Development of a risk prediction model of potentially avoidable readmission for patients hospitalised with community-acquired pneumonia: study protocol and population

Effect of Exercise Training on Prognosis in Community-acquired Pneumonia: A Randomized Controlled Trial

OBJECTIVE

To investigate the effect of standard care (SoC) combined with supervised in-bed cycling (Bed-Cycle) or booklet exercises (Book-Exe) versus SoC in community-acquired pneumonia (CAP).

METHODS

In this randomized controlled trial, 186 patients with CAP were assigned to SoC (n = 62), Bed-Cycle (n = 61), or Book-Exe (n = 63). Primary outcome length of stay (LOS) was analyzed with analysis of covariance. Secondary outcomes, 90-day readmission, and 180-day mortality were analyzed with Cox proportional hazard regression and readmission days with negative-binominal regression.

RESULTS

LOS was -2% (95% CI: -24 to 25) and -1% (95% CI: -22 to 27) for Bed-Cycle and Book-Exe, compared with SoC. Ninety-day readmission was 35.6% for SoC, 27.6% for Bed-Cycle, and 21.3% for Book-Exe. Adjusted hazard ratio (aHR) for 90-day readmission was 0.63 (95% CI: .33-1.21) and 0.54 (95% CI: .27-1.08) for Bed-Cycle and Book-Exe compared with SoC. aHR for 90-day readmission for combined exercise was 0.59 (95% CI: .33-1.03) compared with SoC. aHR for 180-day mortality was 0.84 (95% CI: .27-2.60) and 0.82 (95% CI: .26-2.55) for Bed-Cycle and Book-Exe compared with SoC. Number of readmission days was 226 for SoC, 161 for Bed-Cycle, and 179 for Book-Exe. Incidence rate ratio for readmission days was 0.73 (95% CI: .48-1.10) and 0.77 (95% CI: .51-1.15) for Bed-Cycle and Book-Exe compared with SoC.

CONCLUSIONS

Although supervised exercise training during admission with CAP did not reduce LOS or mortality, this trial suggests its potential to reduce readmission risk and number of readmission days.

CLINICAL TRIALS REGISTRATION

NCT04094636.

Readmission following hospital admission for community-acquired pneumonia in England

INTRODUCTION

Readmission rates following hospital admission with community-acquired pneumonia (CAP) have increased in the UK over the past decade. The aim of this work was to describe the cohort of patients with emergency 30-day readmission following hospitalisation for CAP in England and explore the reasons for this.

METHODS

A retrospective analysis of cases from the British Thoracic Society national adult CAP audit admitted to hospitals in England with CAP between 1 December 2018 and 31 January 2019 was performed. Cases were linked with corresponding patient level data from Hospital Episode statistics, providing data on the primary diagnosis treated during readmission and mortality. Analyses were performed describing the cohort of patients readmitted within 30 days, reasons for readmission and comparing those readmitted and primarily treated for pneumonia with other diagnoses.

RESULTS

Of 8136 cases who survived an index admission with CAP, 1304 (15.7%) were readmitted as an emergency within 30 days of discharge. The main problems treated on readmission were pneumonia in 516 (39.6%) patients and other respiratory disorders in 284 (21.8%). Readmission with pneumonia compared with all other diagnoses was associated with significant inpatient mortality (15.9% vs 6.5%; aOR 2.76, 95% CI 1.86 to 4.09, p<0.001). A diagnosis of hospital-acquired infection was more frequent in readmissions treated for pneumonia than other diagnoses (22.1% vs 3.9%, p<0.001).

CONCLUSION

Pneumonia is the most common condition treated on readmission following hospitalisation with CAP and carries a higher mortality than both the index admission or readmission due to other diagnoses. Strategies to reduce readmissions due to pneumonia are required.

Machine learning methods to predict 30-day hospital readmission outcome among US adults with pneumonia: analysis of the national readmission database

Background

Hospital readmissions for pneumonia are a growing concern in the US, with significant consequences for costs and quality of care. This study developed the rule-based model and other machine learning (ML) models to predict 30-day readmission risk in patients with pneumonia and compared model performance.

Methods

This population-based study involved patients aged ≥ 18 years hospitalized with pneumonia from January 1, 2016, through November 30, 2016, using the Healthcare Cost and Utilization Project-National Readmission Database (HCUP-NRD). Rule-based algorithms and other ML algorithms, specifically decision trees, random forest, extreme gradient descent boosting (XGBoost), and Least Absolute Shrinkage and Selection Operator (LASSO), were used to model all-cause readmissions 30 days post-discharge from index pneumonia hospitalization. A total of 61 clinically relevant variables were included for ML model development. Models were trained on randomly partitioned 50% of the data and evaluated using the remaining dataset. Model hyperparameters were tuned using the ten-fold cross-validation on the resampled training dataset. The area under the receiver operating curves (AUROC) and area under precision-recall curves (AUPRC) were calculated for the testing set to evaluate the model performance.

