Development of a risk prediction model for hospital-onset Clostridium difficile infection in patients receiving systemic antibiotics

A Simple Prediction Model for Clostridioides difficile Infection: A Hospital-Based Administrative Database Study

BACKGROUND AND AIM

Few prediction scores for Clostridioides difficile infection (CDI), a potentially life-threatening nosocomial diarrhea, combine high accuracy with simplicity. A simple prediction score for routine clinical practice is needed.

METHODS

We conducted a retrospective cohort study of all inpatients aged ≥ 18 at a secondary care hospital in Japan. The derivation and validation cohorts consisted of patients from January 2016 to December 2020 and January 2021 to September 2022, respectively. Demographic and clinical data were retrieved using electronic medical records and an administrative database. The primary outcome was to derive and validate an accurate, simple prediction score for primary hospital-onset CDI. A derived prediction score by logistic regression analysis was calibrated and validated.

RESULTS

CDI developed in 102 of 25 517 and 25 of 6259 patients in the derived and validation cohorts (2.7 cases/10 000 patient-days). The derived model for predicting CDI, including antibiotic use, acid suppressant (proton pump inhibitors or vonoprazan) use, Charlson comorbidity index, and Barthel index, yielded c-statistics of 0.89 and 0.82 in the derivation and validation cohort. The model was well calibrated.

CONCLUSIONS

This simple prediction score enables early medical intervention and modification of treatment plans to reduce the risk of developing primary hospital-onset CDI.

Artificial intelligence applications in the diagnosis and treatment of bacterial infections

The diagnosis and treatment of bacterial infections in the medical and public health field in the 21st century remain significantly challenging. Artificial Intelligence (AI) has emerged as a powerful new tool in diagnosing and treating bacterial infections. AI is rapidly revolutionizing epidemiological studies of infectious diseases, providing effective early warning, prevention, and control of outbreaks. Machine learning models provide a highly flexible way to simulate and predict the complex mechanisms of pathogen-host interactions, which is crucial for a comprehensive understanding of the nature of diseases. Machine learning-based pathogen identification technology and antimicrobial drug susceptibility testing break through the limitations of traditional methods, significantly shorten the time from sample collection to the determination of result, and greatly improve the speed and accuracy of laboratory testing. In addition, AI technology application in treating bacterial infections, particularly in the research and development of drugs and vaccines, and the application of innovative therapies such as bacteriophage, provides new strategies for improving therapy and curbing bacterial resistance. Although AI has a broad application prospect in diagnosing and treating bacterial infections, significant challenges remain in data quality and quantity, model interpretability, clinical integration, and patient privacy protection. To overcome these challenges and, realize widespread application in clinical practice, interdisciplinary cooperation, technology innovation, and policy support are essential components of the joint efforts required. In summary, with continuous advancements and in-depth application of AI technology, AI will enable doctors to more effectivelyaddress the challenge of bacterial infection, promoting the development of medical practice toward precision, efficiency, and personalization; optimizing the best nursing and treatment plans for patients; and providing strong support for public health safety.

Development of a Prediction Model to Identify the Risk of Clostridioides difficile Infection in Hospitalized Patients Receiving at Least One Dose of Antibiotics

OBJECTIVE

This study's objective was to develop a risk-prediction model to identify hospitalized patients at risk of Clostridioides difficile infection (CDI) who had received at least one dose of systemic antibiotics in a large tertiary hospital.

PATIENTS AND METHODS

This was a retrospective case-control study that included patients hospitalized for more than 2 days who received antibiotic therapy during hospitalization. The study included two groups: patients diagnosed with hospital CDI and controls without hospital CDI. Cases were matched 1:3 with assigned controls by age and sex. Descriptive statistics were used to identify the study population by comparing cases with controls. Continuous variables were stated as the means and standard deviations. A multivariate analysis was built to identify the significantly associated covariates between cases and controls for CDI.

