Predicting non-elective hospital readmissions: a multi-site study. Department of Veterans Affairs Cooperative Study Group on Primary Care and Readmissions

Interpretable machine learning models for hospital readmission prediction: a two-step extracted regression tree approach

BACKGROUND

Advanced machine learning models have received wide attention in assisting medical decision making due to the greater accuracy they can achieve. However, their limited interpretability imposes barriers for practitioners to adopt them. Recent advancements in interpretable machine learning tools allow us to look inside the black box of advanced prediction methods to extract interpretable models while maintaining similar prediction accuracy, but few studies have investigated the specific hospital readmission prediction problem with this spirit.

METHODS

Our goal is to develop a machine-learning (ML) algorithm that can predict 30- and 90- day hospital readmissions as accurately as black box algorithms while providing medically interpretable insights into readmission risk factors. Leveraging a state-of-art interpretable ML model, we use a two-step Extracted Regression Tree approach to achieve this goal. In the first step, we train a black box prediction algorithm. In the second step, we extract a regression tree from the output of the black box algorithm that allows direct interpretation of medically relevant risk factors. We use data from a large teaching hospital in Asia to learn the ML model and verify our two-step approach.

RESULTS

The two-step method can obtain similar prediction performance as the best black box model, such as Neural Networks, measured by three metrics: accuracy, the Area Under the Curve (AUC) and the Area Under the Precision-Recall Curve (AUPRC), while maintaining interpretability. Further, to examine whether the prediction results match the known medical insights (i.e., the model is truly interpretable and produces reasonable results), we show that key readmission risk factors extracted by the two-step approach are consistent with those found in the medical literature.

CONCLUSIONS

The proposed two-step approach yields meaningful prediction results that are both accurate and interpretable. This study suggests a viable means to improve the trust of machine learning based models in clinical practice for predicting readmissions through the two-step approach.

Development of Prediction Models for Unplanned Hospital Readmission within 30 Days Based on Common Data Model: A Feasibility Study

Background  Unplanned hospital readmission after discharge reflects low satisfaction and reliability in care and the possibility of potential medical accidents, and is thus indicative of the quality of patient care and the appropriateness of discharge plans.

Objectives  The purpose of this study was to develop and validate prediction models for all-cause unplanned hospital readmissions within 30 days of discharge, based on a common data model (CDM), which can be applied to multiple institutions for efficient readmission management.

Methods  Retrospective patient-level prediction models were developed based on clinical data of two tertiary general university hospitals converted into a CDM developed by Observational Medical Outcomes Partnership. Machine learning classification models based on the LASSO logistic regression model, decision tree, AdaBoost, random forest, and gradient boosting machine (GBM) were developed and tested by manipulating a set of CDM variables. An internal 10-fold cross-validation was performed on the target data of the model. To examine its transportability, the model was externally validated. Verification indicators helped evaluate the model performance based on the values of area under the curve (AUC).

Results  Based on the time interval for outcome prediction, it was confirmed that the prediction model targeting the variables obtained within 30 days of discharge was the most efficient (AUC of 82.75). The external validation showed that the model is transferable, with the combination of various clinical covariates. Above all, the prediction model based on the GBM showed the highest AUC performance of 84.14 ± 0.015 for the Seoul National University Hospital cohort, yielding in 78.33 in external validation.

Conclusions  This study showed that readmission prediction models developed using machine-learning techniques and CDM can be a useful tool to compare two hospitals in terms of patient-data features.

Hospital-at-Home Interventions vs In-Hospital Stay for Patients With Chronic Disease Who Present to the Emergency Department: A Systematic Review and Meta-analysis

Importance

Hospitalizations are costly and may lead to adverse events; hospital-at-home interventions could be a substitute for in-hospital stays, particularly for patients with chronic diseases who use health services more than other patients. Despite showing promising results, heterogeneity in past systematic reviews remains high.

Objective

To systematically review and assess the association between patient outcomes and hospital-at-home interventions as a substitute for in-hospital stay for community-dwelling patients with a chronic disease who present to the emergency department and are offered at least 1 home visit from a nurse and/or physician.

Data Sources

Databases were searched from date of inception to March 4, 2019. The databases were Ovid MEDLINE, Ovid Embase, Ovid PsycINFO, CINAHL, Health Technology Assessment, the Cochrane Library, OVID Allied and Complementary Medicine Database, the World Health Organization International Clinical Trials Registry Platform, and ClinicalTrials.gov.

Study Selection

Randomized clinical trials in which the experimental group received hospital-at-home interventions and the control group received the usual in-hospital care. Patients were 18 years or older with a chronic disease who presented to the emergency department and received home visits from a nurse or physician.

Data Extraction and Synthesis

Risk of bias was assessed, and a meta-analysis was conducted for outcomes that were reported by at least 2 studies using comparable measures. Risk ratios (RRs) were reported for binary outcomes and mean differences for continuous outcomes. Narrative synthesis was performed for other outcomes.

Main Outcomes and Measures

Outcomes of interest were patient outcomes, which included mortality, long-term care admission, readmission, length of treatment, out-of-pocket costs, depression and anxiety, quality of life, patient satisfaction, caregiver stress, cognitive status, nutrition, morbidity due to hospitalization, functional status, and neurological deficits.

Results

Nine studies were included, providing data on 959 participants (median age, 71.0 years [interquartile range, 70.0-79.9 years]; 613 men [63.9%]; 346 women [36.1%]). Mortality did not differ between the hospital-at-home and the in-hospital care groups (RR, 0.84; 95% CI, 0.61-1.15; I2 = 0%). Risk of readmission was lower (RR, 0.74; 95% CI, 0.57-0.95; I2 = 31%) and length of treatment was longer in the hospital-at-home group than in the in-hospital group (mean difference, 5.45 days; 95% CI, 1.91-8.97 days; I2 = 87%). In addition, the hospital-at-home group had a lower risk of long-term care admission than the in-hospital care group (RR, 0.16; 95% CI, 0.03-0.74; I2 = 0%). Patients who received hospital-at-home interventions had lower depression and anxiety than those who remained in-hospital, but there was no difference in functional status. Other patient outcomes showed mixed results.

Conclusions and Relevance

The results of this systematic review and meta-analysis suggest that hospital-at-home interventions represent a viable substitute to an in-hospital stay for patients with chronic diseases who present to the emergency department and who have at least 1 visit from a nurse or physician. Although the heterogeneity of the findings remained high for some outcomes, particularly for length of treatment, the heterogeneity of this study was comparable to that of past reviews and further explored.

Using Health Plan Authorization Data to Predict Hospitalizations and Readmissions in a Combined Commercial and Medicaid Population

Health plans develop predictive models to predict key clinical events (eg, admissions, readmissions, emergency department visits). The authors developed predictive models of admissions and readmissions for a quality improvement organization with many large government and private health plan clients. Its membership and authorization data were used to develop models predicting 2019 inpatient stays, and 2019 readmissions following 2019 admissions, based on patients' age and sex, diagnoses identified and procedures requested in 2018 authorizations, and 2018 admission authorizations. In addition to testing multivariate models, risk scores were calculated for admission and readmission for all patients in the model. The admissions model (C = 0.8491) is much more accurate than the readmissions model (C = 0.6237). Measures of risk score central tendency and skewness indicate that the vast majority of members had little risk of hospitalization in 2019; the mean (standard deviation) was 0.042 (0.074), and the median was 0.018. These risk scores can be used to identify members at risk of admission and to support proactive risk management (eg, design of health management programs). Different risk thresholds can be used to identify different subsets of members for follow-up, depending on overall strategy and available resources. This model development project was novel in employing authorization data rather than utilization data. Advantages of authorization data are their timeliness, and the fact that they are sometimes the only data available, but disadvantages of authorization data are that authorized services are not always actually performed, and diagnoses are often "rule out" rather than final diagnoses.

Implementation of machine learning algorithms to create diabetic patient re-admission profiles

BACKGROUND

Machine learning is a branch of Artificial Intelligence that is concerned with the design and development of algorithms, and it enables today's computers to have the property of learning. Machine learning is gradually growing and becoming a critical approach in many domains such as health, education, and business.

