Brompton Harefield Infection Score (BHIS): Development and validation of a stratification tool for predicting risk of surgical site infection after coronary artery bypass grafting

Developing a calculable risk prediction model for sternal wound infection after median sternotomy: a retrospective study

Background

Diagnosing sternal wound infection (SWI) following median sternotomy remains laborious and troublesome, resulting in high mortality rates and great harm to patients. Early intervention and prevention are critical and challenging. This study aimed to develop a simple risk prediction model to identify high-risk populations of SWI and to guide examination programs and intervention strategies.

Methods

A retrospective analysis was conducted on the clinical data obtained from 6715 patients who underwent median sternotomy between January 2016 and December 2020. The least absolute shrink and selection operator (LASSO) regression method selected the optimal subset of predictors, and multivariate logistic regression helped screen the significant factors. The nomogram model was built based on all significant factors. Area under the curve (AUC), calibration curve and decision curve analysis (DCA) were used to assess the model's performance.

Results

LASSO regression analysis selected an optimal subset containing nine predictors that were all statistically significant in multivariate logistic regression analysis. Independent risk factors of SWI included female [odds ratio (OR) = 3.405, 95% confidence interval (CI) = 2.535-4.573], chronic obstructive pulmonary disease (OR = 4.679, 95% CI = 2.916-7.508), drinking (OR = 2.025, 95% CI = 1.437-2.855), smoking (OR = 7.059, 95% CI = 5.034-9.898), re-operation (OR = 3.235, 95% CI = 1.087-9.623), heart failure (OR = 1.555, 95% CI = 1.200-2.016) and repeated endotracheal intubation (OR = 1.975, 95% CI = 1.405-2.774). Protective factors included bone wax (OR = 0.674, 95% CI = 0.538-0.843) and chest physiotherapy (OR = 0.446, 95% CI = 0.248-0.802). The AUC of the nomogram was 0.770 (95% CI = 0.745-0.795) with relatively good sensitivity (0.798) and accuracy (0.620), exhibiting moderately good discernment. The model also showed an excellent fitting degree on the calibration curve. Finally, the DCA presented a remarkable net benefit.

Conclusions

A visual and convenient nomogram-based risk calculator built on disease-associated predictors might help clinicians with the early identification of high-risk patients of SWI and timely intervention.

Sternal wound infections following internal mammary artery grafts for a coronary bypass: A meta-analysis

The meta-analysis aims to evaluate and compare the sternal wound infections following internal mammary artery grafts for a coronary bypass. Examinations comparing bilateral internal mammary artery to single internal mammary artery for coronary artery bypass grafting were among the meta-analyses from various languages that met the inclusion criteria. Using dichotomous random- or fixed-effect models, the results of these investigations were examined, and the Odd Ratio (OR) with 95% confidence intervals (CIs) was computed. A total of 31 examinations from 2001 to 2023 were recruited for the current analysis including 181 503 personals with coronary artery bypass grafting. Bilateral internal mammary artery had significantly higher sternal wound infection (OR, 1.51; 95% CI, 1.37-1.68, p < 0.001), superficial sternal wound infection (OR, 1.72; 95% CI, 1.16-2.56, p = 0.007), deep sternal wound infection (OR, 1.62; 95% CI, 1.41-1.86, p < 0.001), sternal wound infection in diabetics (OR, 1.48; 95% CI, 1.16-1.90, p = 0.002), sternal wound infection in elderly (OR, 1.38; 95% CI, 1.22-1.57, p < 0.001), sternal wound infection in pedicled preparation (OR, 1.70; 95% CI, 1.30-2.23, p < 0.001) and sternal wound infection in skeletonized preparation (OR, 1.40; 95% CI, 1.09-1.81, p = 0.009) compared to single internal mammary artery in personals with coronary artery bypass grafting. Bilateral internal mammary artery grafting is linked to a higher risk of impaired wound healing, particularly in diabetic individuals, elderly, pedicled preparation, and skeletonized preparation. Nevertheless, caution should be exercised while interacting with its values since examinations were performed by different surgeons with different skills on different types of personals.

