Computer algorithms to detect bloodstream infections

Colorectal cancer and association with anaerobic bacteraemia: A Danish nationwide population-based cohort study

OBJECTIVES

We aimed to identify specific anaerobic bacteria causing bacteraemia and a subsequent diagnosis of colorectal cancer.

METHODS

A nationwide population-based cohort study, which included all episodes of defined specific anaerobic bacteraemia from 2010 (5,534,738 inhabitants) through 2020 (5,822,763 inhabitants) and all cases of colorectal cancer diagnosed from 2010 through 2021 in Denmark. We calculated the incidence and risk of colorectal cancer after bacteraemia with specific anaerobic bacteria using Escherichia coli bacteraemia as reference.

RESULTS

Nationwide data on colorectal cancer and specific anaerobic bacteraemia (100% complete, representing 11,124 episodes). The frequencies of colorectal cancer within one year following anaerobic bacteraemia were higher for species, which almost exclusively reside in the colon, such as Phocaeicola vulgatus/dorei (5.5%), Clostridium septicum (24.2%), and Ruminococcus gnavus (4.6%) compared to 0.6% in 50,650 E. coli bacteraemia episodes. Bacteroides spp. had a subhazard ratio for colorectal cancer of 3.9 (95% confidence interval [CI], 3.0 to 5.1) and for Clostridium spp. it was 8.9 (95% CI, 6.7 to 11.8, with C. septicum 50.0 [95% CI, 36.0 to 69.5]) compared to E. coli (reference).

CONCLUSION

This nationwide study identified specific colorectal cancer-associated anaerobic bacteria, which almost exclusively reside in the colon. Bacteraemia with these bacteria could be an indicator of colorectal cancer.

Multirole of the internet of medical things (IoMT) in biomedical systems for managing smart healthcare systems: An overview of current and future innovative trends

Internet of Medical Things (IoMT) is an emerging subset of Internet of Things (IoT), often called as IoT in healthcare, refers to medical devices and applications with internet connectivity, is exponentially gaining researchers' attention due to its wide-ranging applicability in biomedical systems for Smart Healthcare systems. IoMT facilitates remote health biomedical system and plays a crucial role within the healthcare industry to enhance precision, reliability, consistency and productivity of electronic devices used for various healthcare purposes. It comprises a conceptualized architecture for providing information retrieval strategies to extract the data from patient records using sensors for biomedical analysis and diagnostics against manifold diseases to provide cost-effective medical solutions, quick hospital treatments, and personalized healthcare. This article provides a comprehensive overview of IoMT with special emphasis on its current and future trends used in biomedical systems, such as deep learning, machine learning, blockchains, artificial intelligence, radio frequency identification, and industry 5.0.

Systematic scoping review of automated systems for the surveillance of healthcare-associated bloodstream infections related to intravascular catheters

INTRODUCTION

Intravascular catheters are crucial devices in medical practice that increase the risk of healthcare-associated infections (HAIs), and related health-economic adverse outcomes. This scoping review aims to provide a comprehensive overview of published automated algorithms for surveillance of catheter-related bloodstream infections (CRBSI) and central line-associated bloodstream infections (CLABSI).

METHODS

We performed a scoping review based on a systematic search of the literature in PubMed and EMBASE from 1 January 2000 to 31 December 2021. Studies were included if they evaluated predictive performance of automated surveillance algorithms for CLABSI/CRBSI detection and used manually collected surveillance data as reference. We assessed the design of the automated systems, including the definitions used to develop algorithms (CLABSI versus CRBSI), the datasets and denominators used, and the algorithms evaluated in each of the studies.

RESULTS

We screened 586 studies based on title and abstract, and 99 were assessed based on full text. Nine studies were included in the scoping review. Most studies were monocentric (n = 5), and they identified CLABSI (n = 7) as an outcome. The majority of the studies used administrative and microbiological data (n = 9) and five studies included the presence of a vascular central line in their automated system. Six studies explained the denominator they selected, five of which chose central line-days. The most common rules and steps used in the algorithms were categorized as hospital-acquired rules, infection rules (infection versus contamination), deduplication, episode grouping, secondary BSI rules (secondary versus primary BSI), and catheter-associated rules.

CONCLUSION

The automated surveillance systems that we identified were heterogeneous in terms of definitions, datasets and denominators used, with a combination of rules in each algorithm. Further guidelines and studies are needed to develop and implement algorithms to detect CLABSI/CRBSI, with standardized definitions, appropriate data sources and suitable denominators.

The accuracy of fully-automated algorithms for the surveillance of central venous catheter-related bloodstream infection in hospitalised patients

BACKGROUND

Continuous surveillance for healthcare-associated infections such as central venous catheter-related bloodstream infections (CVC-BSI) is crucial for prevention. However, traditional surveillance methods are resource-intensive and prone to bias. This study aimed to develop and validate fully-automated surveillance algorithms for CVC-BSI.

METHODS

Two algorithms were developed using electronic health record data from 1000 admissions with a positive blood culture (BCx) at Karolinska University Hospital from 2017: (1) Combining microbiological findings in BCx and CVC cultures with BSI symptoms; (2) Only using microbiological findings. These algorithms were validated in 5170 potential CVC-BSI-episodes from all admissions in 2018-2019, and results extrapolated to all potential CVC-BSI-episodes within this period (n = 181,354). The reference standard was manual record review according to ECDC's definition of microbiologically confirmed CVC-BSI (CRI3-CVC).

RESULTS

In the potential CVC-BSI-episodes, 51 fulfilled ECDC's definition and the algorithms identified 47 and 49 episodes as CVC-BSI, respectively. Both algorithms performed well in assessing CVC-BSI. Overall, algorithm 2 performed slightly better with in the total period a sensitivity of 0.880 (95%-CI 0.783-0.959), specificity of 1.000 (95%-CI 0.999-1.000), PPV of 0.918 (95%-CI 0.833-0.981) and NPV of 1.000 (95%-CI 0.999-1.000). Incidence according to the reference and algorithm 2 was 0.33 and 0.31 per 1000 in-patient hospital-days, respectively.

CONCLUSIONS

Both fully-automated surveillance algorithms for CVC-BSI performed well and could effectively replace manual surveillance. The simpler algorithm, using only microbiology data, is suitable when BCx testing adheres to recommendations, otherwise the algorithm using symptom data might be required. Further validation in other settings is necessary to assess the algorithms' generalisability.

Healthcare-associated infections and conditions in the era of digital measurement

As the third edition of the Compendium of Strategies to Prevent Healthcare-Associated Infections in Acute Care Hospitals is released with the latest recommendations for the prevention and management of healthcare-associated infections (HAIs), a new approach to reporting HAIs is just beginning to unfold. This next generation of HAI reporting will be fully electronic and based largely on existing data in electronic health record (EHR) systems and other electronic data sources. It will be a significant change in how hospitals report HAIs and how the Centers for Disease Control and Prevention (CDC) and other agencies receive this information. This paper outlines what that future electronic reporting system will look like and how it will impact HAI reporting.

Predictive performance of automated surveillance algorithms for intravascular catheter bloodstream infections: a systematic review and meta-analysis

BACKGROUND

Intravascular catheter infections are associated with adverse clinical outcomes. However, a significant proportion of these infections are preventable. Evaluations of the performance of automated surveillance systems for adequate monitoring of central-line associated bloodstream infection (CLABSI) or catheter-related bloodstream infection (CRBSI) are limited.

OBJECTIVES

We evaluated the predictive performance of automated algorithms for CLABSI/CRBSI detection, and investigated which parameters included in automated algorithms provide the greatest accuracy for CLABSI/CRBSI detection.

METHODS

We performed a meta-analysis based on a systematic search of published studies in PubMed and EMBASE from 1 January 2000 to 31 December 2021. We included studies that evaluated predictive performance of automated surveillance algorithms for CLABSI/CRBSI detection and used manually collected surveillance data as reference. We estimated the pooled sensitivity and specificity of algorithms for accuracy and performed a univariable meta-regression of the different parameters used across algorithms.

RESULTS

The search identified five full text studies and 32 different algorithms or study populations were included in the meta-analysis. All studies analysed central venous catheters and identified CLABSI or CRBSI as an outcome. Pooled sensitivity and specificity of automated surveillance algorithm were 0.88 [95%CI 0.84-0.91] and 0.86 [95%CI 0.79-0.92] with significant heterogeneity (I2 = 91.9, p < 0.001 and I2 = 99.2, p < 0.001, respectively). In meta-regression, algorithms that include results of microbiological cultures from specific specimens (respiratory, urine and wound) to exclude non-CRBSI had higher specificity estimates (0.92, 95%CI 0.88-0.96) than algorithms that include results of microbiological cultures from any other body sites (0.88, 95% CI 0.81-0.95). The addition of clinical signs as a predictor did not improve performance of these algorithms with similar specificity estimates (0.92, 95%CI 0.88-0.96).