RESULTS

A total of 364 patients were included and distributed between the two groups. The control group included 273 patients, and the case group included 91 patients. The risk factors for CDI were investigated, with only significant risks identified and included in the risk assessment model: age older than 70 years (p = 0.034), chronic kidney disease (p = 0.043), solid organ transplantation (p = 0.021), and lymphoma or leukemia (p = 0.019). A risk score of ≥2 showed the best sensitivity, specificity, and accuracy of 78.02%, 45.42%, and 78.02, respectively, with an area under the curve of 0.6172.

CONCLUSION

We identified four associated risk factors in the risk-prediction model. The tool showed good discrimination that might help predict, identify, and evaluate hospitalized patients at risk of developing CDI.

Diagnostic prediction models to identify patients at risk for healthcare-facility-onset Clostridioides difficile: A systematic review of methodology and reporting

OBJECTIVE

To systematically review the methodology, performance, and generalizability of diagnostic models for predicting the risk of healthcare-facility-onset (HO) Clostridioides difficile infection (CDI) in adult hospital inpatients (aged ≥18 years).

BACKGROUND

CDI is the most common cause of healthcare-associated diarrhea. Prediction models that identify inpatients at risk of HO-CDI have been published; however, the quality and utility of these models remain uncertain.

METHODS

Two independent reviewers evaluated articles describing the development and/or validation of multivariable HO-CDI diagnostic models in an inpatient setting. All publication dates, languages, and study designs were considered. Model details (eg, sample size and source, outcome, and performance) were extracted from the selected studies based on the CHARMS checklist. The risk of bias was further assessed using PROBAST.

RESULTS

Of the 3,030 records evaluated, 11 were eligible for final analysis, which described 12 diagnostic models. Most studies clearly identified the predictors and outcomes but did not report how missing data were handled. The most frequent predictors across all models were advanced age, receipt of high-risk antibiotics, history of hospitalization, and history of CDI. All studies reported the area under the receiver operating characteristic curve (AUROC) as a measure of discriminatory ability. However, only 3 studies reported the model calibration results, and only 2 studies were externally validated. All of the studies had a high risk of bias.

CONCLUSION

The studies varied in their ability to predict the risk of HO-CDI. Future models will benefit from the validation on a prospective external cohort to maximize external validity.

A Systematic Literature Review on Risk Factors for and Timing of Clostridioides difficile Infection in the United States

INTRODUCTION

Clostridioides difficile infection (CDI) is a major public health threat. Up to 40% of patients with CDI experience recurrent CDI (rCDI), which is associated with increased morbidity. This study aimed to define an at-risk population by obtaining a detailed understanding of the different factors leading to CDI, rCDI, and CDI-related morbidity and of time to CDI.

METHODS

We conducted a systematic literature review (SLR) of MEDLINE (using PubMed) and EMBASE for relevant articles published between January 1, 2016, and November 11, 2022, covering the US population.

RESULTS

Of the 1324 articles identified, 151 met prespecified inclusion criteria. Advanced patient age was a likely risk factor for primary CDI within a general population, with significant risk estimates identified in nine of 10 studies. Older age was less important in specific populations with comorbidities usually diagnosed at earlier age, such as bowel disease and cancer. In terms of comorbidities, the established factors of infection, kidney disease, liver disease, cardiovascular disease, and bowel disease along with several new factors (including anemia, fluid and electrolyte disorders, and coagulation disorders) were likely risk factors for primary CDI. Data on diabetes, cancer, and obesity were mixed. Other primary CDI risk factors were antibiotics, proton pump inhibitors, female sex, prior hospitalization, and the length of stay in hospital. Similar factors were identified for rCDI, but evidence was limited. Older age was a likely risk factor for mortality. Timing of primary CDI varied depending on the population: 2-3 weeks in patients receiving stem cell transplants, within 3 weeks for patients undergoing surgery, and generally more than 3 weeks following solid organ transplant.

CONCLUSION

This SLR uses recent evidence to define the most important factors associated with CDI, confirming those that are well established and highlighting new ones that could help to identify patient populations at high risk.

