Novel Method to Flag Cardiac Implantable Device Infections by Integrating Text Mining With Structured Data in the Veterans Health Administration's Electronic Medical Record

Innovative Techniques for Infection Control and Surveillance in Hospital Settings and Long-Term Care Facilities: A Scoping Review

Healthcare-associated infections (HAIs) pose significant challenges in healthcare systems, with preventable surveillance playing a crucial role. Traditional surveillance, although effective, is resource-intensive. The development of new technologies, such as artificial intelligence (AI), can support traditional surveillance in analysing an increasing amount of health data or meeting patient needs. We conducted a scoping review, following the PRISMA-ScR guideline, searching for studies of new digital technologies applied to the surveillance, control, and prevention of HAIs in hospitals and LTCFs published from 2018 to 4 November 2023. The literature search yielded 1292 articles. After title/abstract screening and full-text screening, 43 articles were included. The mean study duration was 43.7 months. Surgical site infections (SSIs) were the most-investigated HAI and machine learning was the most-applied technology. Three main themes emerged from the thematic analysis: patient empowerment, workload reduction and cost reduction, and improved sensitivity and personalization. Comparative analysis between new technologies and traditional methods showed different population types, with machine learning methods examining larger populations for AI algorithm training. While digital tools show promise in HAI surveillance, especially for SSIs, challenges persist in resource distribution and interdisciplinary integration in healthcare settings, highlighting the need for ongoing development and implementation strategies.

Factors associated with uptake of guideline-recommended cardiovascular implantable electronic device management: a nationwide, retrospective cohort study

Clinical guidelines recommend device removal for cardiovascular implantable electronic device (CIED) infection management. In this retrospective, nationwide cohort, 60.8% of CIED infections received guideline-concordant care. One-year mortality was higher among those without procedural management (25% vs 16%). Factors associated with receipt of device procedures included pocket infections and positive microbiology.

Use of electronic health record data mining for heart failure subtyping

OBJECTIVE

To assess whether electronic health record (EHR) data text mining can be used to improve register-based heart failure (HF) subtyping. EHR data of 43,405 individuals from two Finnish hospital biobanks were mined for unstructured text mentions of ejection fraction (EF) and validated against clinical assessment in two sets of 100 randomly selected individuals. Structured laboratory data was then incorporated for a categorization by HF subtype (HF with mildly reduced EF, HFmrEF; HF with preserved EF, HFpEF; HF with reduced EF, HFrEF; and no HF).

RESULTS

In 86% of the cases, the algorithm-identified EF belonged to the correct HF subtype range. Sensitivity, specificity, PPV and NPV of the algorithm were 94-100% for HFrEF, 85-100% for HFmrEF, and 96%, 67%, 53% and 98% for HFpEF. Survival analyses using the traditional diagnosis of HF were in concordance with the algorithm-based ones. Compared to healthy individuals, mortality increased from HFmrEF (hazard ratio [HR], 1.91; 95% confidence interval [CI], 1.24-2.95) to HFpEF (2.28; 1.80-2.88) to HFrEF group (2.63; 1.97-3.50) over a follow-up of 1.5 years. We conclude that quantitative EF data can be efficiently extracted from EHRs and used with laboratory data to subtype HF with reasonable accuracy, especially for HFrEF.

A Novel Informatics Tool to Detect Periprocedural Antibiotic Allergy Adverse Events for Near Real-time Surveillance to Support Audit and Feedback

Importance

Standardized processes for identifying when allergic-type reactions occur and linking reactions to drug exposures are limited.

Objective

To develop an informatics tool to improve detection of antibiotic allergic-type events.

Design, Setting, and Participants

This retrospective cohort study was conducted from October 1, 2015, to September 30, 2019, with data analyzed between July 1, 2021, and January 31, 2022. The study was conducted across Veteran Affairs hospitals among patients who underwent cardiovascular implantable electronic device (CIED) procedures and received periprocedural antibiotic prophylaxis. The cohort was split into training and test cohorts, and cases were manually reviewed to determine presence of allergic-type reaction and its severity. Variables potentially indicative of allergic-type reactions were selected a priori and included allergies entered in the Veteran Affair's Allergy Reaction Tracking (ART) system (either historical [reported] or observed), allergy diagnosis codes, medications administered to treat allergic reactions, and text searches of clinical notes for keywords and phrases indicative of a potential allergic-type reaction. A model to detect allergic-type reaction events was iteratively developed on the training cohort and then applied to the test cohort. Algorithm test characteristics were assessed.