CONCLUSIONS

Performance of automated algorithms for detection of intravascular catheter infections in comparison to manual surveillance seems encouraging. The development of automated algorithms should consider the inclusion of results of microbiological cultures from specific specimens to exclude non-CRBSI, while the inclusion of clinical data may not have an added-value. Trail Registration Prospectively registered with International prospective register of systematic reviews (PROSPERO ID CRD42022299641; January 21, 2022). https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022299641.

Detection of infections by computerized capture of peaks in longitudinally measured C-reactive protein levels

We developed four algorithms for the automatic capture of C-reactive protein (CRP) peaks in 296 adult patients with acute myeloid leukemia who had bloodstream infection (BSI) episodes, negative blood cultures (BCs) or possible infections where no BCs were performed. The algorithms detected CRP peaks for 418-446 of the 586 documented BSI episodes (71.3-76.1%) and 2714-3118 of the 4382 negative BCs (61.9-71.2%). The four algorithms captured 382-789 CRP peaks in which there were neither BSI episodes nor negative BCs. We conclude that automatic capture of CRP peaks is a tool for the monitoring of BSI episodes and possibly other infections in patients with acute myeloid leukemia.

Effectiveness of enhanced check during acute phase to reduce central venous catheters-associated bloodstream infections: a before-after, real-world study

BACKGROUND

To evaluate the effectiveness of enhanced check to the duration of the central venous catheters associated bloodstream infections (CABSIs), and the impact on infection rates.

METHODS

A before-after, real-world study in six adult intensive care units was conducted. All adult patients who had only one central venous catheter were included during two consecutive periods. The intervention period, added cross-check that all patients with central venous catheter (CVC) need to be performed, and included nurses' checks for insertion practices and doctors' checks for maintenance practices. Propensity scores matching were used to account for potential confounding, and restricted cubic spline was served as visualizing the CABSI risk.

RESULTS

A total of 2906 patients with 26,157 CVC-days were analyzed. After intervention, the density incidence of CABSI decreased from 10.24 to 6.33/1,000 CVC-days (P < 0.001), and the acute period of rapid increase in CABSI risk was shortened, 6.5 to 5 days for femoral-vein catheterization and 7 to 5.5 days for subclavian-vein catheterization. For jugular-vein catheterization, the acute onset period disappeared.

CONCLUSION

Enhanced check during the first 7 calendar days after CVC insertion shortens the duration of the CABSI acute phase and tends to decrease CABSI rate.

Predictive performance of comorbidity for 30-day and 1-year mortality in patients with bloodstream infection visiting the emergency department: a retrospective cohort study

OBJECTIVES

To investigate whether the Charlson Comorbidity Index (CCI) predicted short-term and long-term mortality in patients with a bloodstream infection visiting the emergency department (ED) and compare it to the often-validated National Early Warning Score (NEWS).

DESIGN

A retrospective cohort study.

SETTING

A tertiary hospital in the Netherlands.

PARTICIPANTS

Adult patients attending the ED with a blood culture-proven infection between 2012 and 2017 were included. We collected the comorbidities from the CCI and the vital signs from the NEWS.

MAIN OUTCOMES

Short-term mortality (30-day) and long-term mortality (1 year). We assessed the predictive performance by discrimination, expressed as the area under the curve (AUC).

RESULTS

We included 1039 patients with a blood culture-proven infection. Mortality was 10.4% within 30 days and 27.8% within 1 year. On average patients had two comorbidities (ranging from 0 to 6). Highly prevalent comorbidities were malignancy (30.2%) and diabetes mellitus (20.5%). The predictive performance of the CCI was highest for 1-year mortality (AUC 0.696 (95%CI) (0.660 to 0.732)) and better compared with the NEWS (AUC (95% CI) 0.594 (0.555 to 0.632)). For prediction of 30-day mortality, the NEWS was superior (AUC (95% CI) 0.706 (0.656 to 0.756)) to the comorbidities of the CCI (AUC (95% CI) 0.568 (0.507 to 0.628)).

CONCLUSIONS

We found that presenting comorbidity (ie, the CCI) is most useful to prognosticate long-term outcome in patients with bloodstream infection in the ED. Short-term mortality is more accurately predicted by deviating vital signs (ie, the NEWS).

Bacteremia with Anaerobic Bacteria and Association with Colorectal Cancer: A Population-based Cohort Study

BACKGROUND

There is a well-described association between bacteremia with bovis group streptococci or Clostridium septicum and an increased risk of a colorectal cancer (CRC) diagnosis. We wanted to investigate the possible existence of a similar association between CRC and bacteremia with other bacteria belonging to the gut microbiota.

METHODS

A population based cohort study in a population about 2 million people including 45,774 bacteremia episodes and 231,387 blood culture negative cases was performed in the Region of Southern Denmark and Region Zealand (Denmark) from 2007-2016. Episodes of bacteremia were combined with the Danish central register for CRC. We performed Cox's regression analysis with hazard ratios (HRs) and 95% confidence intervals (CIs).

RESULTS

The study results confirmed previous findings of an increased risk of a CRC diagnosis after bacteremia with the bovis group streptococci (risk within a year: 4.3%; HR [95% CI]: 8.46 [3.51-20.4]) or C. septicum (20.8%; 76.2 [42.0-138]). Furthermore, Bacteroides ovatus (6.7%; 20.3 [5.04-81.8]), Bacteroides uniformis (5.4%; 16.2 [4.02-65.7]), Clostridium tertium (3.6 %; 13.9 [1.96-99.4]), Fusobacterium spp. (excluding F. necrophorum) (3.0 %; 8.51 [2.73-26.5]), and Gram-positive anaerobic cocci (3.6 %; 10.9 [4.50-26.3]) were also associated with an increased risk of a CRC diagnosis compared to patients with negative blood cultures (0.4%).

CONCLUSIONS

Bacteremia with several specific gut microbiota anaerobic bacteria is associated with a high risk of a diagnosis of CRC, indicating the need for colorectal workup in such cases. Importantly, this strategy also holds the possible additional benefit of detecting adenomas or other premalignant conditions, which were not included in the present study.

The association of body temperature with antibiotic therapy and mortality in patients attending the emergency department with suspected infection

BACKGROUND AND IMPORTANCE

Previous studies found that septic patients with normothermia have higher mortality than patients with fever. We hypothesize that antibiotic therapy is less frequently initiated if infectious patients present with normothermia to the emergency department (ED).

OBJECTIVES

To examine the association of body temperature with the initiation of antibiotic therapy in patients attending the ED with suspected and proven infection. Additionally, the association of temperature with 30-day mortality was assessed.

DESIGN, SETTINGS AND PARTICIPANTS

We conducted a retrospective cohort study between 2012 and 2016 at a tertiary university hospital. Adult patients attending the ED with a blood culture taken (i.e. suspected infection) and a positive blood culture (i.e. proven bacteremia) were included.

EXPOSURE

Tympanic temperature at arrival was categorized as hypothermia (<36.1°C), normothermia (36.1-38.0°C) or hyperthermia (>38.0°C).

OUTCOME MEASURES AND ANALYSIS

Primary outcome was the initiation of antibiotic therapy. A secondary outcome was 30-day mortality. Multivariable logistic regression was used to control for covariates.

MAIN RESULTS

Of 5997 patients with a suspected infection, 45.8% had normothermia, 44.6% hyperthermia and 5.6% hypothermia. Patients with hyperthermia received more often antibiotic therapy (53.5%) compared to normothermic patients (27.6%, adjusted odds ratio [95% confidence interval], 2.59 [2.27-2.95]). Patients with hyperthermia had lower mortality (4.7%) than those with normothermia (7.4%, adjusted odds ratio [95% confidence interval], 0.50 [0.39-0.64]). Sensitivity analyses in patients with proven bacteremia (n = 934) showed similar results.

CONCLUSION

Normothermia in patients presenting with infection was associated with receiving less antibiotic therapy in the ED compared to presentations with hyperthermia. Moreover, normothermia was associated with a higher mortality risk than hyperthermia.