METHODS

In this paper, we applied machine learning to the diabetes dataset with the aim of recognizing patterns and combinations of factors that characterizes or explain re-admission among diabetes patients. The classifiers used include Linear Discriminant Analysis, Random Forest, k-Nearest Neighbor, Naïve Bayes, J48 and Support vector machine.

RESULTS

Of the 100,000 cases, 78,363 were diabetic and over 47% were readmitted.Based on the classes that models produced, diabetic patients who are more likely to be readmitted are either women, or Caucasians, or outpatients, or those who undergo less rigorous lab procedures, treatment procedures, or those who receive less medication, and are thus discharged without proper improvements or administration of insulin despite having been tested positive for HbA1c.

CONCLUSION

Diabetic patients who do not undergo vigorous lab assessments, diagnosis, medications are more likely to be readmitted when discharged without improvements and without receiving insulin administration, especially if they are women, Caucasians, or both.

Vital Sign Abnormalities on Discharge Do Not Predict 30-Day Readmission

OBJECTIVE

Hospital readmissions are common and expensive. Risk factors for hospital readmission may include vital sign abnormalities (VSA) at the time of discharge. The study aimed to validate VSA at the time of discharge as a useful predictor of hospital readmission within 30 days of discharge. VSA was compared to the validated HOSPITAL score and LACE index readmission risk prediction models.

DESIGN

All adult medical patients discharged from internal medicine hospitalist service were studied retrospectively. Variables such as age, gender, diagnoses, vital signs at discharge, 30-day hospital readmission, and components for the HOSPITAL score and LACE index were extracted from the electronic health record for analysis.

SETTINGS

A 507-bed university-affiliated tertiary care center.

PARTICIPANTS

During the 2-year study period, a cohort of 1,916 discharges for the hospitalist service were evaluated. The final analysis was based on the data from 1,781 hospital discharges that met the inclusion criteria.

RESULTS

VSA was found in 13% of the study population. Only one abnormal vital sign was present in a higher proportion readmitted to the hospital within 30 days of discharge. No discharges had three or more unstable vital signs. Receiver operating characteristic (ROC) comparisons of the HOSPITAL score (C statistic of 0.67, P < 0.001), LACE index (C statistic of 0.61, P < 0.001), and VSA (C statistic of 0.52, P = 0.318) indicated that VSA at time of discharge was not a useful predictor of hospital readmission within 30 days of discharge.

CONCLUSION

Our study indicated that VSA at the time of discharge is not a useful predictor of 30-day hospital readmission at a university-affiliated teaching hospital. The more complex and validated HOSPITAL score and LACE index were useful predictors of hospital readmission in this patient population.

Predictors of 30-day hospital readmission: The direct comparison of number of discharge medications to the HOSPITAL score and LACE index

Effective hospital readmission risk prediction tools exist, but do not identify actionable items that could be modified to reduce the risk of readmission. Polypharmacy has attracted attention as a potentially modifiable risk factor for readmission, showing promise in a retrospective study. Polypharmacy is a very complex issue, reflecting comorbidities and healthcare resource utilisation patterns. This investigation compares the predictive ability of polypharmacy alone to the validated HOSPITAL score and LACE index readmission risk assessment tools for all adult admissions to an academic hospitalist service at a moderate sized university-affiliated hospital in the American Midwest over a 2-year period. These results indicate that the number of discharge medications alone is not a useful tool in identifying patients at high risk of hospital readmission within 30 days of discharge. Further research is needed to explore the impact of polypharmacy as a risk predictor for hospital readmission.

Derivation and validation of a predictive model for the readmission of patients with diabetes mellitus treated in internal medicine departments

OBJECTIVES

We developed a predictive model for the hospital readmission of patients with diabetes. The objective was to identify the frail population that requires additional strategies to prevent readmissions at 90 days.

METHODS

Using data collected from 1977 patients in 3 studies on the national prevalence of diabetes (2015-2017), we developed and validated a predictive model of readmission at 90 days for patients with diabetes.

RESULTS

A total of 704 (36%) readmissions were recorded. There were no differences in the readmission rates over the course of the 3 studies. The hospitals with more than 500 beds showed significantly (p=.02) higher readmission rates than those with fewer beds. The main reasons for readmission were infectious diseases (29%), cardiovascular diseases (24) and respiratory diseases (14%). Readmissions directly related to diabetic decompensations accounted for only 2% of all readmissions. The independent variables associated with hospital readmission were patient's age, degree of comorbidity, estimated glomerular filtration rate, degree of disability, presence of previous episodes of hypoglycaemia, use of insulin in treating diabetes and the use of systemic glucocorticoids. The predictive model showed an area under the ROC curve (AUC) of 0.676 (95% confidence interval [95% CI] 0.642-0.709; p=.001) in the referral cohort. In the validation cohort, the model showed an AUC of 0.661 (95% CI 0.612-0.710; p=.001).

CONCLUSION

The model we developed for predicting readmissions for hospitalised patients with type 2 diabetes helps identify a subgroup of frail patients with a high risk of readmission.

Correction to: Hospital Readmission of Patients with Diabetes

This article was originally published with errors that were introduced during the editing process. The corrected version of this article appears below.

Mobile Compression Reduces Bleeding-related Readmissions and Wound Complications After THA and TKA

BACKGROUND

The use of chemoprophylaxis to prevent thromboembolic disease after primary THA and TKA can be associated with postoperative bleeding complications. Mechanical prophylaxis has been studied as an alternative to chemoprophylaxis with greater safety in patients undergoing THA, but no data have been published comparing the safety of chemoprophylaxis versus mechanical methods for patients undergoing TKA. The risk of readmission resulting from bleeding and venous thromboembolism (VTE) has also not been determined for patients undergoing THA or TKA when treated with low-molecular-weight heparin (LMWH) alone compared with mechanical prophylaxis plus aspirin (ASA).

QUESTION/PURPOSES

We sought to answer four questions: For the THA and TKA cohorts, respectively, (1) was the incidence of readmission resulting from VTE and bleeding complications higher with LMWH than mobile compression plus ASA; and (2) was the incidence of wound bleeding complications higher with LMWH than mechanical compression plus ASA? For the TKA cohort specifically, (3) was the frequency of systemic bleeding events and complications related to chemical prophylaxis higher with LMWH compared with mechanical compression plus ASA? (4) Was there a difference in symptomatic VTEs between LMWH and mechanical compression plus ASA?

METHODS

Between November 2008 and April 2011, 632 patients underwent primary THA and TKA. Seventy-two patients (11%) were identified before surgery as being at high risk for VTE (31 patients) or bleeding (41 patients) and were excluded from the study. Five hundred sixty patients (89%) were considered to be at standard risk for VTE and bleeding and comprise the study cohort. Between November 2008 and November 2009, 252 patients (76 THAs, 176 TKAs) underwent THA and TKA and were treated with LMWH (5 mg dalteparin given subcutaneously daily for 14 days) and in-hospital nonmobile mechanical compression. Between November 2009 and April 2011, a total of 308 patients undergoing THA and TKA (108 THAs, 200 TKAs) were treated using a mobile compression device plus oral aspirin once daily for 2 weeks after surgery. All complications and readmissions that occurred within 6 weeks of surgery were noted. There were no differences between the VTE treatment groups with regard to age, sex, or body mass index.

RESULTS

For the THA cohort, there was no difference in the frequency of readmission for a bleeding complication (wound or systemic) between the two groups (2.6% for LMWH versus 0.9% for mobile compression; p = 0.57; odds ratio [OR], 2.9). Patients undergoing TKA treated with LMWH had higher readmission rates within 6 weeks of surgery because of a bleeding complication, a wound infection, or the development of a VTE (6.8% for LMWH versus 1.5% for mobile compression; p = 0.015; OR, 4.8). For the THA cohort, there was higher wound bleeding complication frequency with LMWH (9.2% for LMWH versus 0.9% for mechanical compression; p = 0.009; OR, 10.9). Patients undergoing TKA treated with LMWH had a higher frequency of wound bleeding complications or infection (3.9% for LMWH versus 0.5% for mobile compression; p = 0.028; OR, 8.2). Patients undergoing TKA treated with LMWH had higher rates of systemic bleeding or a complication secondary to LMWH administration (2.8% for LMWH versus 0% for mobile compression; p = 0.022; OR, 12.8). No difference was noted in the rate of symptomatic VTEs between either group (for THA: 2.6% for the LMWH group versus 1.9% for the mechanical compression group; p = 1; for TKA: 1.1% versus 0%, respectively; p = 0.22).