An integrated pipeline for prediction of Clostridioides difficile infection

With the expansion of electronic health records(EHR)-linked genomic data comes the development of machine learning-enable models. There is a pressing need to develop robust pipelines to evaluate the performance of integrated models and minimize systemic bias. We developed a prediction model of symptomatic Clostridioides difficile infection(CDI) by integrating common EHR-based and genetic risk factors(rs2227306/IL8). Our pipeline includes (1) leveraging phenotyping algorithm to minimize temporal bias, (2) performing simulation studies to determine the predictive power in samples without genetic information, (3) propensity score matching to control for the confoundings, (4) selecting machine learning algorithms to capture complex feature interactions, (5) performing oversampling to address data imbalance, and (6) optimizing models and ensuring proper bias-variance trade-off. We evaluate the performance of prediction models of CDI when including common clinical risk factors and the benefit of incorporating genetic feature(s) into the models. We emphasize the importance of building a robust integrated pipeline to avoid systemic bias and thoroughly evaluating genetic features when integrated into the prediction models in the general population and subgroups.

Development of Risk Prediction Model for Grade 2 or Higher Hypocalcemia in Bone Metastasis Patients Treated with Denosumab plus Cholecalciferol (Vitamin D3)/Ca Supplement

Denosumab-induced hypocalcemia is sometimes severe, and although a natural vitamin D/Ca combination is used to prevent hypocalcemia, some patients rapidly develop severe hypocalcemia even under supplementation. It is clinically important to predict this risk. This study aimed to develop a risk prediction model for grade 2 or higher hypocalcemia within 28 days after the first denosumab dose under natural vitamin D/Ca supplementation. Using a large database containing multicenter practice data, 2,399 bone metastasis patients who were treated with denosumab between June 2013 and May 2020 were retrospectively analyzed. Background factors in patients who developed grade 2 or higher hypocalcemia within 28 days after the first denosumab dose and those who did not were compared by univariate analysis. Multivariate analysis was conducted to develop a risk prediction model. The model was evaluated for discriminant performance (ROC-AUC: receiver operating characteristic - area under the curve, sensitivity, specificity) and predictive performance (calibration slope). A total of 124 patients in the hypocalcemia group and 1,191 patients in the non-hypocalcemia group were extracted. A risk prediction model consisting of sex, Ca, albumin, alkaline phosphatase, osteoporosis, breast cancer, gastric cancer, proton pump inhibitor combination, and pretreatment with zoledronic acid was developed. The ROC-AUC was 0.87. Sensitivity and specificity were 83% and 81%, respectively, and the calibration slope indicated acceptable agreement between observed and predicted risk. This model appears to be useful to predict the risk of denosumab-induced hypocalcemia and thus should be helpful for risk management of denosumab treatment in patients with bone metastases. This article is protected by copyright. All rights reserved.

Can prediction scores be used to identify patients at risk of Clostridioides difficile infection?

PURPOSE OF REVIEW

To describe the current state of literature on modeling risk of incident and recurrent Clostridioides difficile infection (iCDI and rCDI), to underscore limitations, and to propose a path forward for future research.

RECENT FINDINGS

There are many published risk factors and models for both iCDI and rCDI. The approaches include scores with a limited list of variables designed to be used at the bedside, but more recently have also included automated tools that take advantage of the entire electronic health record. Recent attempts to externally validate scores have met with mixed success.

SUMMARY

For iCDI, the performance largely hinges on the incidence, which even for hospitalized patients can be low (often <1%). Most scores fail to achieve high accuracy and/or are not externally validated. A challenge in predicting rCDI is the significant overlap with risk factors for iCDI, reducing the discriminatory ability of models. Automated electronic health record-based tools show promise but portability to other centers is challenging. Future studies should include external validation and consider biomarkers to augment performance.

Evaluation of a risk assessment model to predict infection with healthcare facility-onset Clostridioides difficile

Disclaimer

In an effort to expedite the publication of articles related to the COVID-19 pandemic, AJHP is posting these manuscripts online as soon as possible after acceptance. Accepted manuscripts have been peer-reviewed and copyedited, but are posted online before technical formatting and author proofing. These manuscripts are not the final version of record and will be replaced with the final article (formatted per AJHP style and proofed by the authors) at a later time.