Clinical prediction models and risk tools for early detection of patients at risk of surgical site infection and surgical wound dehiscence: a scoping review

OBJECTIVE

Despite advances in surgical techniques, intraoperative practice and a plethora of advanced wound therapies, surgical wound complications (SWCs), such as surgical site infection (SSI) and surgical wound dehiscence (SWD), continue to pose a considerable burden to the patient and healthcare setting. Predicting those patients at risk of a SWC may give patients and healthcare providers the opportunity to implement a tailored prevention plan or potentially ameliorate known risk factors to improve patient postoperative outcomes.

METHOD

A scoping review of the literature for studies which reported predictive power and internal/external validity of risk tools for clinical use in predicting patients at risk of SWCs after surgery was conducted. An electronic search of three databases and two registries was carried out with date restrictions. The search terms included 'prediction surgical site infection' and 'prediction surgical wound dehiscence'.

RESULTS

A total of 73 records were identified from the database search, of which six studies met the inclusion criteria. Of these, the majority of validated risk tools were predominantly within the cardiothoracic domain, and targeted morbidity and mortality outcomes. There were four risk tools specifically targeting SWCs following surgery.

CONCLUSION

The findings of this review have highlighted an absence of well-developed risk tools specifically for SSI and/or SWD in most surgical populations. This review suggests that further research is required for the development and clinical implementation of rigorously validated and fit-for-purpose risk tools for predicting patients at risk of SWCs following surgery. The ability to predict such patients enables the implementation of preventive strategies, such as the use of prophylactic antibiotics, delayed timing of surgery, or advanced wound therapies following a procedure.

Impact of Antibiotic Prophylaxis on Surgical Site Infections in Cardiac Surgery

(1) Background: Cephalosporins (CA) are the first-line antibiotic prophylaxis recommended to prevent surgical site infection (SSI) after cardiac surgery. The combination of vancomycin/gentamicin (VGA) might represent a good alternative, but few studies have evaluated its efficacy in SSI prevention. (2) Methods: A single-centre retrospective study was conducted over a 13-year period in all consecutive adult patients undergoing elective cardiac surgery. Patients were stratified according to the type of antibiotic prophylaxis. CA served as the first-line prophylaxis, and VGA was used as the second-line prophylaxis. The primary endpoint was SSI occurrence at 90 days, which was defined as the need for reoperation due to SSI. (3) Results: In total, 14,960 adult patients treated consecutively from 2006 to 2019 were included in this study, of whom 1774 (12%) received VGA and 540 (3.7%) developed SSI. VGA patients had higher severity with increased 90-day mortality. Nevertheless, the frequency of SSI was similar between CA and VGA patients. However, the microbiological aetiologies were different, with more Gram-negative bacteria noted in the VGA group. (4) Conclusions: VGA seems to be as effective as CA in preventing SSI.

A single-patient-use ECG system for cardiothoracic surgery admissions in the UK: A cost-consequence analysis

Background

Deep sternal wound infections (DSWI) are severe complications in up to 1.36% of coronary artery bypass grafting (CABG) procedures in the United Kingdom. Each event adds between £4,000 and £11,000 in healthcare costs, owing primarily to prolonged hospitalisations. ECG devices have been shown to convey infection throughout perioperative CABG. On the other hand, single-patient ECG devices (spECG) can effectively reduce the incidence of surgical site infections (SSI), including DSWI, but no assessment of spECG impact in NHS cardiac units has been conducted.

Methods

To estimate the impact of spECG on NHS cardiac units, we conducted a cost-consequence analysis modeling the CABG care pathway in the United Kingdom using Simul8 software for a probabilistic, individual-patient simulation. The simulation time was 1 year, with each patient followed from admission through 30 days post-discharge. The base case simulation mirrors the cardiac unit of Bart Health NHS Trust, London. A total of 2,183 patients are generated with demographic and clinical attributes from probabilistic distributions informed by hospital-specific inputs from NHS Digital Data. The Brompton Harefield Infection Score (BHIS) is allocated to gauge the risk of SSI. Results are averaged across 50 independent and randomly seeded iterations.