CONCLUSIONS

Based on these results, we advocate for routine use of mobile mechanical compression devices in the prevention of VTEs and complications associated with more potent chemical anticoagulants. However, more focused randomized clinical trials are needed to validate these findings.

LEVEL OF EVIDENCE

Level III, therapeutic study.

Reducing COPD readmissions through predictive modeling and incentive-based interventions

This paper introduces a case study at a community hospital to develop a predictive model to quantify readmission risks for patients with chronic obstructive pulmonary disease (COPD), and use it to support decision making for appropriate incentive-based interventions. Data collected from the community hospital's database are analyzed to identify risk factors and a logistic regression model is developed to predict the readmission risk within 30 days post-discharge of an individual COPD patient. By targeting on the high-risk patients, we investigate the implementability of the incentive policy which encourages patients to take interventions and helps them to overcome the compliance barrier. Specifically, the conditions and scenarios are identified for either achieving the desired readmission rate while minimizing the total cost, or reaching the lowest readmission rate under incentive budget constraint. Currently, such models are under consideration for a pilot study at the community hospital.

Impact of a Patient Navigator Program on Hospital-Based and Outpatient Utilization Over 180 Days in a Safety-Net Health System

BACKGROUND

With emerging global payment structures, medical systems need to understand longer-term impacts of care transition strategies.

OBJECTIVE

To determine the effect of a care transition program using patient navigators (PNs) on health service utilization among high-risk safety-net patients over a 180-day period.

DESIGN

Randomized controlled trial conducted October 2011 through April 2013.

PARTICIPANTS

Patients admitted to the general medicine service with ≥1 readmission risk factor: (1) age ≥ 60; (2) in-network inpatient admission within prior 6 months; (3) index length of stay ≥ 3 days; or (4) admission diagnosis of heart failure or (5) chronic obstructive pulmonary disease. The analytic sample included 739 intervention patients, 1182 controls.

INTERVENTIONS

Through hospital visits and 30 days of post-discharge telephone outreach, PNs provided coaching and assistance with medications, appointments, transportation, communication with primary care, and self-care.

MAIN MEASURES

Primary outcomes: (1) hospital-based utilization, a composite of ED visits and hospital admissions; (2) hospital admissions; (3) ED visits; and (4) outpatient visits. We evaluated outcomes following an index discharge, stratified by patient age (≥ 60 and < 60 years), using a 180-day time frame divided into six 30-day periods.

KEY RESULTS

The PN program produced starkly different outcomes by patient age. Among older PN patients, hospital-based utilization was consistently lower than controls, producing an 18.7% cumulative decrease at 180 days (p = 0.038); outpatient visits increased in the critical first 30-day period (p = 0.006). Among younger PN patients, hospital-based utilization was 31.7% (p = 0.038) higher at 180 days, largely reflecting sharply higher utilization in the initial 30 days (p = 0.002), with non-significant changes thereafter; outpatient visits experienced no significant changes.

CONCLUSIONS

A PN program serving high-risk safety-net patients differentially impacted patients based on age, and among younger patients, outcomes varied over time. Our findings highlight the importance for future research to evaluate care transition programs among different subpopulations and over longer time periods.

The Relationship Between Index Hospitalizations, Sepsis, and Death or Transition to Hospice Care During 30-Day Hospital Readmissions

BACKGROUND

Hospital readmissions are common, expensive, and increasingly used as a metric for assessing quality of care. The relationship between index hospitalizations and specific outcomes among those readmitted remains largely unknown.

OBJECTIVES

Identify risk factors present during the index hospitalization associated with death or transition to hospice care during 30-day readmissions and examine the contribution of infection in readmissions resulting in death.

RESEARCH DESIGN

Retrospective cohort study.

SUBJECTS

A total of 17,716 30-day readmissions in an academic health system.

MEASURES

We used mixed-effects multivariable logistic regression models to identify risk factors associated with the primary outcome, in-hospital death, or transition to hospice during 30-day readmissions.

RESULTS

Of 17,716 30-day readmissions, 1144 readmissions resulted in death or transition to hospice care (6.5%). Risk factors identified included: age, burden, and type of comorbid conditions, recent hospitalizations, nonelective index admission type, outside hospital transfer, low discharge hemoglobin, low discharge sodium, high discharge red blood cell distribution width, and disposition to a setting other than home. Sepsis (OR=1.33; 95% CI, 1.02-1.72; P=0.03) and shock (OR=1.78; 95% CI, 1.22-2.58; P=0.002) during the index admission were associated with the primary outcome, and in-hospital mortality specifically. In patients who died, infection was the primary cause for readmission in 51.6% of readmissions after sepsis and 28.6% of readmissions after a nonsepsis hospitalization (P=0.009).

CONCLUSIONS

We identified factors, including sepsis and shock during the index hospitalization, associated with death or transition to hospice care during readmission. Infection was frequently implicated as the cause of a readmission that ended in death.

The HOSPITAL score and LACE index as predictors of 30 day readmission in a retrospective study at a university-affiliated community hospital

Introduction

Hospital readmissions are common, expensive, and a key target of the Medicare Value Based Purchasing (VBP) program. Validated risk assessment tools such as the HOSPITAL score and LACE index have been developed to identify patients at high risk of hospital readmission so they can be targeted for interventions aimed at reducing the rate of readmission. This study aims to evaluate the utility of HOSPITAL score and LACE index for predicting hospital readmission within 30 days in a moderate-sized university affiliated hospital in the midwestern United States.

Materials and Methods

All adult medical patients who underwent one or more ICD-10 defined procedures discharged from the SIU-SOM Hospitalist service from Memorial Medical Center (MMC) from October 15, 2015 to March 16, 2016, were studied retrospectively to determine if the HOSPITAL score and LACE index were a significant predictors of hospital readmission within 30 days.

Results

During the study period, 463 discharges were recorded for the hospitalist service. The analysis includes data for the 432 discharges. Patients who died during the hospital stay, were transferred to another hospital, or left against medical advice were excluded. Of these patients, 35 (8%) were readmitted to the same hospital within 30 days. A receiver operating characteristic evaluation of the HOSPITAL score for this patient population shows a C statistic of 0.75 (95% CI [0.67–0.83]), indicating good discrimination for hospital readmission. The Brier score for the HOSPITAL score in this setting was 0.069, indicating good overall performance. The Hosmer–Lemeshow goodness of fit test shows a χ² value of 3.71 with a p value of 0.59. A receiver operating characteristic evaluation of the LACE index for this patient population shows a C statistic of 0.58 (95% CI [0.48–0.68]), indicating poor discrimination for hospital readmission. The Brier score for the LACE index in this setting was 0.082, indicating good overall performance. The Hosmer–Lemeshow goodness of fit test shows a χ² value of 4.97 with a p value of 0.66.

Discussion

This single center retrospective study indicates that the HOSPITAL score has superior discriminatory ability when compared to the LACE index as a predictor of hospital readmission within 30 days at a medium-sized university-affiliated teaching hospital.

Conclusions

The internationally validated HOSPITAL score may be superior to the LACE index in moderate-sized community hospitals to identify patients at high risk of hospital readmission within 30 days.