Computer Informatics for Infection Control

Computer informatics have the potential to improve infection control outcomes in surveillance, prevention, and public health. Surveillance activities include surveillance of infections, device use, and facility/ward outbreak detection and investigation. Prevention activities include awareness of multidrug-resistant organism carriage on admission, identification of high-risk individuals or populations, reducing device use, and antimicrobial stewardship. Enhanced communication with public health and other health care facilities across networks includes automated electronic communicable disease reporting, syndromic surveillance, and regional outbreak detection. Computerized public health networks may represent the next major evolution in infection control. This article reviews the use of informatics for infection control.

Do we need to change catheter-related bloodstream infection surveillance in the Netherlands? A qualitative study among infection prevention professionals

OBJECTIVES

Catheter-related bloodstream infections (CRBSI) are a common healthcare-associated infection and therefore targeted by surveillance programmes in many countries. Concerns, however, have been voiced regarding the reliability and construct validity of CRBSI surveillance and the connection with the current diagnostic procedures. The aim of this study was to explore the experiences of infection control practitioners (ICPs) and medical professionals with the current CRBSI surveillance in the Netherlands and their suggestions for improvement.

DESIGN

Qualitative study using focus group discussions (FGDs) with ICPs and medical professionals separately, followed by semistructured interviews to investigate whether the points raised in the FGDs were recognised and confirmed by the interviewees. Analyses were performed using thematic analyses.

SETTING

Basic, teaching and academic hospitals in the Netherlands.

PARTICIPANTS

24 ICPs and 9 medical professionals.

RESULTS

Main themes derived from experiences with current surveillance were (1) ICPs' doubt regarding the yield of surveillance given the low incidence of CRBSI, the high workload and IT problems; (2) the experienced lack of leadership and responsibility for recording information needed for surveillance and (3) difficulties with applying and interpreting the CRBSI definition. Suggestions were made to simplify the surveillance protocol, expand the follow-up and surveillance to homecare settings, simplify the definition and customise it for specific patient groups. Participants reported hoping for and counting on automatisation solutions to support future surveillance.

CONCLUSIONS

This study reveals several problems with the feasibility and acceptance of the current CRBSI surveillance and proposes several suggestions for improvement. This provides valuable input for future surveillance activities, thereby taking into account automation possibilities.

PRAISE: providing a roadmap for automated infection surveillance in Europe

INTRODUCTION

Healthcare-associated infections (HAI) are among the most common adverse events of medical care. Surveillance of HAI is a key component of successful infection prevention programmes. Conventional surveillance - manual chart review - is resource intensive and limited by concerns regarding interrater reliability. This has led to the development and use of automated surveillance (AS). Many AS systems are the product of in-house development efforts and heterogeneous in their design and methods. With this roadmap, the PRAISE network aims to provide guidance on how to move AS from the research setting to large-scale implementation, and how to ensure the delivery of surveillance data that are uniform and useful for improvement of quality of care.

METHODS

The PRAISE network brings together 30 experts from ten European countries. This roadmap is based on the outcome of two workshops, teleconference meetings and review by an independent panel of international experts.

RESULTS

This roadmap focuses on the surveillance of HAI within networks of healthcare facilities for the purpose of comparison, prevention and quality improvement initiatives. The roadmap does the following: discusses the selection of surveillance targets, different organizational and methodologic approaches and their advantages, disadvantages and risks; defines key performance requirements of AS systems and suggestions for their design; provides guidance on successful implementation and maintenance; and discusses areas of future research and training requirements for the infection prevention and related disciplines. The roadmap is supported by accompanying documents regarding the governance and information technology aspects of implementing AS.

CONCLUSIONS

Large-scale implementation of AS requires guidance and coordination within and across surveillance networks. Transitions to large-scale AS entail redevelopment of surveillance methods and their interpretation, intensive dialogue with stakeholders and the investment of considerable resources. This roadmap can be used to guide future steps towards implementation, including designing solutions for AS and practical guidance checklists.

Decrease in All-Cause 30-Day Mortality after Bacteraemia over a 15-Year Period: A Population-Based Cohort Study in Denmark in 2000-2014

INTRODUCTION

Bacteraemia is a frequent infectious condition that strongly affects morbidity and mortality. The incidence is increasing worldwide. This study explores all-cause 30-day mortality after bacteraemia in two out of Denmark's five healthcare regions with approximately 2.4 million inhabitants.

METHODS

Clinically significant bacteraemia episodes (n = 55,257) were identified from a geographically well-defined background population between 2000 and 2014, drawing on population-based data regarding bacterial species and vital status. All-cause 30-day mortality was assessed in relation to bacteraemia episodes, number of patients with analysed blood cultures and the background population.

RESULTS

We observed a decreasing trend of all-cause 30-day mortality between 2000 and 2014, both in relation to the number of bacteraemia episodes and the background population. Mortality decreased from 22.7% of the bacteraemia episodes in 2000 to 17.4% in 2014 (annual IRR [95% CI]: 0.983 [0.979-0.987]). In relation to the background population, there were 41 deaths per 100,000 inhabitants in 2000, decreasing to 39 in 2014 (annual IRR [95% CI]: 0.988 [0.982-0.993]). Numbers of inhabitants, bacteraemia episodes, and analysed persons having BCs increased during the period.

CONCLUSIONS

All-cause 30-day mortality in patients with bacteraemia decreased significantly over a 15-year period.

Hybrid bag of approaches to characterize selection criteria for cohort identification

OBJECTIVE

The 2018 National NLP Clinical Challenge (2018 n2c2) focused on the task of cohort selection for clinical trials, where participating systems were tasked with analyzing longitudinal patient records to determine if the patients met or did not meet any of the 13 selection criteria. This article describes our participation in this shared task.

MATERIALS AND METHODS

We followed a hybrid approach combining pattern-based, knowledge-intensive, and feature weighting techniques. After preprocessing the notes using publicly available natural language processing tools, we developed individual criterion-specific components that relied on collecting knowledge resources relevant for these criteria and pattern-based and weighting approaches to identify "met" and "not met" cases.

RESULTS

As part of the 2018 n2c2 challenge, 3 runs were submitted. The overall micro-averaged F1 on the training set was 0.9444. On the test set, the micro-averaged F1 for the 3 submitted runs were 0.9075, 0.9065, and 0.9056. The best run was placed second in the overall challenge and all 3 runs were statistically similar to the top-ranked system. A reimplemented system achieved the best overall F1 of 0.9111 on the test set.

DISCUSSION

We highlight the need for a focused resource-intensive effort to address the class imbalance in the cohort selection identification task.

CONCLUSION

Our hybrid approach was able to identify all selection criteria with high F1 performance on both training and test sets. Based on our participation in the 2018 n2c2 task, we conclude that there is merit in continuing a focused criterion-specific analysis and developing appropriate knowledge resources to build a quality cohort selection system.

Electronically assisted surveillance systems of healthcare-associated infections: a systematic review

Background

Surveillance of healthcare-associated infections (HAI) is the basis of each infection control programme and, in case of acute care hospitals, should ideally include all hospital wards, medical specialties as well as all types of HAI. Traditional surveillance is labour intensive and electronically assisted surveillance systems (EASS) hold the promise to increase efficiency.

Objectives

To give insight in the performance characteristics of different approaches to EASS and the quality of the studies designed to evaluate them.

Methods

In this systematic review, online databases were searched and studies that compared an EASS with a traditional surveillance method were included. Two different indicators were extracted from each study, one regarding the quality of design (including reporting efficiency) and one based on the performance (e.g. specificity and sensitivity) of the EASS presented.

Results

A total of 78 studies were included. The majority of EASS (n = 72) consisted of an algorithm-based selection step followed by confirmatory assessment. The algorithms used different sets of variables. Only a minority (n = 7) of EASS were hospital-wide and designed to detect all types of HAI. Sensitivity of EASS was generally high (> 0.8), but specificity varied (0.37–1). Less than 20% (n = 14) of the studies presented data on the efficiency gains achieved.

Conclusions

Electronically assisted surveillance of HAI has yet to reach a mature stage and to be used routinely in healthcare settings. We recommend that future studies on the development and implementation of EASS of HAI focus on thorough validation, reproducibility, standardised datasets and detailed information on efficiency.

Validation of a semi-automated surveillance system for surgical site infections: Improving exhaustiveness, representativeness, and efficiency

OBJECTIVES

To assess whether electronic records data could improve the efficiency, exhaustiveness, and representativeness of SSI surveillance by selecting a group of high-risk patients for manual review.