Results

Simulation results indicate a base-case savings of £388 per patient, determined by the incidence of infections rather than the number of CABG procedures. In the base-case simulation, the mean cost of care with rECG was £13,096, whereas the mean cost with spECG was £12,708, resulting in a cost saving of £388 (2021 GBP). The simulation yielded an overall 8.6% SSI incidence rECG, whereas the incidence of SSIs with spECG was 6.9%. The model was most sensitive to changes in general ward and ICU costs, and infection incidence was a stronger predictor of potential per-patient savings than annual CABG volume.

Conclusion

Single-patient ECG is a sustainable and effective alternative to reusable ECG cables and lead wires in terms of patient safety and resource allocation.

Surgical Site Infection in Cardiac Surgery

Surgical site infections (SSIs) are one of the most significant complications in surgical patients and are strongly associated with poorer prognosis. Due to their aggressive character, cardiac surgical procedures carry a particular high risk of postoperative infection, with infection incidence rates ranging from a reported 3.5% and 26.8% in cardiac surgery patients. Given the specific nature of cardiac surgical procedures, sternal wound and graft harvesting site infections are the most common SSIs. Undoubtedly, DSWIs, including mediastinitis, in cardiac surgery patients remain a significant clinical problem as they are associated with increased hospital stay, substantial medical costs and high mortality, ranging from 3% to 20%. In SSI prevention, it is important to implement procedures reducing preoperative risk factors, such as: obesity, hypoalbuminemia, abnormal glucose levels, smoking and S. aureus carriage. For decolonisation of S. aureus carriers prior to cardiac surgery, it is recommended to administer nasal mupirocin, together with baths using chlorhexidine-based agents. Perioperative management also involves antibiotic prophylaxis, surgical site preparation, topical antibiotic administration and the maintenance of normal glucose levels. SSI treatment involves surgical intervention, NPWT application and antibiotic therapy.

Development and Validation of a Risk Score for Predicting Post-acute Myocardial Infarction Infection in Patients Undergoing Percutaneous Coronary Intervention: Study Protocol for an Observational Study

Background: Post-acute myocardial infarction (post-AMI) infection is an infrequent but important and serious complication in patients with ST-segment elevation myocardial infarction (STEMI) treated with percutaneous coronary intervention (PCI). Predicting its occurrence is essential for future prevention. However, little is known about the prediction of post-AMI infection in such patients to date. This study aims to develop and validate a new risk score based on risk factors for early prediction of infection in STEMI patients undergoing PCI. Methods: This prospective, multi-center and observational study assesses the predictive value of risk score for post-AMI infection among a cohort of patients hospitalized due to STEMI. The STEMI patients undergoing PCI enrolled between January 1st 2010 and May 31st 2016 were served as a development cohort while those enrolled from June 1st 2016 to May 31st 2018 were served as validation cohort. The primary endpoint was post-AMI infection during hospitalization, defined as infection requiring antibiotics (reflecting the clinical influence of infection compatible with the necessity for additional treatment), and all-cause death and major adverse cardiovascular events (MACE) including all-cause death, recurrent myocardial infarction, target vessel revascularization, and stroke were considered as secondary endpoints. The risk score model based on risk factors was established using stepwise logistic regression, and will be validated in other centers and external patients with non-ST-elevation acute coronary syndrome (NSTE-ACS). Results: This study will provide evidence on prognostic property, reliability of scoring, comparative performance, and suitability of the novel model for screening purpose in order to be recommended for clinical practice. Discussion: Our study is designed to develop and validate a clinical risk score for predicting infection in participants with STEMI who have undergone PCI. This simple tool may therefore improve evaluation of post-AMI infection and enhance future researches into the best practices to prevent or reduce infection in such patients. Clinical Trial Registration: www.chictr.org.cn, identifier: ChiCTR1900028278.