A study to reduce readmissions after surgery in the Veterans Health Administration: design and methodology

BACKGROUND

Hospital readmissions are associated with higher resource utilization and worse patient outcomes. Causes of unplanned readmission to the hospital are multiple with some being better targets for intervention than others. To understand risk factors for surgical readmission and their incremental contribution to current Veterans Health Administration (VA) surgical quality assessment, the study, Improving Surgical Quality: Readmission (ISQ-R), is being conducted to develop a readmission risk prediction tool, explore predisposing and enabling factors, and identify and rank reasons for readmission in terms of salience and mutability.

METHODS

Harnessing the rich VA enterprise data, predictive readmission models are being developed in data from patients who underwent surgical procedures within the VA 2007-2012. Prospective assessment of psychosocial determinants of readmission including patient self-efficacy, cognitive, affective and caregiver status are being obtained from a cohort having colorectal, thoracic or vascular procedures at four VA hospitals in 2015-2017. Using these two data sources, ISQ-R will develop readmission categories and validate the readmission risk prediction model. A modified Delphi process will convene surgeons, non-surgeon clinicians and quality improvement nurses to rank proposed readmission categories vis-à-vis potential preventability.

DISCUSSION

ISQ-R will identify promising avenues for interventions to facilitate improvements in surgical quality, informing specifications for surgical workflow managers seeking to improve care and reduce cost. ISQ-R will work with Veterans Affairs Surgical Quality Improvement Program (VASQIP) to recommend potential new elements VASQIP might collect to monitor surgical complications and readmissions which might be preventable and ultimately improve surgical care.

The Situation-Specific Theory of Heart Failure Self-Care: Revised and Updated

BACKGROUND

Since the situation-specific theory of heart failure (HF) self-care was published in 2008, we have learned much about how and why patients with HF take care of themselves. This knowledge was used to revise and update the theory.

OBJECTIVE

The purpose of this article was to describe the revised, updated situation-specific theory of HF self-care.

RESULT

Three major revisions were made to the existing theory: (1) a new theoretical concept reflecting the process of symptom perception was added; (2) each self-care process now involves both autonomous and consultative elements; and (3) a closer link between the self-care processes and the naturalistic decision-making process is described. In the revised theory, HF self-care is defined as a naturalistic decision-making process with person, problem, and environmental factors that influence the everyday decisions made by patients and the self-care actions taken. The first self-care process, maintenance, captures those behaviors typically referred to as treatment adherence. The second self-care process, symptom perception, involves body listening, monitoring signs, as well as recognition, interpretation, and labeling of symptoms. The third self-care process, management, is the response to symptoms when they occur. A total of 5 assumptions and 8 testable propositions are specified in this revised theory.

CONCLUSION

Prior research illustrates that all 3 self-care processes (ie, maintenance, symptom perception, and management) are integral to self-care. Further research is greatly needed to identify how best to help patients become experts in HF self-care.

Satisfaction of Nigerian stroke survivors with outpatient physiotherapy care

OBJECTIVE

To investigate the satisfaction of stroke survivors with outpatient physiotherapy care.

METHODS

Sixty stroke survivors were surveyed using the European Physiotherapy Treatment Outpatient Satisfaction Survey (EPTOPS). Focus group discussion (FGD) was also conducted with four stroke survivors from the same sample. Data were analyzed using the Kruskal Wallis test and Spearman's correlation coefficients at p = 0.05. FGD was transcribed and thematically analyzed.

RESULTS

Nearly all the participants (98.3%) indicated one of good, very good, and excellent improvement in their clinical conditions with physiotherapy. Majority expressed satisfaction with their physiotherapy care, the modal response being very good (59.3%). Patients' satisfaction and socio-demographics were not significantly correlated (p > 0.05). Overarching themes from FGD were physiotherapy in stroke rehabilitation, satisfaction with physiotherapy care, cost, and lack of continuity of care as sources of dissatisfaction. Physiotherapists' demeanor was a facilitator of satisfaction.

CONCLUSION

The stroke survivors were generally satisfied with outpatient physiotherapy care. However, lack of continuity and cost of care were sources of dissatisfaction among patients. Delivery of physiotherapy to stroke survivors in Nigeria should be structured to allow for continuity of care as this may enhance satisfaction. Implementation of inexpensive rehabilitation strategies may help reduce cost of physiotherapy.

The HOSPITAL score as a predictor of 30 day readmission in a retrospective study at a university affiliated community hospital

Introduction

Hospital readmissions are common, expensive, and a key target of the Medicare Value Based Purchasing (VBP) program. Risk assessment tools have been developed to identify patients at high risk of hospital readmission so they can be targeted for interventions aimed at reducing the rate of readmission. One such tool is the HOSPITAL score that uses seven readily available clinical variables to predict the risk of readmission within 30 days of discharge. The HOSPITAL score has been internationally validated in large academic medical centers. This study aims to determine if the HOSPITAL score is similarly useful in a moderate sized university affiliated hospital in the midwestern United States.

Materials and Methods

All adult medical patients discharged from the SIU-SOM Hospitalist service from Memorial Medical Center (MMC) from October 15, 2015 to March 16, 2016, were studied retrospectively to determine if the HOSPITAL score was a significant predictor of hospital readmission within 30 days.

Results

During the study period, 998 discharges were recorded for the hospitalist service. The analysis includes data for the 931 discharges. Patients who died during the hospital stay, were transferred to another hospital, or left against medical advice were excluded. Of these patients, 109 (12%) were readmitted to the same hospital within 30 days. The patients who were readmitted were more likely to have a length of stay greater than or equal to 5 days (55% vs. 41%, p = 0.005) and were more likely to have been admitted more than once to the hospital within the last year (100% vs. 49%, p < 0.001). A receiver operating characteristic evaluation of the HOSPITAL score for this patient population shows a C statistic of 0.77 (95% CI [0.73–0.81]), indicating good discrimination for hospital readmission. The Brier score for the HOSPITAL score in this setting was 0.10, indicating good overall performance. The Hosmer–Lemeshow goodness of fit test shows a χ² value of 1.63 with a p value of 0.20.

Discussion

This single center retrospective study indicates that the HOSPITAL score has good discriminatory ability to predict hospital readmissions within 30 days for a medical hospitalist service at a university-affiliated hospital. This data for all causes of hospital readmission is comparable to the discriminatory ability of the HOSPITAL score in the international validation study (C statistics of 0.72 vs. 0.77) conducted at considerably larger hospitals (975 average beds vs. 507 at MMC) for potentially avoidable hospital readmissions.

Conclusions

The internationally validated HOSPITAL score may be a useful tool in moderate sized community hospitals to identify patients at high risk of hospital readmission within 30 days. This easy to use scoring system using readily available data can be used as part of interventional strategies to reduce the rate of hospital readmission.

Risk Factors for Readmission in Patients With Cancer Comanaged by Hospitalists

Little is known about which variables put patients with cancer at risk for 30-day hospital readmission. Comanagement of this often complex patient population by specialists and hospitalists has become increasingly common. This retrospective study examined inpatients with cancer comanaged by hospitalists, hematologists, and oncologists to determine the rate of readmission and factors associated with readmission. Patients in this cohort had a readmission rate of 23%. Patients who were discharged to a skilled nursing facility (odds ratio [OR] = 0.34) or hospice (OR = 0.11) were less likely to have 30-day readmissions, whereas patients who had surgery (OR = 3.16) during their index admission were more likely. Other factors, including patient demographics, cancer types, and hospitalization interventions and events, did not differ between patients who were readmitted and those who were not. These findings contribute to a growing body of literature identifying risk factors for readmission in medical oncology and hematology patients.

Factors Related to Chronic Obstructive Pulmonary Disease Readmission in Taiwan

This study examines the relationships among physiological, psychological, and social factors and hospital readmission to develop a model predicting chronic obstructive pulmonary disease (COPD) readmission for 145 patients with COPD following hospital discharge at 14 days and 90 days in Taiwan. Daily functioning, comorbidity, severity of illness, self-efficacy, depressive symptoms, and perceived informal support were regressed on hospital readmission. Daily functioning was the only significant variable to predict COPD readmission at 90 days in the Taiwan population living in a rural area. Age was significantly correlated with 14 days readmission. Post hoc analyses examined differences in three ethnic groups. Mainlanders perceived less family support, had higher depressive symptoms and lower daily functioning than the majority culture Fukiens and Hakkas, or the Aborigines. The study reinforced the need for identification of cultural differences and low functioning as risk factors for early readmission so they can be addressed in discharge planning.