METHODS

Colorectal surgeries (2016-2018) and cholecystectomies (2017-2018) were selected. Post-surgical antibiotic use, positive culture, C-reactive protein (CRP) values, body temperature, leukocyte count, surgical re-intervention, admission to the emergency room, and hospital readmission were retrieved. For representativeness, procedures registered in HAI-Net were compared with non-included procedures, and the validity of each variable (or combination) was tested considering the presence of SSI as the gold standard. The proportion of procedures flagged for manual review by each criterion was estimated.

RESULTS

Little more than 50% of procedures were included in HAI-Net (SSI risk: 10.6% for colorectal and 2.9% for cholecystectomies). Non-included procedures showed higher proportions of infection markers. Antibiotic use and CRP >100 mg/dl presented the highest sensitivity for both surgical groups, while antibiotic use achieved the highest positive predictive value in both groups (22% and 21%, respectively) and flagged fewer colorectal procedures (47.7%).

CONCLUSIONS

Current SSI surveillance has major limitations. Thus, the reported incidence seems unreliable and underestimated. Antibiotic use appears to be the best criterion to select a sub-sample of procedures for manual review, improving the exhaustiveness and efficiency of the system.

Non-adherence to antimicrobial guidelines in patients with bloodstream infection visiting the emergency department

OBJECTIVE

Non-adherence to antimicrobial guidelines in patients with bloodstream infection can result in undertreatment, overtreatment, or equivalent treatment, and could lead to suboptimal care. Our aim was to examine the association between non-adherence and appropriate coverage as well as to assess the impact of non-adherence on 30-day mortality.

METHODS

We conducted a retrospective cohort study between 2012 and 2017 at a tertiary university hospital. Adult patients attending the emergency department with a bloodstream infection were included. Adherence was defined as guideline-recommended antibiotic therapy. Non-adherence was either undertreatment (too narrow-spectrum), overtreatment (too broad-spectrum), or equivalent treatment. Outcomes were appropriate coverage (i.e. antibiotic therapy that matches in vitro susceptibility of the isolated bacteria) and 30-day mortality.

RESULTS

We included 909 patients of whom 395 (43.5%) were treated adherently, 355 (39.1%) were undertreated, 87 (9.6%) were overtreated, and 72 (7.9%) received an equivalent treatment. Overtreated patients were more severely ill, whilst undertreated patients had more favorable patient characteristics. Overtreatment did not result in higher appropriate coverage, whereas undertreatment was associated with lower coverage (OR[95%CI]: 0.18 [0.12; 0.26]). Overtreatment and undertreatment were not associated with 30-day mortality.

CONCLUSIONS

Guideline adherence likely depends on disease severity, because overtreatment was more often observed in patients with high disease severity and undertreatment in less severely ill patients. Undertreatment was associated lower appropriate coverage but not with higher mortality. However, this can be the result of residual confounding . Overtreatment did not result in higher appropriate antibiotic coverage nor a survival benefit . Therefore, overtreatment seems not justifiable.

Longitudinal trajectory patterns of plasma albumin and C-reactive protein levels around diagnosis, relapse, bacteraemia, and death of acute myeloid leukaemia patients

Background

No study has evaluated C-reactive protein (CRP) and plasma albumin (PA) levels longitudinally in patients with acute myeloid leukaemia (AML).

Methods

We studied defined events in 818 adult patients with AML in relation to 60,209 CRP and PA measures. We investigated correlations between CRP and PA levels and daily CRP and PA levels in relation to AML diagnosis, AML relapse, or bacteraemia (all ±30 days), and death (─30–0 days).

Results

On the AML diagnosis date (D0), CRP levels increased with higher WHO performance score (PS), e.g. patients with PS 3/4 had 68.1 mg/L higher CRP compared to patients with PS 0, adjusted for relevant covariates. On D0, the PA level declined with increasing PS, e.g. PS 3/4 had 7.54 g/L lower adjusted PA compared to PS 0. CRP and PA levels were inversely correlated for the PA interval 25–55 g/L (R = − 0.51, p < 10–5), but not for ≤24 g/L (R = 0.01, p = 0.57). CRP increases and PA decreases were seen prior to bacteraemia and death, whereas no changes occurred up to AML diagnosis or relapse. CRP increases and PA decreases were also found frequently in individuals, unrelated to a pre-specified event.

Conclusions

PA decrease is an important biomarker for imminent bacteraemia in adult patients with AML.

C-reactive protein and albumin kinetics before community-acquired bloodstream infections - a Danish population-based cohort study

Early changes in biomarker levels probably occur before bloodstream infection (BSI) is diagnosed. However, this issue has not been fully addressed. We aimed at evaluating the kinetics of C-reactive protein (CRP) and plasma albumin (PA) in the 30 days before community-acquired (CA) BSI diagnosis. From a population-based BSI database we identified 658 patients with at least one measurement of CRP or PA from day −30 (D–30) through day −1 (D–1) before the day of CA-BSI (D0) and a measurement of the same biomarker at D0 or D1. Amongst these, 502 had both CRP and PA measurements which fitted these criteria. CRP and PA concentrations began to change inversely some days before CA-BSI diagnosis, CRP increasing by day −3.1 and PA decreasing by day −1.3. From D–30 to D–4, CRP kinetics (expressed as slopes – rate of concentration change per day) was −1.5 mg/l/day. From D–3 to D1, the CRP slope increased to 36.3 mg/l/day. For albumin, the slope between D–30 to D–2 was 0.1 g/l/day and changed to −1.8 g/l/day between D–1 and D1. We showed that biomarker levels begin to change some days before the CA-BSI diagnosis, CRP 3.1 days and PA 1.3 days before.

Appropriate empirical antibiotic therapy and mortality: Conflicting data explained by residual confounding

Objective

Clinical practice universally assumes that appropriate empirical antibiotic therapy improves survival in patients with bloodstream infection. However, this is not generally supported by previous studies. We examined the association between appropriate therapy and 30-day mortality, while minimizing bias due to confounding by indication.

Methods

We conducted a retrospective cohort study between 2012 and 2017 at a tertiary university hospital in the Netherlands. Adult patients with bloodstream infection attending the emergency department were included. Based on in vitro susceptibility, antibiotic therapy was scored as appropriate or inappropriate. Primary outcome was 30-day mortality. To control for confounding, we performed conventional multivariable logistic regression and propensity score methods. Additionally, we performed an analysis in a more homogeneous subgroup (i.e. antibiotic monotherapy).

Results

We included 1.039 patients, 729 (70.2%) received appropriate therapy. Overall 30-day mortality was 10.4%. Appropriately treated patients had more unfavorable characteristics, indicating more severe illness. Despite adjustments, we found no association between appropriate therapy and mortality. For the antibiotic monotherapy subgroup (n = 449), patient characteristics were more homogeneous. Within this subgroup, appropriate therapy was associated with lower mortality (Odds Ratios [95% Confidence Intervals] ranging from: 0.31 [0.14; 0.67] to 0.40 [0.19; 0.85]).

Conclusions

Comparing heterogeneous treatment groups distorts associations despite use of common methods to prevent bias. Consequently, conclusions of such observational studies should be interpreted with care. If possible, future investigators should use our method of attempting to identify and analyze the most homogeneous treatment groups nested within their study objective, because this minimizes residual confounding.

Evaluation of a web-based tool for labelling potential hospital outbreaks: a mixed methods study

BACKGROUND

Labelling outbreaks in surveillance data is necessary to train advanced analytical methods for outbreak detection, but there is a lack of software tools dedicated to this task.

AIM

To evaluate the usability of a web-based tool by infection control practitioners for labelling potential outbreaks.

METHODS

A mixed methods design was used to evaluate how 25 experts from France and Canada interacted with a web-based application to identify potential outbreaks. Each expert used the application to retrospectively review 11-12 1-year incidence time series from 23 different types of micro-organism. The interactions between the users and the application were recorded and analysed using mixed effect models. The users' comments were analysed via qualitative methods.

FINDINGS

From the 240 reviews completed, 439 potential outbreaks were labelled, approximately half with a high probability. Significant heterogeneity was observed between users regarding their answers and behaviours (evaluation time, usage of the different options). A significant learning effect was also observed for the experts' interactions with the tool, but this did not seem to impact their answers. The content analysis of the comments highlighted the difficulty of early outbreak identification for practitioners, but also the potential utility of web applications such as that evaluated for routine surveillance.