Higher Surgery and Recovery Room Air Pressures Associated with Reduced Surgical Site Infection Risk

BACKGROUND

Incisional surgical site infections (SSIs) following coronary artery bypass grafting (CABG) prolong hospital stays, elevate healthcare costs and increase likelihood of further complications. High air pressure deactivates bacteria and is utilized for commercial food preservation, assuring microbiologically safe pharmaceuticals and sanitizing instruments. However, research on utilizing air pressure deactivation thresholds in surgical and postoperative rooms to reduce rates of SSIs is lacking.

METHODS

A case-control study of 801 CABG patients, 128 SSI cases and 673 controls was conducted from January 1, 2006 through March 31, 2009 in Yerevan, Armenia. Patient and surgery characteristics, air pressure measurements and seasons were selected as independent variables with SSI rates as the outcome. The novel threshold regression analysis was used to determine potential air pressure bacterial deactivation thresholds. A final multivariate logistic regression model adjusted for confounders.

RESULTS

Overall, bacterial deactivation air pressure threshold was 694.2 mmHg, with the presence of infection for higher air pressure values not statistically significant from zero. Individual deactivation thresholds for Staphylococcus epidermidis (threshold = 694.2 mmHg) and Escherichia coli (threshold = 689.2) showed similar patterns. Multivariate logistic regression showed air pressure above the deactivation threshold was highly protective against SSIs with adjOR = 0.27 (p-value = 0.009, 95%CI: 0.10-0.72). Other SSI risk factors included female sex, adjOR = 2.12 (p-value = 0.006, 95%CI: 1.24-3.62), diabetes, adjOR = 2.61 (p-value < 0.001, 95%CI: 1.72-3.96) and longer time on ventilator, adjOdds = 1.01 (p-value = 0.012, 95%CI: 1.00-1.02).

CONCLUSION

Maintaining air pressures in operating and postoperative rooms exceeding bacterial-deactivation thresholds might substantially reduce SSI rates following surgery. Further research should identify specific bacterial-deactivation air pressure thresholds in surgical and postoperative rooms to reduce SSI rates, especially for drug-resistant bacteria.

Risk stratification tool for all surgical site infections after coronary artery bypass grafting

OBJECTIVE

To develop a risk score for surgical site infections (SSIs) after coronary artery bypass grafting (CABG).

DESIGN

Retrospective study.

SETTING

University hospital.

PATIENTS

A derivation sample of 7,090 consecutive isolated or combined CABG patients and 2 validation samples (2,660 total patients).

METHODS

Predictors of SSIs were identified by multivariable analyses from the derivation sample, and a risk stratification tool (additive and logistic) for all SSIs after CABG (acronym, ASSIST) was created. Accuracy of prediction was evaluated with C-statistic and compared 1:1 (using the Hanley-McNeil method) with most relevant risk scores for SSIs after CABG. Both internal (1,000 bootstrap replications) and external validation were performed.

RESULTS

SSIs occurred in 724 (10.2%) cases and 2 models of ASSIST were created, including either baseline patient characteristics alone or combined with other perioperative factors. Female gender, body mass index >29.3 kg/m2, diabetes, chronic obstructive pulmonary disease, extracardiac arteriopathy, angina at rest, and nonelective surgical priority were predictors of SSIs common to both models, which outperformed (P < .0001) 6 specific risk scores (10 models) for SSIs after CABG. Although ASSIST performed differently in the 2 validation samples, in both, as well as in the derivation data set, the combined model outweighed (albeit not always significantly) the preoperative-only model, both for additive and logistic ASSIST.

CONCLUSIONS

In the derivation data set, ASSIST outperformed specific risk scores in predicting SSIs after CABG. The combined model had a higher accuracy of prediction than the preoperative-only model both in the derivation and validation samples. Additive and logistic ASSIST showed equivalent performance.