Readmissions of medical patients: an external validation of two existing prediction scores

BACKGROUND

Hospital readmissions are increasingly used as a quality indicator with a belief that they are a marker of poor care and have led to financial penalties in UK and USA. Risk scoring systems, such as LACE and HOSPITAL, have been proposed as tools for identifying patients at high risk of readmission but have not been validated in international populations.

AIM

To perform an external independent validation of the HOSPITAL and LACE scores.

DESIGN

An unplanned secondary cohort study.

METHODS

Patients admitted to the medical admission unit at the Hospital of South West Jutland (10/2008-2/2009; 2/2010-5/2010) and the Odense University Hospital (6/2009-8/2011) were analysed. Validation of the scores using 30 day readmissions as the endpoint was performed.

RESULTS

A total of 19 277 patients fulfilled the inclusion criteria. Median age was 67 (range 18-107) years and 8977 (46.6%) were female. The LACE score had a discriminatory power of 0.648 with poor calibration and the HOSPITAL score had a discriminatory power of 0.661 with poor calibration. The HOSPITAL score was significantly better than the LACE score for identifying patients at risk of 30 day readmission (P < 0.001). The discriminatory power of both scores decreased with increasing age.

CONCLUSION

Readmissions are a complex phenomenon with not only medical conditions contributing but also system, cultural and environmental factors exerting a significant influence. It is possible that the heterogeneity of the population and health care systems may prohibit the creation of a simple prediction tool that can be used internationally.

Predicting 30-day Hospital Readmission with Publicly Available Administrative Database. A Conditional Logistic Regression Modeling Approach

INTRODUCTION

This article is part of the Focus Theme of Methods of Information in Medicine on "Big Data and Analytics in Healthcare".

BACKGROUND

Hospital readmissions raise healthcare costs and cause significant distress to providers and patients. It is, therefore, of great interest to healthcare organizations to predict what patients are at risk to be readmitted to their hospitals. However, current logistic regression based risk prediction models have limited prediction power when applied to hospital administrative data. Meanwhile, although decision trees and random forests have been applied, they tend to be too complex to understand among the hospital practitioners.

OBJECTIVES

Explore the use of conditional logistic regression to increase the prediction accuracy.

METHODS

We analyzed an HCUP statewide inpatient discharge record dataset, which includes patient demographics, clinical and care utilization data from California. We extracted records of heart failure Medicare beneficiaries who had inpatient experience during an 11-month period. We corrected the data imbalance issue with under-sampling. In our study, we first applied standard logistic regression and decision tree to obtain influential variables and derive practically meaning decision rules. We then stratified the original data set accordingly and applied logistic regression on each data stratum. We further explored the effect of interacting variables in the logistic regression modeling. We conducted cross validation to assess the overall prediction performance of conditional logistic regression (CLR) and compared it with standard classification models.

RESULTS

The developed CLR models outperformed several standard classification models (e.g., straightforward logistic regression, stepwise logistic regression, random forest, support vector machine). For example, the best CLR model improved the classification accuracy by nearly 20% over the straightforward logistic regression model. Furthermore, the developed CLR models tend to achieve better sensitivity of more than 10% over the standard classification models, which can be translated to correct labeling of additional 400 - 500 readmissions for heart failure patients in the state of California over a year. Lastly, several key predictor identified from the HCUP data include the disposition location from discharge, the number of chronic conditions, and the number of acute procedures.

CONCLUSIONS

It would be beneficial to apply simple decision rules obtained from the decision tree in an ad-hoc manner to guide the cohort stratification. It could be potentially beneficial to explore the effect of pairwise interactions between influential predictors when building the logistic regression models for different data strata. Judicious use of the ad-hoc CLR models developed offers insights into future development of prediction models for hospital readmissions, which can lead to better intuition in identifying high-risk patients and developing effective post-discharge care strategies. Lastly, this paper is expected to raise the awareness of collecting data on additional markers and developing necessary database infrastructure for larger-scale exploratory studies on readmission risk prediction.

Low serum albumin and total lymphocyte count as predictors of 30 day hospital readmission in patients 65 years of age or older

Introduction. Hospital readmission within 30 days of discharge is a target for health care cost savings through the medicare Value Based Purchasing initiative. Because of this focus, hospitals and health systems are investing considerable resources into the identification of patients at risk of hospital readmission and designing interventions to reduce the rate of hospital readmission. Malnutrition is a known risk factor for hospital readmission.

Materials and Methods. All medical patients 65 years of age or older discharged from Memorial Medical Center from January 1, 2012 to March 31, 2012 who had a determination of serum albumin level and total lymphocyte count on hospital admission were studied retrospectively. Admission serum albumin levels and total lymphocyte counts were used to classify the nutritional status of all patients in the study. Patients with a serum albumin less than 3.5 grams/dL and/or a TLC less than 1,500 cells per mm3 were classified as having protein energy malnutrition. The primary outcome investigated in this study was hospital readmission for any reason within 30 days of discharge.

Results. The study population included 1,683 hospital discharges with an average age of 79 years. The majority of the patients were female (55.9%) and had a DRG weight of 1.22 (0.68). 219 patients (13%) were readmitted within 30 days of hospital discharge. Protein energy malnutrition was common in this population. Low albumin was found in 973 (58%) patients and a low TLC was found in 1,152 (68%) patients. Low albumin and low TLC was found in 709 (42%) of patients. Kaplan–Meier analysis shows any laboratory evidence of PEM is a significant (p < 0.001) predictor of hospital readmission. Low serum albumin (p < 0.001) and TLC (p = 0.018) show similar trends. Cox proportional-hazards regression analysis showed low serum albumin (Hazard Ratio 3.27, 95% CI [2.30–4.63]) and higher DRG weight (Hazard Ratio 1.19, 95% CI [1.03–1.38]) to be significant independent predictors of hospital readmission within 30 days.

Discussion. This study investigated the relationship of PEM to the rate of hospital readmission within 30 days of discharge in patients 65 years of age or older. These results indicate that laboratory markers of PEM can identify patients at risk of hospital readmission within 30 days of discharge. This risk determination is simple and identifies a potentially modifiable risk factor for readmission: protein energy malnutrition.

Functional Status and Hospital Readmissions Using the Medical Expenditure Panel Survey

BACKGROUND

Hospital readmissions are expensive and they may signal poor quality of care. Whether functional status is related to hospital readmissions using a representative U.S sample remains unexplored .

OBJECTIVE

We aimed to assess the relationship between functional status and all-cause 30-day hospital readmissions using a representative sample of the US population.

DESIGN

This was a retrospective observational study (2003-2011).

PATIENTS

The study included 3,772 patients who completed the SF-12 before being hospitalized. Three hundred and eighteen (8.4%) were readmitted within 30 days after being discharged.

MEASUREMENTS

The Medical Expenditure Panel Survey (MEPS) was employed. Functional status was measured with the Short-Form 12-Item Health Survey Version 2® (SF-12). The probability of being readmitted was estimated using a logistic model controlling for demographic characteristics, comorbid conditions, insurance coverage, physical (PCS) and mental (MCS) summaries of the SF-12, reason for hospitalization, length of hospital stay, region, and residential area.

RESULTS

A one-unit difference in PCS reduced the odds of readmission by 2% (odds ratio 0.98 [95% CI, 0.97 to 0.99]; p < 0.001), which implies an 18% reduction in the odds of readmissions for a ten-unit difference (one standard deviation) in PCS. The c-statistic of the model was 0.72.

CONCLUSION

Baseline physical function is associated with hospital readmissions. The SF-12 improves the ability to identify patients at high risk of hospital readmission.

A Patient Navigator Intervention to Reduce Hospital Readmissions among High-Risk Safety-Net Patients: A Randomized Controlled Trial

BACKGROUND

Evidence-based interventions to reduce hospital readmissions may not generalize to resource-constrained safety-net hospitals.