Exposure

Preprocedural and postprocedural prophylactic antibiotic administration.

Main Outcomes and Measures

Antibiotic allergic-type reactions.

Results

The cohort of 36 344 patients included 34 703 CIED procedures with antibiotic exposures (mean [SD] age, 72 [10] years; 34 008 [98%] male patients); median duration of postprocedural prophylaxis was 4 days (IQR, 2-7 days; maximum, 45 days). The final algorithm included 7 variables: entries in the Veteran Affair's hospitals ART, either historic (odds ratio [OR], 42.37; 95% CI, 11.33-158.43) or observed (OR, 175.10; 95% CI, 44.84-683.76); PheCodes for "symptoms affecting skin" (OR, 8.49; 95% CI, 1.90-37.82), "urticaria" (OR, 7.01; 95% CI, 1.76-27.89), and "allergy or adverse event to an antibiotic" (OR, 11.84, 95% CI, 2.88-48.69); keyword detection in clinical notes (OR, 3.21; 95% CI, 1.27-8.08); and antihistamine administration alone or in combination (OR, 6.51; 95% CI, 1.90-22.30). In the final model, antibiotic allergic-type reactions were identified with an estimated probability of 30% or more; positive predictive value was 61% (95% CI, 45%-76%); and sensitivity was 87% (95% CI, 70%-96%).

Conclusions and Relevance

In this retrospective cohort study of patients receiving periprocedural antibiotic prophylaxis, an algorithm with a high sensitivity to detect incident antibiotic allergic-type reactions that can be used to provide clinician feedback about antibiotic harms from unnecessarily prolonged antibiotic exposures was developed.

Development of a Periprocedure Trigger for Outpatient Interventional Radiology Procedures in the Veterans Health Administration

OBJECTIVES

Interventional radiology (IR) is the newest medical specialty. However, it lacks robust quality assurance metrics, including adverse event (AE) surveillance tools. Considering the high frequency of outpatient care provided by IR, automated electronic triggers offer a potential catalyst to support accurate retrospective AE detection.

METHODS

We programmed previously validated AE triggers (admission, emergency visit, or death up to 14 days after procedure) for elective, outpatient IR procedures performed in Veterans Health Administration surgical facilities between fiscal years 2017 and 2019. We then developed a text-based algorithm to detect AEs that explicitly occurred in the periprocedure time frame: before, during, and shortly after the IR procedure. Guided by the literature and clinical expertise, we generated clinical note keywords and text strings to flag cases with high potential for periprocedure AEs. Flagged cases underwent targeted chart review to measure criterion validity (i.e., the positive predictive value), to confirm AE occurrence, and to characterize the event.

RESULTS

Among 135,285 elective outpatient IR procedures, the periprocedure algorithm flagged 245 cases (0.18%); 138 of these had ≥1 AE, yielding a positive predictive value of 56% (95% confidence interval, 50%-62%). The previously developed triggers for admission, emergency visit, or death in 14 days flagged 119 of the 138 procedures with AEs (73%). Among the 43 AEs detected exclusively by the periprocedure trigger were allergic reactions, adverse drug events, ischemic events, bleeding events requiring blood transfusions, and cardiac arrest requiring cardiopulmonary resuscitation.

CONCLUSIONS

The periprocedure trigger performed well on IR outpatient procedures and offers a complement to other electronic triggers developed for outpatient AE surveillance.

Automating surveillance for healthcare-associated infections: Rationale and current realities (Part I/III)

Infection surveillance is one of the cornerstones of infection prevention and control. Measurement of process metrics and clinical outcomes, such as detection of healthcare-associated infections (HAIs), can be used to support continuous quality improvement. HAI metrics are reported as part of the CMS Hospital-Acquired Conditions Program, and they influence facility reputation and financial outcomes.

