical analysis suggested that for general purposes comorbidity is currently best described by the set of 30 Elixhauser comorbidities plus dementia. Model discrimination was often high for mortality and poor, or at best moderate, for other outcomes, for examplec = 0.62 for readmission andc = 0.73 for death following stroke. Calibration was often good for procedure groups but poorer for diagnosis groups, with overprediction of low risk a common cause. The machine learning methods we investigated offered little beyond LR for their greater complexity and implementation difficulties. For HF, some patient-level predictors differed by primary diagnosis of readmission but not by length of follow-up. Prior non-attendance at outpatient appointments was a useful, strong predictor of readmission. Hospital-level readmission rates for HF did not correlate with readmission rates for non-HF; hospital performance on national audit process measures largely correlated only with HF readmission rates.

Conclusions

Many practical risk-prediction or casemix adjustment models can be generated from HES data using LR, though an extra step is often required for accurate calibration. Including outpatient data in readmission models is useful. The three machine learning methods we assessed added little with these data. Readmission rates for HF patients should be divided by diagnosis on readmission when used for quality improvement.

Future work

As HES data continue to develop and improve in scope and accuracy, they can be used more, for instance A&E records. The return to theatre metric appears promising and could be extended to other index procedures and specialties. While our data did not warrant the testing of a larger number of machine learning methods, databases augmented with physiological and pathology information, for example, might benefit from methods such as boosted trees. Finally, one could apply the HF readmissions analysis to other chronic conditions.

Funding

The National Institute for Health Research Health Services and Delivery Research programme.

Using the LACE index to predict hospital readmissions in congestive heart failure patients

Background

The LACE index has been used to predict the risk of unplanned readmission within 30 days after hospital discharge in both medical and surgical patients. The aim of this study is to validate the accuracy of using the LACE index in CHF patients.

Methods

This was a retrospective study. The LACE index score was calculated on each patient who was admitted to hospital due to an acute CHF exacerbation. Operational and clinical variables were collected from patients including basic clinical characteristics, length of hospitalization, comorbidities, number of previous ED visits in the past 6 months before the index admission, and the number of post discharge ED revisits at 30, 60, and 90 days. All variables were analyzed by multivariate logistic regression to determine the association between clinical variables and the hospital unplanned readmissions. C-statistic was used to discriminate those patients with high risk of readmissions.

Results

Of the 253 patients included in the study, 24.50% (62/253) experienced unplanned readmission to hospital within 30 days after discharge. The LACE index was slightly higher in patients readmitted versus patients not readmitted (12.17 ± 2.22 versus 11.80 ± 1.92, p = 0.199). Adjusted odds ratios based on logistic regression of all clinical variables showed only the number of previous ED visits (OR 1.79, 95% CI 1.30-2.47, p < 0.001), history of myocardial infarction (OR 2.51, 95% CI 1.02-6.21, p = 0.045), and history of peripheral vascular disease (OR 10.75, 95% CI 1.52-75.73, p = 0.017) increased the risk of unplanned readmission within 30 days of hospital discharge. However, patients with high LACE scores (≥10) had a significantly higher rate of ED revisits (15.04% vs 0%) within 30 days from the index discharge than those with low LACE scores (p = 0.030).

Conclusion

The LACE index may not accurately predict unplanned readmissions within 30 days from hospital discharge in CHF patients. The LACE high risk index may have utility as a screening tool to predict high risk ED revisits after hospital discharge.