OBJECTIVE

To determine if an intervention by patient navigators (PNs), hospital-based Community Health Workers, reduces readmissions among high risk, low socioeconomic status patients.

DESIGN

Randomized controlled trial.

PARTICIPANTS

General medicine inpatients having at least one of the following readmission risk factors: (1) age ≥60 years, (2) any in-network inpatient admission within the past 6 months, (3) length of stay ≥3 days, (4) admission diagnosis of heart failure, or (5) chronic obstructive pulmonary disease. The analytic sample included 585 intervention patients and 925 controls.

INTERVENTIONS

PNs provided coaching and assistance in navigating the transition from hospital to home through hospital visits and weekly telephone outreach, supporting patients for 30 days post-discharge with discharge preparation, medication management, scheduling of follow-up appointments, communication with primary care, and symptom management.

MAIN MEASURES

The primary outcome was in-network 30-day hospital readmissions. Secondary outcomes included rates of outpatient follow-up. We evaluated outcomes for the entire cohort and stratified by patient age >60 years (425 intervention/584 controls) and ≤60 years (160 intervention/341 controls).

KEY RESULTS

Overall, 30-day readmission rates did not differ between intervention and control patients. However, the two age groups demonstrated marked differences. Intervention patients >60 years showed a statistically significant adjusted absolute 4.1% decrease [95% CI: -8.0%, -0.2%] in readmission with an increase in 30-day outpatient follow-up. Intervention patients ≤60 years showed a statistically significant adjusted absolute 11.8% increase [95% CI: 4.4%, 19.0%] in readmission with no change in 30-day outpatient follow-up.

CONCLUSIONS

A patient navigator intervention among high risk, safety-net patients decreased readmission among older patients while increasing readmissions among younger patients. Care transition strategies should be evaluated among diverse populations, and younger high risk patients may require novel strategies.

Demographic and clinical predictors of unplanned hospital utilisation among chronically ill patients: a prospective cohort study

BACKGROUND

In urban Australia, patients with serious and continuing illnesses make frequent use of hospital emergency department (ED) services. However, the risk factors for hospital utilisation among the broad population of people with chronic illness are not well known. The aim of this study was to assess the predictors of hospital utilisation (either inpatient admissions or ED visits) in a cohort of 308 patients with chronic illness.

METHODS

We studied patients with serious and continuing chronic illnesses presenting to an ED in a large periurban hospital in western Sydney, Australia, between 2010 and 2013. ED presentations and hospital admissions were observed over two years. Multivariate negative-binomial regression analyses were used to identify risk factors for the number of presentations to hospital.

RESULTS

The main risk factors for hospital utilisation were having a live-in carer, and a history of hospital utilisation. Having a live-in carer was associated with an increase in number of ED presentations by 88% (RR 1.88; 95% CI 1.41-2.51), and of admissions by 116% (RR 2.16; 95% CI 1.61-2.92). Seventy-seven percent of hospital utilisation in the cohort was attributable to carer status. Each additional ED presentation that a person had in the 12 months prior to the study led to an increased risk of an ED presentation in the follow-up period by 6% (RR = 1.06, 95% CI = 1.03-1.08). Between 20% and 25% of variability in hospital utilisation in the cohort was attributable to the number of hospital admissions or ED presentations in the previous 12 months.

CONCLUSIONS

Patients with a live-in carer and with a history of hospital utilisation are at high risk for future hospital use.

Hospital readmission of patients with diabetes

Hospital readmission is a high-priority health care quality measure and target for cost reduction. Despite broad interest in readmission, relatively little research has focused on patients with diabetes. The burden of diabetes among hospitalized patients, however, is substantial, growing, and costly, and readmissions contribute a significant portion of this burden. Reducing readmission rates of diabetic patients has the potential to greatly reduce health care costs while simultaneously improving care. Risk factors for readmission in this population include lower socioeconomic status, racial/ethnic minority, comorbidity burden, public insurance, emergent or urgent admission, and a history of recent prior hospitalization. Hospitalized patients with diabetes may be at higher risk of readmission than those without diabetes. Potential ways to reduce readmission risk are inpatient education, specialty care, better discharge instructions, coordination of care, and post-discharge support. More studies are needed to test the effect of these interventions on the readmission rates of patients with diabetes.

A tool for prediction of risk of rehospitalisation and mortality in the hospitalised elderly: secondary analysis of clinical trial data

OBJECTIVES

To construct and internally validate a risk score, the '80+ score', for revisits to hospital and mortality for older patients, incorporating aspects of pharmacotherapy. Our secondary aim was to compare the discriminatory ability of the score with that of three validated tools for measuring inappropriate prescribing: Screening Tool of Older Person's Prescriptions (STOPP), Screening Tool to Alert doctors to Right Treatment (START) and Medication Appropriateness Index (MAI).

SETTING

Two acute internal medicine wards at Uppsala University hospital. Patient data were used from a randomised controlled trial investigating the effects of a comprehensive clinical pharmacist intervention.

PARTICIPANTS

Data from 368 patients, aged 80 years and older, admitted to one of the study wards.

PRIMARY OUTCOME MEASURE

Time to rehospitalisation or death during the year after discharge from hospital. Candidate variables were selected among a large number of clinical and drug-specific variables. After a selection process, a score for risk estimation was constructed. The 80+ score was internally validated, and the discriminatory ability of the score and of STOPP, START and MAI was assessed using C-statistics.

RESULTS

Seven variables were selected. Impaired renal function, pulmonary disease, malignant disease, living in a nursing home, being prescribed an opioid or being prescribed a drug for peptic ulcer or gastroesophageal reflux disease were associated with an increased risk, while being prescribed an antidepressant drug (tricyclic antidepressants not included) was linked to a lower risk of the outcome. These variables made up the components of the 80+ score. The C-statistics were 0.71 (80+), 0.57 (STOPP), 0.54 (START) and 0.63 (MAI).

CONCLUSIONS

We developed and internally validated a score for prediction of risk of rehospitalisation and mortality in hospitalised older people. The score discriminated risk better than available tools for inappropriate prescribing. Pending external validation, this score can aid in clinical identification of high-risk patients and targeting of interventions.

Predictors of 30 day hospital readmission in patients with type 2 diabetes: a retrospective, case-control, database study

OBJECTIVE

To assess factors predictive of all-cause, 30 day hospital readmission among patients with type 2 diabetes in the United States.

METHODS

A retrospective, case-control study using deidentified Humedica electronic health record data was conducted to identify patients ≥18 years old with ≥6 months of data prior to index hospitalization (pre-period) and ≥30 days of data after discharge (post-period). Combined methods of bootstrap resampling and stepwise logistic regression were used to identify factors associated with readmission.

RESULTS

Among 52,070 patients with type 2 diabetes and an initial hospitalization for any reason, 5201 (10.0%) patients were readmitted within 30 days and 46,869 (90.0%) patients showed no evidence of readmission. Diabetic treatment escalation; race; type 2 diabetes diagnosis prior to the index stay; pre-period heart failure; and number of pre-period, inpatient healthcare visits were among the strongest predictors of 30 day readmission. From a receiver-operating characteristic plot (mean area under curve of 0.693), the predictive accuracy of the final logistic regression model is considered modest. This result might be due to the unavailability of some variables or data.

CONCLUSIONS

These results highlight the importance of the appropriate recognition of and treatment for type 2 diabetes, prior to and during hospitalization and following discharge, in order to impact a subsequent hospitalization. In our analysis, escalation of diabetic treatments (especially those escalated from having no records of anti-diabetic medications to treatment with insulin) was the strongest predictor of 30 day readmission. Limitations of this study include the fact that hospitalizations and other encounters, outside the Humedica network, were not captured in this analysis.

Postdischarge complications are an important predictor of postoperative readmissions

BACKGROUND

Thirty-day readmissions are common in general surgery patients and affect long-term outcomes including mortality. We sought to determine the effect of complication timing on postoperative readmissions.