CONCLUSION

The interactive web application was both usable and useful for infection control practitioners. Its implementation in routine practice could help professionals to identify potential outbreaks while creating data to train automated detection algorithms.

WMSS: A Web-Based Multitiered Surveillance System for Predicting CLABSI

Central-line-associated bloodstream infection (CLABSI) rates are a key quality metric for comparing hospital quality and safety. Manual surveillance systems for CLABSIs are time-consuming and often limited to intensive care units (ICUs). A computer-automated method of CLABSI detection can improve the validity of surveillance. A new web-based, multitiered surveillance system for predicting and reducing CLABSI is proposed. The system has the capability to collect patient-related data from hospital databases and hence predict the patient infection automatically based on knowledge discovery rules and CLABSI decision standard algorithms. In addition, the system has a built-in simulator for generating patients' data records, when needed, offering the capability to train nurses and medical staff for enhancing their qualifications. Applying the proposed system, both CLABSI rates and patient treatment costs can be reduced significantly. The system has many benefits, among which there is the following: it is a web-based system that can collect real patients' data from many IT resources using iPhone, iPad, laptops, Internet, scanners, and hospital databases. These facilities help to collect patients' actual data quickly and safely in electronic format and hence predict CLABSI efficiently. Automation of the patients' data diagnosis process helps in reducing CLABSI detection times. The system is multimedia-based; it uses text, colors, and graphics to enhance patient healthcare report generation and charts. It helps healthcare decision makers to review and approve policies and surveillance plans to reduce and prevent CLABSI.

Impact of appropriate empirical antibiotic treatment on recurrence and mortality in patients with bacteraemia: a population-based cohort study

Background

Data on the impact of empirical antibiotic treatment (EAT) on patient outcome in a population-based setting are sparse. We assessed the association between EAT and the risk of recurrence within one year, short-term- (2–30 days) and long-term (31–365 days) mortality in a Danish cohort of bacteraemia patients.

Methods

A cohort study including all patients hospitalized with incident bacteraemia during 2007–2008 in the Copenhagen City and County areas and the North Denmark Region. EAT was defined as the antibiotic treatment given at the 1^st notification of a positive blood culture. The definition of recurrence took account of pathogen species, site of infection, and time frame and was not restricted to homologous pathogens. The vital status was determined through the civil registration system. Association estimates between EAT and the outcomes were estimated by Cox and logistic regression models.

Results

In 6483 eligible patients, 712 (11%) had a recurrent episode. A total of 3778 (58%) patients received appropriate EAT, 1290 (20%) received inappropriate EAT, while EAT status was unrecorded for 1415 (22%) patients. The 2–30 day mortality was 15.1%, 17.4% and 19.2% in patients receiving appropriate EAT, inappropriate EAT, and unknown EAT, respectively. Among patients alive on day 30, the 31–365 day mortality was 22.3% in patients given appropriate EAT compared to 30.7% in those given inappropriate EAT. Inappropriate EAT was independently associated with recurrence (HR 1.25; 95% CI = 1.03–1.52) and long-term mortality (OR 1.35; 95% CI = 1.10–1.60), but not with short-term mortality (OR 0.85; 95% CI = 0.70–1.02) after bacteraemia.

Conclusions

Our data indicate that appropriate EAT is associated with reduced incidence of recurrence and lower long-term mortality following bacteraemia.

Electronic supplementary material

The online version of this article (doi:10.1186/s12879-017-2233-z) contains supplementary material, which is available to authorized users.

Guidance on Public Reporting of Healthcare-Associated Infections: Recommendations of the Healthcare Infection Control Practices Advisory Committee

Consumer demand for healthcare information, including data about the performance of healthcare providers, has increased steadily during the past decade. Many state and national initiatives are under way to mandate or induce healthcare organizations to publicly disclose information regarding institutional and physician performance. Mandatory public reporting of healthcare performance is intended to enable stakeholders, including consumers, to make more informed choices on healthcare issues.

Public reporting of healthcare performance information has taken several forms. Healthcare performance reports (report cards and honor rolls) typically describe the outcomes of medical care in terms of mortality, selected complications, or medical errors and, to a lesser extent, economic outcomes. Increasingly, process measures (ie, measurement of adherence to recommended healthcare practices, such as hand hygiene) are being used as an indicator of how well an organization adheres to established standards of practice with the implicit assumption that good processes lead to good healthcare outcomes. National healthcare quality improvement initiatives, notably those of the Joint Commission on the Accreditation of Healthcare Organizations (JCAHO), the Centers for Medicare & Medicaid Services (CMS), and the Hospital Quality Alliance, use process measures in their public reporting initiatives.

Sampling for Collection of Central Line–Day Denominators in Surveillance of Healthcare-Associated Bloodstream Infections

Objective.

To determine the feasibility of estimating the number of central line-days at a hospital from a sample of months or individual days in a year, for surveillance of healthcare-associated bloodstream infections.

Design.

We used data reported to the National Nosocomial Infections Surveillance system in the adult and pediatric intensive care unit component for 1995-2003 and data from a sample of hospitals' daily counts of device use for 12 consecutive months. We calculated the percentile error as the central line-associated bloodstream infection percentile based on rates per line-days minus the percentile based on rates per estimated line-days.

Setting and Participants.

A total of 247 hospitals were used for sampling whole months and 12 hospitals were used for sampling individual days.

Results.

For a 1-month sample of central line–days data, the median percentile error was 3.3 (75th percentile, 7.9; 90th percentile, 15.4). The percentile error decreased with an increase in the number of months sampled. For a 3-month sample, the median percentile error was 1.4 (75th percentile, 4.3; 95th percentile, 8.3). Sampling individual days throughout the year yielded lower percentile errors than sampling an equivalent fraction of whole months. With 1 weekday sampled per week, the median percentile error ranged from 0.65 to 1.40, and the 90th percentile ranged from 2.8 to 5.0. Thus, for 90% of units, collecting data on line-days once a week provides an estimate within ± 5 percentile points of the true line-day rate.

Conclusion.

Sample-based estimates of central line-days can yield results that are acceptable for surveillance of healthcare-associated bloodstream infections.

The Validation of a Novel Surveillance System for Monitoring Bloodstream Infections in the Calgary Zone

Background. Electronic surveillance systems (ESSs) that utilize existing information in databases are more efficient than conventional infection surveillance methods. The objective was to assess an ESS for bloodstream infections (BSIs) in the Calgary Zone for its agreement with traditional medical record review. Methods. The ESS was developed by linking related data from regional laboratory and hospital administrative databases and using set definitions for excluding contaminants and duplicate isolates. Infections were classified as hospital-acquired (HA), healthcare-associated community-onset (HCA), or community-acquired (CA). A random sample of patients from the ESS was then compared with independent medical record review. Results. Among the 308 patients selected for comparative review, the ESS identified 318 episodes of BSI of which 130 (40.9%) were CA, 98 (30.8%) were HCA, and 90 (28.3%) were HA. Medical record review identified 313 episodes of which 136 (43.4%) were CA, 97 (30.9%) were HCA, and 80 (25.6%) were HA. Episodes of BSI were concordant in 304 (97%) cases. Overall, there was 85.5% agreement between ESS and medical record review for the classification of where BSIs were acquired (kappa = 0.78, 95% Confidence Interval: 0.75-0.80). Conclusion. This novel ESS identified and classified BSIs with a high degree of accuracy. This system requires additional linkages with other related databases.

Seasonal Variation of Escherichia coli, Staphylococcus aureus, and Streptococcus pneumoniae Bacteremia According to Acquisition and Patient Characteristics: A Population-Based Study

OBJECTIVE Seasonal variation is a characteristic of many infectious diseases, but relatively little is known about determinants thereof. We studied the impact of place of acquisition and patient characteristics on seasonal variation of bacteremia caused by the 3 most common pathogens. DESIGN Seasonal variation analysis. METHODS In 3 Danish health regions (2.3 million total inhabitants), patients with bacteremia were identified from 2000 through 2011 using information from laboratory information systems. Analyses were confined to Escherichia coli, Staphylococcus aureus, and Streptococcus pneumoniae. Additional data were obtained from the Danish National Hospital Registry for the construction of admission histories and calculation of the Charlson comorbidity index (CCI). Bacteremias were categorized as community acquired, healthcare associated (HCA), and hospital acquired. We defined multiple subgroups by combining the following characteristics: species, acquisition, age group, gender, CCI level, and location of infection. Assuming a sinusoidal model, seasonal variation was assessed by the peak-to-trough (PTT) ratio with a 95% confidence interval (CI). RESULTS In total, we included 16,006 E. coli, 6,924 S. aureus, and 4,884 S. pneumoniae bacteremia cases. For E. coli, the seasonal variation was highest for community-acquired cases (PTT ratio, 1.24; 95% CI, 1.17-1.32), was diminished for HCA (PTT ratio, 1.14; 95% CI, 1.04-1.25), and was missing for hospital-acquired cases. No seasonal variation was observed for S. aureus. S. pneumoniae showed high seasonal variation, which did not differ according to acquisition (overall PTT ratio, 3.42; 95% CI, 3.10-3.83). CONCLUSIONS Seasonal variation was mainly related to the species although the place of acquisition was important for E. coli. Infect Control Hosp Epidemiol 2016;37:946-953.