Purpose

We evaluated a previously published risk model (Novant model) to identify patients at risk for healthcare facility–onset Clostridioides difficile infection (HCFO-CDI) at 2 hospitals within a large health system and compared its predictive value to that of a new model developed based on local findings.

Methods

We conducted a retrospective case-control study including adult patients admitted from July 1, 2016, to July 1, 2018. Patients with HCFO-CDI who received systemic antibiotics were included as cases and were matched 1 to 1 with controls (who received systemic antibiotics without developing HCFO-CDI). We extracted chart data on patient risk factors for CDI, including those identified in prior studies and those included in the Novant model. We applied the Novant model to our patient population to assess the model’s utility and generated a local model using logistic regression–based prediction scores. A receiver operating characteristic area under the curve (ROC-AUC) score was determined for each model.

Results

We included 362 patients, with 161 controls and 161 cases. The Novant model had a ROC-AUC of 0.62 in our population. Our local model using risk factors identifiable at hospital admission included hospitalization within 90 days of admission (adjusted odds ratio [OR], 3.52; 95% confidence interval [CI], 2.06-6.04), hematologic malignancy (adjusted OR, 12.87; 95% CI, 3.70-44.80), and solid tumor malignancy (adjusted OR, 4.76; 95% CI, 1.27-17.80) as HCFO-CDI predictors and had a ROC-AUC score of 0.74.

Conclusion

The Novant model evaluating risk factors identifiable at admission poorly predicted HCFO-CDI in our population, while our local model was a fair predictor. These findings highlight the need for institutions to review local risk factors to adjust modeling for their patient population.

Effectiveness of Oral Vancomycin for Prevention of Healthcare Facility-Onset Clostridioides difficile Infection in Targeted Patients During Systemic Antibiotic Exposure

BACKGROUND

Limited retrospective data suggest prophylactic oral vancomycin may prevent Clostridioides difficile infection (CDI). We sought to evaluate the effectiveness of oral vancomycin for the prevention of healthcare facility-onset CDI (HCFO-CDI) in targeted patients.

METHODS

We conducted a randomized, prospective, open-label study at Novant Health Forsyth Medical Center in Winston-Salem, North Carolina, between October 2018 and April 2019. Included patients were randomized 1:1 to either oral vancomycin (dosed at 125 mg once daily while receiving systemic antibiotics and continued for 5 days postcompletion of systemic antibiotics [OVP]) or no prophylaxis. The primary endpoint was incidence of HCFO-CDI. Secondary endpoints included incidence of community-onset healthcare facility-associated CDI (CO-HCFA-CDI), incidence of vancomycin-resistant Enterococci (VRE) colonization after receiving OVP, adverse effects, and cost of OVP.

RESULTS

A total of 100 patients were evaluated, 50 patients in each arm. Baseline and hospitalization characteristics were similar, except antibiotic exposure. No events of HCFO-CDI were noted in the OVP group compared with 6 (12%) in the no-prophylaxis group (P = .03). CO-HCFA-CDI was identified in 2 patients who were previously diagnosed with HCFO-CDI. No patients developed new VRE colonization, with only 1 patient reporting mild gastrointestinal side effects to OVP. A total of 600 doses of OVP were given during the study, with each patient receiving an average of 12 doses. Total acquisition cost of OVP was $1302, $26.04 per patient.

CONCLUSION

OVP appears to protect against HCFO-CDI during in-patient stay in targeted patients during systemic antibiotic exposure. Further prospective investigation is warranted.