METHODS

Patients from our institutional American College of Surgeons National Surgical Quality Improvement Project database who underwent general surgery procedures from 2006 to 2011 were included. The primary outcome of interest was 30-day hospital readmission.

RESULTS

Patients diagnosed with postdischarge complications were significantly more likely to be readmitted (56%) compared with patients diagnosed with complications before discharge (7%, P < .001). Independent predictors of postdischarge complications included laparoscopic case, short hospital stay, preoperative dyspnea, and independent functional status. Gastrointestinal complications and surgical site infection were the most common reasons for readmission.

CONCLUSIONS

The development of complications after hospital discharge places patients at significant risk for readmission. Early identification and treatment of gastrointestinal complications and surgical site infections in the outpatient setting may decrease postoperative readmission rates.

Institution specific risk factors for 30 day readmission at a community hospital: a retrospective observational study

BACKGROUND

As of October 1, 2012, hospitals in the United States with excess readmissions based on the Centers for Medicare and Medicaid Services (CMS) risk-adjusted ratio began being penalized. Given the impact of high readmission rates to hospitals nationally, it is important for individual hospitals to identify which patients may be at highest risk of readmission. The objective of this study was to assess the association of institution specific factors with 30-day readmission.

METHODS

The study is a retrospective observational study using administrative data from January 1, 2009 through December 31, 2010 conducted at a 257 bed community hospital in Massachusetts. The patients included inpatient medical discharges from the hospitalist service with the primary diagnoses of congestive heart failure, pneumonia or chronic obstructive pulmonary disease. The outcome was 30-day readmission rates. After adjusting for known factors that impact readmission, provider associated factors (i.e. hours worked and census on the day of discharge) and hospital associated factors (i.e. floor of discharge, season) were compared.

RESULTS

Over the study time period, there were 3774 discharges by hospitalists, with 637 30-day readmissions (17% readmission rate). By condition, readmission rates were 19.6% (448/2284) for congestive heart failure, 13.0% (141/1083) for pneumonia, and 14.7% (200/1358) for chronic obstructive lung disease. After adjusting for known risk factors (gender, age, length of stay, Elixhauser sum score, admission in the previous year, insurance, disposition, primary diagnosis), we found that patients discharged in the winter remained significantly more likely to be readmitted compared to the summer (OR 1.54, p = 0.0008). Patients discharged from the cardiac floor had a trend toward decreased readmission compared a medical/oncology floor (OR 0.85, p = 0.08). Hospitalist work flow factors (census and hours on the day of discharge) were not associated with readmission.

CONCLUSIONS

We found that 30 day hospital readmissions may be associated with institution specific risk factors, even after adjustment for patient factors. These institution specific risk factors may be targets for interventions to prevent readmissions.

A single-item self-report medication adherence question predicts hospitalisation and death in patients with heart failure

AIMS AND OBJECTIVES

To determine whether a single-item self-report medication adherence question predicts hospitalisation and death in patients with heart failure.

BACKGROUND

Poor medication adherence is associated with increased morbidity and mortality. Having a simple means of identifying suboptimal medication adherence could help identify at-risk patients for interventions.

DESIGN

We performed a prospective cohort study in 592 participants with heart failure within a four-site randomised trial.

METHODS

Self-report medication adherence was assessed at baseline using a single-item question: 'Over the past seven days, how many times did you miss a dose of any of your heart medication?' Participants who reported no missing doses were defined as fully adherent, and those missing more than one dose were considered less than fully adherent. The primary outcome was combined all-cause hospitalisation or death over one year and the secondary endpoint was heart failure hospitalisation. Outcomes were assessed with blinded chart reviews, and heart failure outcomes were determined by a blinded adjudication committee. We used negative binomial regression to examine the relationship between medication adherence and outcomes.

RESULTS

Fifty-two percent of participants were 52% male, mean age was 61 years, and 31% were of New York Heart Association class III/IV at enrolment; 72% of participants reported full adherence to their heart medicine at baseline. Participants with full medication adherence had a lower rate of all-cause hospitalisation and death (0·71 events/year) compared with those with any nonadherence (0·86 events/year): adjusted-for-site incidence rate ratio was 0·83, fully adjusted incidence rate ratio 0·68. Incidence rate ratios were similar for heart failure hospitalisations.

CONCLUSION

A single medication adherence question at baseline predicts hospitalisation and death over one year in heart failure patients.

RELEVANCE TO CLINICAL PRACTICE

Medication adherence is associated with all-cause and heart failure-related hospitalisation and death in heart failure. It is important for clinicians to assess patients' medication adherence on a regular basis at their clinical follow-ups.

Patterns of engagement with the health care system and risk of subsequent hospitalization amongst patients with diabetes

Background

Re-hospitalization is common among patients with diabetes, and may be related to aspects of health care use. We sought to determine the association between patterns of health care engagement and risk of subsequent hospitalization within one year of discharge for patients with diabetes.

Methods

We identified adults with incident diabetes in Alberta, Canada, who had at least one hospitalization following their diabetes diagnosis between January 1, 2004 and March 31, 2011. We used Cox regression to estimate the association between factors related to health care engagement (prior emergency department use, primary care visits, and discharge disposition (i.e. whether the patient left against medical advice)) and the risk of subsequent all-cause hospitalization within one year.

Results

Of the 33811 adults with diabetes and at least one hospitalization, 11095 (32.8%) experienced a subsequent all-cause hospitalization within a mean (standard deviation) follow-up time of 0.68 (0.3) years. Compared to patients with no emergency department visits, there was a 4 percent increased risk of a subsequent hospitalization for every emergency department visit occurring prior to the index hospitalization (adjusted Hazard Ratio [HR]: 1.04; 95% CI: 1.03–1.05). Limited and increased use of primary care was also associated with increased risk of a subsequent hospitalization. Compared to patients with 1–4 visits, patients with no visits to a primary care physician (adjusted HR: 1.11; 95% CI: 0.99–1.25) and those with 5–9 visits (adjusted HR: 1.06; 95% CI: 1.00–1.12) were more likely to experience a subsequent hospitalization. Finally, compared to patients discharged home, those leaving against medical advice were more likely to have a subsequent hospitalization (adjusted HR: 1.74; 95% CI: 1.50–2.02) and almost 3 times more likely to have a diabetes-specific subsequent event (adjusted HR: 2.86; 95% CI: 1.82–4.49).

Conclusions

Patterns of health care use and the circumstances surrounding hospital discharge are associated with an increased risk of subsequent hospitalization among patients with diabetes. Whether these patterns are related to the health care systems ability to manage complex patients within a primary care setting, or to access to primary care services, remains to be determined.

Identifying risk of hospital readmission among Medicare aged patients: an approach using routinely collected data

Readmission provisions in the Patient Protection and Affordable Care Act of March 2010 have created urgent fiscal accountability requirements for hospitals, dependent upon a better understanding of their specific populations, along with development of mechanisms to easily identify these at-risk patients. Readmissions are disruptive and costly to both patients and the health care system. Effectively addressing hospital readmissions among Medicare aged patients offers promising targets for resources aimed at improved quality of care for older patients. Routinely collected data, accessible via electronic medical records, were examined using logistic models of sociodemographic, clinical, and utilization factors to identify predictors among patients who required rehospitalization within 30 days. Specific comorbidities and discharge care orders in this urban, nonprofit hospital had significantly greater odds of predicting a Medicare aged patient's risk of readmission within 30 days.

Telephone follow-up as a primary care intervention for postdischarge outcomes improvement: a systematic review

OBJECTIVE

Postdischarge telephone follow-up plays an integral part in transitional care efforts in many regions. We systematically reviewed the literature to evaluate the evidence regarding the impact of primary care-based telephone follow-up on postdischarge emergency department visits and hospital readmissions.

METHODS

We performed an electronic database search for relevant telephone follow-up studies originating in adult primary care settings.

RESULTS

Only 3 studies (N=1765) met entry criteria for this review. None of the studies demonstrated evidence of reduced admissions or emergency department visits from primary care-based telephone follow-ups. All 3 studies reported improved primary care office contact as a result of telephone follow-up intervention.