Performance characteristics and associated outcomes for an automated surveillance tool for bloodstream infection

BACKGROUND

The objective of this study was to evaluate performance metrics and associated patient outcomes of an automated surveillance system, the blood Nosocomial Infection Marker (NIM).

METHODS

We reviewed records of 237 patients with and 36,927 patients without blood NIM using the National Healthcare Safety Network (NHSN) definition for laboratory-confirmed bloodstream infection (BSI) as the gold standard. We matched cases with noncases by propensity score and estimated attributable mortality and cost of NHSN-reportable central line-associated bloodstream infections (CLABSIs) and non-NHSN-reportable BSIs.

RESULTS

For patients with central lines (CL), the blood NIM had 73.2% positive predictive value (PPV), 99.9% negative predictive value (NPV), 89.2% sensitivity, and 99.7% specificity. For all patients regardless of CL status, the blood NIM had 53.6% PPV, 99.9% NPV, 84.0% sensitivity, and 99.9% specificity. For CLABSI cases compared with noncases, mortality was 17.5% versus 9.4% (P = .098), and median charge was $143,935 (interquartile range [IQR], $89,794-$257,447) versus $115,267 (IQR, $74,937-$173,053) (P < .01). For non-NHSN-reportable BSI cases compared with noncases, mortality was 23.6% versus 6.7% (P < .0001), and median charge was $86,927 (IQR, $54,728-$156,669) versus $62,929 (IQR, $36,743-$115,693) (P < .0001).

CONCLUSIONS

The NIM is an effective screening tool for BSI. Both NHSN-reportable and nonreportable BSI cases were associated with increased mortality and cost.

Automated surveillance system for hospital-acquired urinary tract infections in Denmark

BACKGROUND

The Danish Hospital-Acquired Infections Database (HAIBA) is an automated surveillance system using hospital administrative, microbiological, and antibiotic medication data.

AIM

To define and evaluate the case definition for hospital-acquired urinary tract infection (HA-UTI) and to describe surveillance data from 2010 to 2014.

METHODS

The HA-UTI algorithm defined a laboratory-diagnosed UTI as a urine culture positive for no more than two micro-organisms with at least one at ≥10(4)cfu/mL, and a probable UTI as a negative urine culture and a relevant diagnosis code or antibiotic treatment. UTI was considered hospital-acquired if a urine sample was collected ≥48h after admission and <48h post discharge. Incidence of HA-UTI was calculated per 10,000 risk-days. For validation, prevalence was calculated for each day and compared to point prevalence survey (PPS) data.

FINDINGS

HAIBA detected a national incidence rate of 42.2 laboratory-diagnosed HA-UTI per 10,000 risk-days with an increasing trend. Compared to PPS the laboratory-diagnosed HA-UTI algorithm had a sensitivity of 50.0% (26/52) and a specificity of 94.2% (1842/1955). There were several reasons for discrepancies between HAIBA and PPS, including laboratory results being unavailable at the time of the survey, the results considered clinically irrelevant by the surveyor due to an indwelling urinary catheter or lack of clinical signs of infection, and UTIs being considered HA-UTI in PPS even though the first sample was taken within 48h of admission.

CONCLUSION

The HAIBA algorithm was found to give valid and valuable information and has, among others, the advantages of covering the whole population and allowing continuous standardized monitoring of HA-UTI.

Text Mining for Precision Medicine: Bringing Structure to EHRs and Biomedical Literature to Understand Genes and Health

The key question of precision medicine is whether it is possible to find clinically actionable granularity in diagnosing disease and classifying patient risk. The advent of next-generation sequencing and the widespread adoption of electronic health records (EHRs) have provided clinicians and researchers a wealth of data and made possible the precise characterization of individual patient genotypes and phenotypes. Unstructured text-found in biomedical publications and clinical notes-is an important component of genotype and phenotype knowledge. Publications in the biomedical literature provide essential information for interpreting genetic data. Likewise, clinical notes contain the richest source of phenotype information in EHRs. Text mining can render these texts computationally accessible and support information extraction and hypothesis generation. This chapter reviews the mechanics of text mining in precision medicine and discusses several specific use cases, including database curation for personalized cancer medicine, patient outcome prediction from EHR-derived cohorts, and pharmacogenomic research. Taken as a whole, these use cases demonstrate how text mining enables effective utilization of existing knowledge sources and thus promotes increased value for patients and healthcare systems. Text mining is an indispensable tool for translating genotype-phenotype data into effective clinical care that will undoubtedly play an important role in the eventual realization of precision medicine.

Data requirements for electronic surveillance of healthcare-associated infections

Electronic surveillance for healthcare-associated infections (HAIs) is increasingly widespread. This is driven by multiple factors: a greater burden on hospitals to provide surveillance data to state and national agencies, financial pressures to be more efficient with HAI surveillance, the desire for more objective comparisons between healthcare facilities, and the increasing amount of patient data available electronically. Optimal implementation of electronic surveillance requires that specific information be available to the surveillance systems. This white paper reviews different approaches to electronic surveillance, discusses the specific data elements required for performing surveillance, and considers important issues of data validation.

A Web-Based, Hospital-Wide Health Care-Associated Bloodstream Infection Surveillance and Classification System: Development and Evaluation

BACKGROUND

Surveillance of health care-associated infections is an essential component of infection prevention programs, but conventional systems are labor intensive and performance dependent.

OBJECTIVE

To develop an automatic surveillance and classification system for health care-associated bloodstream infection (HABSI), and to evaluate its performance by comparing it with a conventional infection control personnel (ICP)-based surveillance system.

METHODS

We developed a Web-based system that was integrated into the medical information system of a 2200-bed teaching hospital in Taiwan. The system automatically detects and classifies HABSIs.

RESULTS

In this study, the number of computer-detected HABSIs correlated closely with the number of HABSIs detected by ICP by department (n=20; r=.999 P<.001) and by time (n=14; r=.941; P<.001). Compared with reference standards, this system performed excellently with regard to sensitivity (98.16%), specificity (99.96%), positive predictive value (95.81%), and negative predictive value (99.98%). The system enabled decreasing the delay in confirmation of HABSI cases, on average, by 29 days.

CONCLUSIONS

This system provides reliable and objective HABSI data for quality indicators, improving the delay caused by a conventional surveillance system.

Low Completeness of Bacteraemia Registration in the Danish National Patient Registry

Bacteraemia is associated with significant morbidity and mortality and timely access to relia-ble information is essential for health care administrators. Therefore, we investigated the complete-ness of bacteraemia registration in the Danish National Patient Registry (DNPR) containing hospital discharge diagnoses and surgical procedures for all non-psychiatric patients. As gold standard we identified bacteraemia patients in three defined areas of Denmark (~2.3 million inhabitants) from 2000 through 2011 by use of blood culture data retrieved from electronic microbiology databases. Diagnoses coded according to the International Classification of Diseases, version 10, and surgical procedure codes were retrieved from the DNPR. The codes were categorized into seven groups, ranked a priori according to the likelihood of bacteraemia. Completeness was analysed by contin-gency tables, for all patients and subgroups. We identified 58,139 bacteraemic episodes in 48,450 patients; 37,740 episodes (64.9%) were covered by one or more discharge diagnoses within the sev-en diagnosis/surgery groups and 18,786 episodes (32.3%) had a code within the highest priority group. Completeness varied substantially according to speciality (from 17.9% for surgical to 36.4% for medical), place of acquisition (from 26.0% for nosocomial to 36.2% for community), and mi-croorganism (from 19.5% for anaerobic Gram-negative bacteria to 36.8% for haemolytic strepto-cocci). The completeness increased from 25.1% in 2000 to 35.1% in 2011. In conclusion, one third of the bacteraemic episodes did not have a relevant diagnosis in the Danish administrative registry recording all non-psychiatric contacts. This source of information should be used cautiously to iden-tify patients with bacteraemia.