Leveraging electronic data to expand infection detection beyond traditional settings and definitions (Part II/III)

The rich and complex electronic health record presents promise for expanding infection detection beyond currently covered settings of care. Here, we review the "how to" of leveraging electronic data sources to expand surveillance to settings of care and infections that have not been the traditional purview of the National Healthcare Safety Network (NHSN), including a discussion of creation of objective and reproducible infection surveillance definitions. In pursuit of a 'fully automated' system, we also examine the promises and pitfalls of leveraging unstructured, free-text data to support infection prevention activities and emerging technological advances that will likely affect the practice of automated infection surveillance. Finally, barriers to achieving a completely 'automated' infection detection system and challenges with intra- and interfacility reliability and missing data are discussed.

Validating administrative data to identify complex surgical site infections following cardiac implantable electronic device implantation: a comparison of traditional methods and machine learning

Background

Cardiac implantable electronic device (CIED) surgical site infections (SSIs) have been outpacing the increases in implantation of these devices. While traditional surveillance of these SSIs by infection prevention and control would likely be the most accurate, this is not practical in many centers where resources are constrained. Therefore, we explored the validity of administrative data at identifying these SSIs.

Methods

We used a cohort of all patients with CIED implantation in Calgary, Alberta where traditional surveillance was done for infections from Jan 1, 2013 to December 31, 2019. We used this infection subgroup as our “gold standard” and then utilized various combinations of administrative data to determine which best optimized the sensitivity and specificity at identifying infection. We evaluated six approaches to identifying CIED infection using administrative data, which included four algorithms using International Classification of Diseases codes and/or Canadian Classification of Health Intervention codes, and two machine learning models. A secondary objective of our study was to assess if machine learning techniques with training of logistic regression models would outperform our pre-selected codes.

Results

We determined that all of the pre-selected algorithms performed well at identifying CIED infections but the machine learning model was able to produce the optimal method of identification with an area under the receiver operating characteristic curve (AUC) of 96.8%. The best performing pre-selected algorithm yielded an AUC of 94.6%.

Conclusions

Our findings suggest that administrative data can be used to effectively identify CIED infections. While machine learning performed the most optimally, in centers with limited analytic capabilities a simpler algorithm of pre-selected codes also has excellent yield. This can be valuable for centers without traditional surveillance to follow trends in SSIs over time and identify when rates of infection are increasing. This can lead to enhanced interventions for prevention of SSIs.

Assessment of Electronic Health Record for Cancer Research and Patient Care Through a Scoping Review of Cancer Natural Language Processing

PURPOSE

The advancement of natural language processing (NLP) has promoted the use of detailed textual data in electronic health records (EHRs) to support cancer research and to facilitate patient care. In this review, we aim to assess EHR for cancer research and patient care by using the Minimal Common Oncology Data Elements (mCODE), which is a community-driven effort to define a minimal set of data elements for cancer research and practice. Specifically, we aim to assess the alignment of NLP-extracted data elements with mCODE and review existing NLP methodologies for extracting said data elements.

METHODS

Published literature studies were searched to retrieve cancer-related NLP articles that were written in English and published between January 2010 and September 2020 from main literature databases. After the retrieval, articles with EHRs as the data source were manually identified. A charting form was developed for relevant study analysis and used to categorize data including four main topics: metadata, EHR data and targeted cancer types, NLP methodology, and oncology data elements and standards.

RESULTS

A total of 123 publications were selected finally and included in our analysis. We found that cancer research and patient care require some data elements beyond mCODE as expected. Transparency and reproductivity are not sufficient in NLP methods, and inconsistency in NLP evaluation exists.

CONCLUSION

We conducted a comprehensive review of cancer NLP for research and patient care using EHRs data. Issues and barriers for wide adoption of cancer NLP were identified and discussed.

Data Science Trends Relevant to Nursing Practice: A Rapid Review of the 2020 Literature

BACKGROUND

The term "data science" encompasses several methods, many of which are considered cutting edge and are being used to influence care processes across the world. Nursing is an applied science and a key discipline in health care systems in both clinical and administrative areas, making the profession increasingly influenced by the latest advances in data science. The greater informatics community should be aware of current trends regarding the intersection of nursing and data science, as developments in nursing practice have cross-professional implications.

OBJECTIVES

This study aimed to summarize the latest (calendar year 2020) research and applications of nursing-relevant patient outcomes and clinical processes in the data science literature.