CONCLUSIONS

Despite the growing use of primary care-based telephone follow-up in the postdischarge period, there are no high-quality studies demonstrating its benefit. However, its positive impact on patient engagement holds potentially meaningful implications. In light of recent national health care legislation, the primary care field is ripe for high-quality studies to evaluate the effectiveness of telephone follow-up for patients in the postdischarge period. Particular areas of research focus are discussed.

Selecting the best prediction model for readmission

Objectives

This study aims to determine the risk factors predicting rehospitalization by comparing three models and selecting the most successful model.

Methods

In order to predict the risk of rehospitalization within 28 days after discharge, 11 951 inpatients were recruited into this study between January and December 2009. Predictive models were constructed with three methods, logistic regression analysis, a decision tree, and a neural network, and the models were compared and evaluated in light of their misclassification rate, root asymptotic standard error, lift chart, and receiver operating characteristic curve.

Results

The decision tree was selected as the final model. The risk of rehospitalization was higher when the length of stay (LOS) was less than 2 days, route of admission was through the out-patient department (OPD), medical department was in internal medicine, 10th revision of the International Classification of Diseases code was neoplasm, LOS was relatively shorter, and the frequency of OPD visit was greater.

Conclusions

When a patient is to be discharged within 2 days, the appropriateness of discharge should be considered, with special concern of undiscovered complications and co-morbidities. In particular, if the patient is admitted through the OPD, any suspected disease should be appropriately examined and prompt outcomes of tests should be secured. Moreover, for patients of internal medicine practitioners, co-morbidity and complications caused by chronic illness should be given greater attention.

Readmission risk factors after hospital discharge among the elderly

Hospital readmission rates among the elderly are attracting increasing attention. Readmission is costly, especially as proposed new guidelines could deny reimbursement for readmissions. Identifying key factors at discharge that can serve as prognostic indicators for readmission is an important step toward developing and targeting interventions to reduce hospital readmissions rates. Published literature has listed predominantly demographic, clinical, and health care utilization characteristics to describe the factors that put the elderly at risk. However, additional factors are proposed that include social, clinical, individual-level, environmental, and system-level factors. Multimodal interventions have been tested and some reduction in readmissions has been shown. Whether these additional factors might lead to a further reduction remains unclear. In addition to possible factors at discharge, factors identified after the patient has been discharged also must be identified and addressed. The patient safety literature characterizes factors that put the elderly at risk for adverse drug events, which function as antecedent factors for readmission and likely include the environmental and system-level factors. Synthesizing these factors from the readmission and patient safety literature provides the basis to develop a more comprehensive conceptual framework to identify research gaps aimed at reducing hospital readmissions among the elderly.

The relationship between hospital admission rates and rehospitalizations

BACKGROUND

Efforts to reduce hospital readmissions have focused primarily on improving transitional care. Yet variation in readmission rates may more closely reflect variation in the underlying hospitalization rates than differences in the quality of care during and after discharge.

METHODS

We used national Medicare data to calculate, for each local hospital referral region (HRR), the 30-day, 60-day, and 90-day readmission rates among patients discharged with congestive heart failure or pneumonia. We also calculated population-based all-cause admission rates among Medicare enrollees in each HRR. We examined the variation in HRR readmission rates that was explained by overall hospitalization rates versus differences in patients' coexisting conditions, quality of discharge planning, physician supply, and bed supply.

RESULTS

HRR readmission rates ranged from 11 to 32% for congestive heart failure and from 8 to 27% for pneumonia. In univariate analyses, all-cause admission rates accounted for the highest proportion of regional variation in readmission rates for congestive heart failure (28%, 34%, and 37% at 30, 60, and 90 days, respectively); the next highest proportions were explained by case mix (11%, 15%, and 18%) and the number of cardiologists per capita (12%, 14%, and 15%). Results for pneumonia were similar, except that the number of pulmonologists per capita accounted for a lower proportion of the variation (6%, 8%, and 7%, respectively). In multivariate analyses, admission rates accounted for 16 to 24% of the variation for congestive heart failure and 11 to 20% for pneumonia; no other factor accounted for more than 6%.

CONCLUSIONS

We found a substantial association between regional rates of rehospitalization and overall admission rates. Programs directed at shared savings from lower utilization of hospital services might be more successful in reducing readmissions than programs initiated to date. (Funded by the Commonwealth Fund.).

Depression is a risk factor for rehospitalization in medical inpatients

BACKGROUND

Rehospitalization occurs in approximately 20% of medical inpatients within 90 days of discharge. Rehospitalization accounts for considerable morbidity, mortality, and costs. Identification of risk factors could lead to interventions to reduce rehospitalization. The objective of the study was to determine if physical and mental health, substance abuse, and social support are risk factors for rehospitalization.

METHOD

This was a prospective cohort study in an innercity population conducted from September 2002 to September 2004. Participants included 144 adult inpatients with at least 1 hospital admission in the past 6 months. Measurements included age, length of stay, number of admissions in the past year, and medical comorbidity as well as measures of depression, alcohol and drug abuse, social support, and health-related quality of life. The outcome studied was the rehospitalization status of participants within 90 days of the index hospitalization.

RESULTS

The mean age of the subjects was 54.8 years; 48% were black and 78% spoke English as a primary language. Subjects were admitted a mean of 2.5 times in the year before the index admission. Sixty-four patients (44%) were subsequently rehospitalized within 90 days after the index admission. In bivariate analysis, rehospitalized patients had more prior admissions (median of 3.0 vs. 2.0 admissions, p = .002), greater medical comorbidity (mean Charlson Comorbidity Index score of 2.6 vs. 2.0, p = .04), and poorer physical functional status (mean SF-12 physical component score of 31.5 vs. 36.2, p = .03). A logistic regression model, including prior admissions in the last year, comorbidity, physical functional status, and depression, showed that depression tripled the odds of rehospitalization (odds ratio = 3.3, 95% CI = 1.2 to 9.3). This model had fair accuracy in identifying patients at greatest risk for rehospitalization (c statistic = 0.72).

CONCLUSIONS

Hospitalized patients with a history of prior hospitalization within 6 months who screen positive for depression are 3 times more likely to be rehospitalized within 90 days in this relatively high-risk population. Screening during hospitalization for depressive symptoms may identify those at risk for rehospitalization.

Unplanned readmission rates, length of hospital stay, mortality, and medical costs of ten common medical conditions: a retrospective analysis of Hong Kong hospital data

Background

Studies on readmissions attributed to particular medical conditions, especially heart failure, have generally not addressed the factors associated with readmissions and the implications for health outcomes and costs. This study aimed to investigate the factors associated with 30-day unplanned readmission for 10 common conditions and to determine the cost implications.

Methods

This population-based retrospective cohort study included patients admitted to all public hospitals in Hong Kong in 2007. The sample consisted of 337,694 hospitalizations in Internal Medicine. The disease-specific risk-adjusted odd ratio (OR), length of stay (LOS), mortality and attributable medical costs for the year were examined for unplanned readmissions for 10 medical conditions, namely malignant neoplasms, heart diseases, cerebrovascular diseases, pneumonia, injury and poisoning, nephritis and nephrosis, diabetes mellitus, chronic liver disease and cirrhosis, septicaemia, and aortic aneurysm.

Results

The overall unplanned readmission rate was 16.7%. Chronic liver disease and cirrhosis had the highest OR (1.62, 95% confidence interval (CI) 1.39-1.87). Patients with cerebrovascular disease had the longest LOS, with mean acute and rehabilitation stays of 6.9 and 3.0 days, respectively. Malignant neoplasms had the highest mortality rate (30.8%) followed by aortic aneurysm and pneumonia. The attributed medical cost of readmission was highest for heart disease (US$3 199 418, 95% CI US$2 579 443-803 393).

Conclusions

Our findings showed variations in readmission rates and mortality for different medical conditions which may suggest differences in the quality of care provided for various medical conditions. In-hospital care, comprehensive discharge planning, and post-discharge community support for patients need to be reviewed to improve the quality of care and patient health outcomes.