Population-based epidemiology and microbiology of community-onset bloodstream infections

Bloodstream infection (BSI) is a major cause of infectious disease morbidity and mortality worldwide. While a positive blood culture is mandatory for establishment of the presence of a BSI, there are a number of determinants that must be considered for establishment of this entity. Community-onset BSIs are those that occur in outpatients or are first identified <48 h after admission to hospital, and they may be subclassified further as health care associated, when they occur in patients with significant prior health care exposure, or community associated, in other cases. The most common causes of community-onset BSI include Escherichia coli, Staphylococcus aureus, and Streptococcus pneumoniae. Antimicrobial-resistant organisms, including methicillin-resistant Staphylococcus aureus and extended-spectrum β-lactamase/metallo-β-lactamase/carbapenemase-producing Enterobacteriaceae, have emerged as important etiologies of community-onset BSI.

An electronic surveillance tool for catheter-associated urinary tract infection in intensive care units

BACKGROUND

Traditional methods of surveillance of catheter-associated urinary tract infections (CAUTIs) are error-prone and resource-intensive. To resolve these issues, we developed a highly sensitive electronic surveillance tool.

OBJECTIVE

To develop an electronic surveillance tool for CAUTIs and assess its performance.

METHODS

The study was conducted at a 947-bed tertiary care center. Patients included adults aged ≥18 years admitted to an intensive care unit between January 10 and June 30, 2012, with an indwelling urinary catheter during their admission. We identified CAUTIs using 4 methods: traditional surveillance (TS) (ie, manual chart review by ICPs), an electronic surveillance (ES) tool, augmented electronic surveillance (AES) (ie, ES with chart review on a subset of cases), and reference standard (RS) (ie, a subset of CAUTIs originally ascertained by TS or ES, confirmed by review). We assessed performance characteristics to RS for reviewed cases.

RESULTS

We identified 417 candidate CAUTIs in 308 patients; 175 (42.0%) of these candidate CAUTIs were selected for review, yielding 32 confirmed CAUTIs in 22 patients (RS). Compared with RS, the sensitivities of TS, ES, and AES were 43.8% (95% confidence interval [CI], 26.4%-62.3%), 100.0% (95% CI, 89.1%-100.0%), and 100.0% (95% CI, 89.1%-100.0%). Specificities were 82.5% (95% CI, 75.3%-88.4%), 2.8% (95% CI, 0.8%-7.0%), and 100.0% (95% CI, 97.5%-100.0%).

CONCLUSIONS

Electronic CAUTI surveillance offers a streamlined approach to improve reliability and resource burden of surveillance.

Accuracy of using automated methods for detecting adverse events from electronic health record data: a research protocol

Background

Adverse events are associated with significant morbidity, mortality and cost in hospitalized patients. Measuring adverse events is necessary for quality improvement, but current detection methods are inaccurate, untimely and expensive. The advent of electronic health records and the development of automated methods for encoding and classifying electronic narrative data, such as natural language processing, offer an opportunity to identify potentially better methods. The objective of this study is to determine the accuracy of using automated methods for detecting three highly prevalent adverse events: a) hospital-acquired pneumonia, b) catheter-associated bloodstream infections, and c) in-hospital falls.

Methods/design

This validation study will be conducted at two large Canadian academic health centres: the McGill University Health Centre (MUHC) and The Ottawa Hospital (TOH). The study population consists of all medical, surgical and intensive care unit patients admitted to these centres between 2008 and 2014. An automated detection algorithm will be developed and validated for each of the three adverse events using electronic data extracted from multiple clinical databases. A random sample of MUHC patients will be used to develop the automated detection algorithms (cohort 1, development set). The accuracy of these algorithms will be assessed using chart review as the reference standard. Then, receiver operating characteristic curves will be used to identify optimal cut points for each of the data sources. Multivariate logistic regression and the areas under curve (AUC) will be used to identify the optimal combination of data sources that maximize the accuracy of adverse event detection. The most accurate algorithms will then be validated on a second random sample of MUHC patients (cohort 1, validation set), and accuracy will be measured using chart review as the reference standard. The most accurate algorithms validated at the MUHC will then be applied to TOH data (cohort 2), and their accuracy will be assessed using a reference standard assessment of the medical chart.

Discussion

There is a need for more accurate, timely and efficient measures of adverse events in acute care hospitals. This is a critical requirement for evaluating the effectiveness of preventive interventions and for tracking progress in patient safety through time.

Introduction: Improving Patient Safety Through Infection Control: A New Healthcare Imperative

Many healthcare organizations, professional associations, government and accrediting agencies, legislators, regulators, payers, and consumer advocacy groups have advanced the prevention of healthcare-associated infections as a national imperative, stimulating the creation of “A Compendium of Strategies to Prevent Healthcare-Associated Infections in Acute Care Hospitals” in this supplement. In this introduction, we provide background and context and discuss the major issues that shaped the recommendations included in the compendium.

Development of a Novel Electronic Surveillance System for Monitoring of Bloodstream Infections

Background.

Electronic surveillance systems (ESSs) that utilize existing information in databases are more efficient than conventional infection surveillance methods.

Objective.

To develop an ESS for monitoring bloodstream infections (BSIs) and assess whether data obtained from the ESS were in agreement with data obtained by traditional manual medical-record review.

Methods.

An ESS was developed by linking data from regional laboratory and hospital administrative databases. Definitions for excluding BSI episodes representing contamination and duplicate episodes were developed and applied. Infections were classified as nosocomial infections, healthcare-associated community-onset infections, or community-acquired infections. For a random sample of episodes, data in the ESS were compared with data obtained by independent medical chart review.

Results.

From the records of the 306 patients whose infections were selected for comparative review, the ESS identified 323 episodes of BSI, of which 107 (33%) were classified as healthcare-associated community-onset infections, 108 (33%) were classified as community-acquired infections, 107 (33%) were classified as nosocomial infections, and 1 (0.3%) could not be classified. In comparison, 310 episodes were identified by use of medical chart review, of which 116 (37%) were classified as healthcare-associated community-onset infections, 95 (31%) as community-acquired infections, and 99 (32%) as nosocomial infections. For 302 episodes of BSI, there was concordance between the findings of the ESS and those of traditional manual chart review. Of the additional 21 discordant episodes that were identified by use of the ESS, 17 (81%) were classified as representing isolation of skin contaminants, by use of chart review. Of the additional 8 discordant episodes further identified by use of chart review, most were classified as repeat or polymicrobial episodes of disease. There was an overall 85% agreement between the findings of the ESS and those of chart review (K = 0.78; standard error, K = 0.04) for classification according to location of acquisition.

Conclusion.

Our novel ESS allows episodes of BSI to be identified and classified with a high degree of accuracy. This system requires validation in other cohorts and settings.

Recommendations For Metrics For Multidrug-Resistant Organisms In Healthcare Settings: SHEA/HICPAC Position Paper

Monitoring multidrug-resistant organisms (MDROs) and the infections they cause in a healthcare setting is important to detect newly emerging antimicrobial resistance profiles, to identify vulnerable patient populations, and to assess the need for and effectiveness of interventions; however, it is unclear which metrics are the best, because most of the metrics are not standardized. This document describes useful and practical metrics and surveillance considerations for measuring MDROs and the infections they cause in the practice of infection prevention and control in healthcare settings. These metrics are designed to aid healthcare workers in documenting trends over time within their facility and should not be used for interfacility comparison.

Electronic Algorithmic Prediction of Central Vascular Catheter Use

Objective.

To develop prediction algorithms for the presence of a central vascular catheter in hospitalized patients with use of data present in an electronic health record. Such algorithms could be used for measurement of device utilization rates and for clinical decision support rules.

Design.

Criterion standard.

Setting.

John H. Stroger, Jr, Hospital of Cook County, a 464-bed public hospital in Chicago, Illinois.

Participants.

Patients admitted to the medical intensive care unit from May 31, 2005 through June 26, 2006 (derivation data set, May 31, 2005-September 28, 2005; validation data set, September 29, 2005-June 28, 2006).

Methods.

Covariates were collected from the electronic medical record for each patient; the outcome variable was presence of a central vascular device. Multivariate models were developed using the derivation set and the generalized estimating equation. Three models, each with increasing database requirements, were validated using the validation set. Device utilization ratios and performance characteristics were calculated.