Results

Of the 372,293 patients with an index hospital hospitalization for pneumonia, 48,280 (12.97%) were readmitted within 30 days. Judged by AUROC in the testing data, rule-based model (0.6591) significantly outperformed decision tree (0.5783, p value < 0.001), random forest (0.6509, p value < 0.01) and LASSO (0.6087, p value < 0.001), but was less superior than XGBoost (0.6606, p value = 0.015). The AUPRC of the rule-based model in the testing data (0.2146) was higher than the decision tree (0.1560), random forest (0.2052), and LASSO (0.2042), but was similar to XGBoost (0.2147). The top risk-predictive rules captured by the rule-based algorithm were comorbidities, illness severity, disposition locations, payer type, age, and length of stay. These predictive risk factors were also identified by other ML models with high variable importance.

Conclusion

The performance of machine learning models for predicting readmission in pneumonia patients varied. The XGboost was better than the rule-based model based on the AUROC. However, important risk factors for predicting readmission remained consistent across ML models.

Incidence of Avoidable 30-Day Readmissions Following Hospitalization for Community-Acquired Pneumonia in France

IMPORTANCE

Rates of 30-day readmissions following hospitalization for pneumonia are used to publicly report on hospital performance and to set financial penalties for the worst-performing hospitals. However, the rate of avoidable readmission following hospitalization for pneumonia is undefined.

OBJECTIVE

To assess how often 30-day readmissions following hospitalization for community-acquired pneumonia (CAP) are avoidable.

DESIGN, SETTING, AND PARTICIPANTS

This cohort study analyzed the results of an independent review of readmissions following hospitalization for CAP within 30 days among patients discharged from 2 large hospitals in France in 2014. Structured clinical records including clinical information (ie, baseline characteristics, physical examination, laboratory findings, x-ray or computed tomography scan findings, discharge plan, and treatments) for both index and readmission stays were independently reviewed by 4 certified board physicians. All consecutive adult patients hospitalized in 2014 with a diagnosis of CAP in our 2 eligible hospitals were eligible. All analyses presented were performed in March 2021.

MAIN OUTCOMES AND MEASURES

Avoidable readmission within 30 days of discharge from index hospitalization. The likelihood that a readmission was avoidable was quantified using latent class analysis based on the independent reviews. A readmission was considered avoidable if Bayes posterior probability exceeded 50%.

RESULTS

The total analytical sample consisted of 1150 index hospital stays with a diagnosis of CAP, which included 651 (56.6%) male patients. The median (IQR) age for all patients was 77.8 (IQR, 62.7-86.4) years. Out of the 1150 index hospital stays, 98 patients (8.5%) died in hospital, and 108 (9.4%) unplanned readmissions were found. Overall, 15 readmissions had a posterior probability of avoidability exceeding 0.50 (13.9% of the 108 unplanned readmissions; 95% CI, 8.0%-21.9%). The median (IQR) delay between the hospital discharge index and readmission was considerably shorter when readmission was deemed avoidable (4 [6-21] days vs 12 [2-18] days; P = .02).

CONCLUSIONS AND RELEVANCE

Only a small number of readmissions following hospitalization for CAP were deemed avoidable, comprising less than 10% of all readmissions. Shorter time interval between hospitalization discharge and readmission was associated with avoidability.

Current Trends in Readmission Prediction: An Overview of Approaches

Hospital readmission shortly after discharge threatens the quality of patient care and leads to increased medical care costs. In the United States, hospitals with high readmission rates are subject to federal financial penalties. This concern calls for incentives for healthcare facilities to reduce their readmission rates by predicting patients who are at high risk of readmission. Conventional practices involve the use of rule-based assessment scores and traditional statistical methods, such as logistic regression, in developing risk prediction models. The recent advancements in machine learning driven by improved computing power and sophisticated algorithms have the potential to produce highly accurate predictions. However, the value of such models could be overrated. Meanwhile, the use of other flexible models that leverage simple algorithms offer great transparency in terms of feature interpretation, which is beneficial in clinical settings. This work presents an overview of the current trends in risk prediction models developed in the field of readmission. The various techniques adopted by researchers in recent years are described, and the topic of whether complex models outperform simple ones in readmission risk stratification is investigated.