A Narrative Review of Antimicrobial Stewardship Interventions within In-patient Settings and Resultant Patient Outcomes

The overuse of antibiotics has led to various healthcare problems such as the emergence of resistance in infectious microbes and mortality due to antibiotic resistant healthcare associated infections (HAIs). An antimicrobial stewardship (AMS) program is the set of interventions used worldwide to enhance the rational use of antibiotics especially for the hospitalized patients. This review aimed to describe the characteristics of the implemented AMS programs in various hospitals of the world mainly focusing on the interventions and patients outcomes. The literature about AMS program was searched through various databases such as PubMed, Google Scholar, Science Direct, Cochran Library, Ovid (Medline), Web of Science and Scopus. In this review the literature pertaining to the AMS programs for hospitalized patients is sorted on the basis of various interventions that are categorized as formulary restriction (pre-authorization), guideline development, clinical pathway development, educative interventions and prospective audit. Moreover a clear emphasis is laid on the patient outcomes obtained as a result of these interventions namely the infection control, drop in readmission rate, mortality control, resistance control and the control of an overall cost of antibiotic treatment obtained mainly by curbing the overuse of antibiotics within the hospital wards. AMS program is an efficient strategy of pharmacovigilance to rationalize the antimicrobial practice for hospitalized patients as it prevents the misuse of antibiotics, which ultimately retards the health threatening effects of various antibiotics.

Modest Clostridiodes difficile infection prediction using machine learning models in a tertiary care hospital

Previous studies have shown promising results of machine learning (ML) models for predicting health outcomes. We develop and test ML models for predicting Clostridioides difficile infection (CDI) in hospitalized patients. This is a retrospective cohort study conducted during 2015-2017. All inpatients tested for C. difficile were included. CDI was defined as having a positive glutamate dehydrogenase and toxin results. We restricted analyses to the first record of C. difficile testing per patient. Of 3514 patients tested, 136 (4%) had CDI. Age and antibiotic use within 90 days before C. difficile testing were associated with CDI (P < 0.01). We tested 10 ML methods with and without resampling. Logistic regression, random forest and naïve Bayes models yielded the highest AUC ROC performance: 0.6. Predicting CDI was difficult in our cohort of patients tested for CDI. Multiple ML models yielded only modest results in a real-world population of hospitalized patients tested for CDI.

Risk factors for contagious gastroenteritis in adult patients with diarrhoea in the emergency department - a prospective observational multicentre study

Background

Infectious gastroenteritis is common in the emergency department (ED). Patients infected with either Norovirus or toxigenic Clostridium difficile require special isolation procedures. The aims were to describe the aetiology of infectious gastroenteritis in the ED, evaluate whether current isolation procedures, based on clinical judgement are sufficient, and to identify information that might be used to identify patients requiring isolation.

Methods

Prospective, observational, multicentre study. We collected information on symptoms, vital signs, travel history, the recent use of antibiotics, and infectious contacts and tested faecal samples for Norovirus, C. difficile, and enteropathogenic bacteria.

Results

The study enrolled 227 patients, of whom 163 (71%) delivered a faecal sample for Norovirus analysis (13% positive), 171 (74%) for C. difficile (13% positive), and 173 (76%) for enteropathogenic bacteria (16% positive). In total 71% of the patients were isolated using strict precautions, 29% of the isolated patient and 14% of the patients who were not isolated had had a highly contagious GE. Risk factors for Norovirus included frequent vomiting (OR 5.5), recent admission of another patient with Norovirus (OR 2.6), and a short duration of diarrhoea. Risk factors for C. difficile infections included older age (OR 6.0), longer duration of diarrhoea (OR 5.2), mucus in stool (OR 3.5), and previous antibiotic use (OR 23.4).

Conclusion

Highly contagious GE occurs in ¼ of the GE patients in the EDs, isolation based on clinical judgement is not very efficient. Several risk factors can predict the presence of Norovirus or toxigenic Clostridium difficile. It is uncertain whether this knowledge can improve isolation practices in ED settings.

Trial registration

This study was retrospectively registered in the Clinical Trials Data Base (NCT02685527) and prospectively approved by the Regional Committees on Health Research Ethics for Southern Denmark (project ID S20140200) and Ethics Committee at the Medical Association of Schleswig-Holstein [“Ethikkommission bei der Ärztekammer Schleswig-Holstein”, project ID 120/15(I)] and registered with the Danish Data Protection Agency (project ID nr. 2008-58-0035/ 1608).