Results.

Although Charlson score and duration of intensive care unit stay were significant predictors in all models, factors that indicated use or presence of a central line were also important. Device utilization rates derived from the algorithmic models were as accurate as those obtained using manual sampling.

Conclusions.

Automated calculation of central vascular catheter use is both feasible and accurate, providing estimates statistically similar to those obtained using manual surveillance. Prediction modeling of central vascular catheter use may enable automated surveillance of bloodstream infections and enhance important prevention interventions, such as timely removal of unnecessary central lines.

Catheter-Associated Bloodstream Infections in General Medical Patients Outside the Intensive Care Unit: A Surveillance Study

Objective.

To determine the incidence of central venous catheter (CVC)-associated bloodstream infection (CA-BSI) among patients admitted to general medical wards outside the intensive care unit (ICU).

Design.

Prospective cohort study performed over a 13-month period, from April 1, 2002, through April 30, 2003.

Setting.

Four selected general medical wards at Barnes-Jewish Hospital, a 1,250-bed teaching hospital in Saint Louis, Missouri.

Patients.

All patients admitted to 4 general medical wards.

Results.

A total of 7,337 catheter-days were observed during 33,174 patient-days. The device utilization ratio (defined as the number of catheter-days divided by the number of patient-days) was 0.22 overall and was similar among the 4 wards (0.21, 0.25, 0.19, and 0.24). Forty-two episodes of CA-BSI were identified (rate, 5.7 infections per 1,000 catheter-days). Twenty-four (57%) of the 42 cases of CA-BSI were caused by gram-positive bacteria: 10 isolates (24%) were coagulase-negative staphylococci, 10 (24%) were Enterococcus species, and 3 (7%) were Staphylococcus aureus. Gram-negative bacteria caused 7 infections (17%). Five CA-BSIs (12%) were caused by Candida albicans, and 5 infections (12%) had a polymicrobial etiology. Thirty-five patients (83%) with CA-BSI had nontunneled CVCs in place.

Conclusions.

Non-ICU medical wards in the study hospital had device utilization rates that were considerably lower than those of medical ICUs, but CA-BSI rates were similar to CA-BSI rates in medical ICUs in the United States. Studies of catheter utilization and on CVC insertion and care should be performed on medical wards. CA-BSI prevention strategies that have been used in ICUs should be studied on medical wards.

Multicenter evaluation of computer automated versus traditional surveillance of hospital-acquired bloodstream infections

OBJECTIVE

Central line-associated bloodstream infection (BSI) rates are a key quality metric for comparing hospital quality and safety. Traditional BSI surveillance may be limited by interrater variability. We assessed whether a computer-automated method of central line-associated BSI detection can improve the validity of surveillance.

DESIGN

Retrospective cohort study.

SETTING

Eight medical and surgical intensive care units (ICUs) in 4 academic medical centers.

METHODS

Traditional surveillance (by hospital staff) and computer algorithm surveillance were each compared against a retrospective audit review using a random sample of blood culture episodes during the period 2004-2007 from which an organism was recovered. Episode-level agreement with audit review was measured with κ statistics, and differences were assessed using the test of equal κ coefficients. Linear regression was used to assess the relationship between surveillance performance (κ) and surveillance-reported BSI rates (BSIs per 1,000 central line-days).

RESULTS

We evaluated 664 blood culture episodes. Agreement with audit review was significantly lower for traditional surveillance (κ [95% confidence interval (CI) = 0.44 [0.37-0.51]) than computer algorithm surveillance (κ [95% CI] = 0.58; P = .001). Agreement between traditional surveillance and audit review was heterogeneous across ICUs (P = .01); furthermore, traditional surveillance performed worse among ICUs reporting lower (better) BSI rates (P = .001). In contrast, computer algorithm performance was consistent across ICUs and across the range of computer-reported central line-associated BSI rates. Conclusions: Compared with traditional surveillance of bloodstream infections, computer automated surveillance improves accuracy and reliability, making interfacility performance comparisons more valid.

Improving the work efficiency of healthcare-associated infection surveillance using electronic medical records

In this study, we developed an integrated hospital-associated urinary tract infection (HAUTI) surveillance information system (called iHAUTISIS) based on existing electronic medical records (EMR) systems for improving the work efficiency of infection control professionals (ICPs) in a 730-bed, tertiary-care teaching hospital in Taiwan. The iHAUTISIS can automatically collect data relevant to HAUTI surveillance from the different EMR systems, and provides a visualization dashboard that helps ICPs make better surveillance plans and facilitates their surveillance work. In order to measure the system performance, we also created a generic model for comparing the ICPs' work efficiency when using existing electronic culture-based surveillance information system (eCBSIS) and iHAUTISIS, respectively. This model can demonstrate a patient's state (unsuspected, suspected, and confirmed) and corresponding time spent on surveillance tasks performed by ICPs for the patient in that state. The study results showed that the iHAUTISIS performed better than the eCBSIS in terms of ICPs' time cost. It reduced the time by 73.27 s, when using iHAUTISIS (114.26 s) and eCBSIS (187.53 s), for each patient on average. With increased adoption of EMR systems, the development of the integrated HAI surveillance information systems would be more and more cost-effective. Moreover, the iHAUTISIS adopted web-based technology that enables ICPs to online access patient's surveillance information using laptops or mobile devices. Therefore, our system can further facilitate the HAI surveillance and reduce ICPs' surveillance workloads.

No specific time window distinguishes between community-, healthcare-, and hospital-acquired bacteremia, but they are prognostically robust

OBJECTIVE

We examined whether specific time windows after hospital admission reflected a sharp transition between community and hospital acquisition of bacteremia. We further examined whether different time windows to distinguish between community acquisition, healthcare association (HCA), and hospital acquisition influenced the results of prognostic models.

DESIGN

Population-based cohort study.

SETTING

Hospitals in 3 areas of Denmark (2.3 million inhabitants) during 2000-2011.

METHODS

We computed graphs depicting proportions of males, absence of comorbidity, microorganisms, and 30-day mortality pertaining to bacteremia 0, 1, 2, …, 30, and 31 days and later after admission. Next, we assessed whether different admission (0-1, 0-2, 0-3, 0-7 days) and HCA (30, 90 days) time windows were associated with changes in odds ratio (OR) and area under the receiver operating characteristic (ROC) curve for 30-day mortality, adjusting for sex, age, comorbidity, and microorganisms.

RESULTS

For 56,606 bacteremic episodes, no sharp transitions were detected on a specific day after admission. Among the 8 combined time windows, ORs for 30-day mortality varied from 1.30 (95% confidence interval [CI], 1.23-1.37) to 1.99 (95% CI, 1.48-2.67) for HCA and from 1.36 (95% CI, 1.24-1.50) to 2.53 (95% CI, 2.01-3.20) for hospital acquisition compared with community acquisition. Area under the ROC curve changed marginally from 0.684 (95% CI, 0.679-0.689) to 0.700 (95% CI, 0.695-0.705).

CONCLUSIONS

No time transitions unanimously distinguished between community and hospital acquisition with regard to sex, comorbidity, or microorganisms, and no difference in 30-day mortality was seen for HCA patients in relation to a 30- or 90-day time window. ORs decreased consistently in the order of hospital acquisition, HCA, and community acquisition, regardless of time window combination, and differences in area under the ROC curve were immaterial.

Clinical microbiology informatics

The clinical microbiology laboratory has responsibilities ranging from characterizing the causative agent in a patient's infection to helping detect global disease outbreaks. All of these processes are increasingly becoming partnered more intimately with informatics. Effective application of informatics tools can increase the accuracy, timeliness, and completeness of microbiology testing while decreasing the laboratory workload, which can lead to optimized laboratory workflow and decreased costs. Informatics is poised to be increasingly relevant in clinical microbiology, with the advent of total laboratory automation, complex instrument interfaces, electronic health records, clinical decision support tools, and the clinical implementation of microbial genome sequencing. This review discusses the diverse informatics aspects that are relevant to the clinical microbiology laboratory, including the following: the microbiology laboratory information system, decision support tools, expert systems, instrument interfaces, total laboratory automation, telemicrobiology, automated image analysis, nucleic acid sequence databases, electronic reporting of infectious agents to public health agencies, and disease outbreak surveillance. The breadth and utility of informatics tools used in clinical microbiology have made them indispensable to contemporary clinical and laboratory practice. Continued advances in technology and development of these informatics tools will further improve patient and public health care in the future.