METHODS

We conducted a rapid review of the literature to identify relevant research published during the year 2020. We explored the following 16 topics: (1) artificial intelligence/machine learning credibility and acceptance, (2) burnout, (3) complex care (outpatient), (4) emergency department visits, (5) falls, (6) health care-acquired infections, (7) health care utilization and costs, (8) hospitalization, (9) in-hospital mortality, (10) length of stay, (11) pain, (12) patient safety, (13) pressure injuries, (14) readmissions, (15) staffing, and (16) unit culture.

RESULTS

Of 16,589 articles, 244 were included in the review. All topics were represented by literature published in 2020, ranging from 1 article to 59 articles. Numerous contemporary data science methods were represented in the literature including the use of machine learning, neural networks, and natural language processing.

CONCLUSION

This review provides an overview of the data science trends that were relevant to nursing practice in 2020. Examinations of such literature are important to monitor the status of data science's influence in nursing practice.

Promoting de-implementation of inappropriate antimicrobial use in cardiac device procedures by expanding audit and feedback: protocol for hybrid III type effectiveness/implementation quasi-experimental study

Background

Despite a strong evidence base and clinical guidelines specifically recommending against prolonged post-procedural antimicrobial use, studies indicate that the practice is common following cardiac device procedures. Formative evaluations conducted by the study team suggest that inappropriate antimicrobial use may be driven by information silos that drive provider belief that antimicrobials are not harmful, in part due to lack of complete feedback about all types of clinical outcomes. De-implementation is recognized as an important area of research that can lead to reductions in unnecessary, wasteful, or harmful practices, such as excess antimicrobial use following cardiac device procedures; however, investigations into strategies that lead to successful de-implementation are limited. The overarching hypothesis to be tested in this trial is that a bundle of implementation strategies that includes audit and feedback about direct patient harms caused by inappropriate prescribing can lead to successful de-implementation of guideline-discordant care.

Methods

We propose a hybrid type III effectiveness-implementation stepped-wedge intervention trial at three high-volume, high-complexity VA medical centers. The main study intervention (an informatics-based, real-time audit-and-feedback tool) was developed based on learning/unlearning theory and formative evaluations and guided by the integrated-Promoting Action on Research Implementation in Health Services (i-PARIHS) Framework. Elements of the bundled and multifaceted implementation strategy to promote appropriate prescribing will include audit-and-feedback reports that include information about antibiotic harms, stakeholder engagement, patient and provider education, identification of local champions, and blended facilitation. The primary study outcome is adoption of evidence-based practice (de-implementation of inappropriate antimicrobial use). Clinical outcomes (cardiac device infections, acute kidney injuries and Clostridioides difficile infections) are secondary. Qualitative interviews will assess relevant implementation outcomes (acceptability, adoption, fidelity, feasibility).

Discussion

De-implementation theory suggests that factors that may have a particularly strong influence on de-implementation include strength of the underlying evidence, the complexity of the intervention, and patient and provider anxiety and fear about changing an established practice. This study will assess whether a multifaceted intervention mapped to identified de-implementation barriers leads to measurable improvements in provision of guideline-concordant antimicrobial use. Findings will improve understanding about factors that impact successful or unsuccessful de-implementation of harmful or wasteful healthcare practices.

Trial registration

ClinicalTrials.govNCT05020418

Incidence and Predictors of Infections and All-Cause Death in Patients with Cardiac Implantable Electronic Devices: The Italian Nationwide RI-AIAC Registry

Background: The incidence of infections associated with cardiac implantable electronic devices (CIEDs) and patient outcomes are not fully known. Aim: To provide a contemporary assessment of the risk of CIEDs infection and associated clinical outcomes. Methods: In Italy, 18 centres enrolled all consecutive patients undergoing a CIED procedure and entered a 12-months follow-up. CIED infections, as well as a composite clinical event of infection or all-cause death were recorded. Results: A total of 2675 patients (64.3% male, age 78 (70–84)) were enrolled. During follow up 28 (1.1%) CIED infections and 132 (5%) deaths, with 152 (5.7%) composite clinical events were observed. At a multivariate analysis, the type of procedure (revision/upgrading/reimplantation) (OR: 4.08, 95% CI: 1.38–12.08) and diabetes (OR: 2.22, 95% CI: 1.02–4.84) were found as main clinical factors associated to CIED infection. Both the PADIT score and the RI-AIAC Infection score were significantly associated with CIED infections, with the RI-AIAC infection score showing the strongest association (OR: 2.38, 95% CI: 1.60–3.55 for each point), with a c-index = 0.64 (0.52–0.75), p = 0.015. Regarding the occurrence of composite clinical events, the Kolek score, the Shariff score and the RI-AIAC Event score all predicted the outcome, with an AUC for the RI-AIAC Event score equal to 0.67 (0.63−0.71) p < 0.001. Conclusions: In this Italian nationwide cohort of patients, while the incidence of CIED infections was substantially low, the rate of the composite clinical outcome of infection or all-cause death was quite high and associated with several clinical factors depicting a more impaired clinical status.

Extracting Significant Comorbid Diseases from MeSH Index of PubMed

Text mining is an important research area to be explored in terms of understanding disease associations and have an insight in disease comorbidities. The reason for comorbid occurrence in any patient may be genetic or molecular interference from any other processes. Comorbidity and multimorbidity may be technically different, yet still are inseparable in studies. They have overlapping nature of associations and hence can be integrated for a more rational approach. The association rule generally used to determine comorbidity may also be helpful in novel knowledge prediction or may even serve as an important tool of assessment in surgical cases. Another approach of interest may be to utilize biological vocabulary resources like UMLS/MeSH across a patient health information and analyze the interrelationship between different health conditions. The protocol presented here can be utilized for understanding the disease associations and analyze at an extensive level.

Association Between Diabetic Foot Infection Wound Culture Positivity and 1-Year Admission for Invasive Infection: A Multicenter Cohort Study

Inpatients with culture-positive diabetic foot infections are at elevated risk for subsequent invasive infection with the same causative organism. In outpatients with index diabetic foot ulcers, we found that wound culture positivity was independently associated with increased odds of 1-year admission for systemic infection when compared with culture-negative wounds.

Prevention of cardiac implantable electronic device infections: guidelines and conventional prophylaxis

Cardiac implantable electronic devices (CIED) are potentially life-saving treatments for several cardiac conditions, but are not without risk. Despite dissemination of recommended strategies for prevention of device infections, such as administration of antibiotics before implantation, infection rates continue to rise resulting in escalating health care costs. New trials conveying important steps for better prevention of device infection and an EHRA consensus paper were recently published. This document will review the role of various preventive measures for CIED infection, emphasizing the importance of adhering to published recommendations. The document aims to provide guidance on how to prevent CIED infections in clinical practice by considering modifiable and non-modifiable risk factors that may be present pre-, peri-, and/or post-procedure.

Development of a fully automated surgical site infection detection algorithm for use in cardiac and orthopedic surgery research

Objective:

To develop a fully automated algorithm using data from the Veterans’ Affairs (VA) electrical medical record (EMR) to identify deep-incisional surgical site infections (SSIs) after cardiac surgeries and total joint arthroplasties (TJAs) to be used for research studies.

Design:

Retrospective cohort study.

Setting:

This study was conducted in 11 VA hospitals.

Participants:

Patients who underwent coronary artery bypass grafting or valve replacement between January 1, 2010, and March 31, 2018 (cardiac cohort) and patients who underwent total hip arthroplasty or total knee arthroplasty between January 1, 2007, and March 31, 2018 (TJA cohort).

Methods:

Relevant clinical information and administrative code data were extracted from the EMR. The outcomes of interest were mediastinitis, endocarditis, or deep-incisional or organ-space SSI within 30 days after surgery. Multiple logistic regression analysis with a repeated regular bootstrap procedure was used to select variables and to assign points in the models. Sensitivities, specificities, positive predictive values (PPVs) and negative predictive values were calculated with comparison to outcomes collected by the Veterans’ Affairs Surgical Quality Improvement Program (VASQIP).

Results:

Overall, 49 (0.5%) of the 13,341 cardiac surgeries were classified as mediastinitis or endocarditis, and 83 (0.6%) of the 12,992 TJAs were classified as deep-incisional or organ-space SSIs. With at least 60% sensitivity, the PPVs of the SSI detection algorithms after cardiac surgeries and TJAs were 52.5% and 62.0%, respectively.

Conclusions:

Considering the low prevalence rate of SSIs, our algorithms were successful in identifying a majority of patients with a true SSI while simultaneously reducing false-positive cases. As a next step, validation of these algorithms in different hospital systems with EMR will be needed.