Predicting cardiac surgical site infection: development and validation of the Barts Surgical Infection Risk tool

OBJECTIVES

The objective of this study was to develop and validate a new risk tool (Barts Surgical Infection Risk (B-SIR)) to predict surgical site infection (SSI) risk after all types of adult cardiac surgery, and compare its predictive ability against existing (but procedure-specific) tools: Brompton-Harefield Infection Score (BHIS), Australian Clinical Risk Index (ACRI), National Nosocomial Infection Surveillance (NNIS).

STUDY DESIGN AND SETTING

Single-center retrospective analysis of prospectively collected data including 2,449 patients undergoing cardiac surgery between January 2016 and December 2017 in a European tertiary hospital. Thirty-four variables associated with SSI risk after cardiac surgery were collated from three local databases. Independent predictors were identified using stepwise multivariable logistic regression. Bootstrap resampling was conducted to validate the model. Hosmer-Lemeshow goodness-of-fit test was performed to assess calibration of scores.

RESULTS

The B-SIR model was constructed from six independent predictors female gender, body mass index >30, diabetes, left ventricular ejection fraction <45%, peripheral vascular disease and operation type, and the risk estimates were derived. The receiver operating characteristics curve for B-SIR was 0.682, vs. 0.603 for BHIS, 0.618 for ACRI, and 0.482 for the NNIS tool.

CONCLUSION

B-SIR provides greater predictive power of SSI risk after cardiac surgery compared with existing tools in our population.

Wound Infection after CABG Using Internal Mammary Artery Grafts: A Meta-Analysis

BACKGROUND

 Internal mammary arteries show better long-term patency rates than venous grafts. The use of both mammary arteries is associated with a higher risk of sternal wound infections. This meta-analysis was designed to assess the incidence of a wound healing disorder after bilateral compared with single mammary artery bypass grafting. Compared with existing meta-analysis this paper includes more current literature and one randomized controlled trial.

METHODS

 A literature search was performed using PubMed and the Cochrane Library. The quality of the articles was assessed by the Newcastle Ottawa Scale. The odds ratio was used as a measure of the chance of developing a wound healing disorder after bilateral internal mammary artery (BIMA) surgery. Meta-analyses were performed for different subgroups.

RESULTS

 Twenty studies met the quality criteria, including one randomized controlled trial. The use of both mammary arteries significantly increased the risk of superficial (odds ratio [OR] 1.72) and deep (OR 1.75) wound healing disorder in the total population (OR 1.80) as well as in the diabetic subgroup (OR 1.38) and with both preparation techniques. The increased risk with BIMA grafting was present independently of the preparation technique (pedicled: OR 1.89, skeletonized: OR 1.37).

CONCLUSION

 Bilateral internal mammary artery grafting, especially in high-risk and diabetic patients, is associated with an increased risk of wound healing impairment. Skeletonized preparation does not eliminate the elevated wound healing disorder risk after BIMA use.

Preoperative risk stratification of deep sternal wound infection after coronary surgery

OBJECTIVE

To develop a risk score for deep sternal wound infection (DSWI) after isolated coronary artery bypass grafting (CABG).

DESIGN

Multicenter, prospective study.

SETTING

Tertiary-care referral hospitals.

PARTICIPANTS

The study included 7,352 patients from the European multicenter coronary artery bypass grafting (E-CABG) registry.

INTERVENTION

Isolated CABG.

METHODS

An additive risk score (the E-CABG DSWI score) was estimated from the derivation data set (66.7% of patients), and its performance was assessed in the validation data set (33.3% of patients).

RESULTS

DSWI occurred in 181 (2.5%) patients and increased 1-year mortality (adjusted hazard ratio, 4.275; 95% confidence interval [CI], 2.804-6.517). Female gender (odds ratio [OR], 1.804; 95% CI, 1.161-2.802), body mass index ≥30 kg/m2 (OR, 1.729; 95% CI, 1.166-2.562), glomerular filtration rate <45 mL/min/1.73 m2 (OR, 2.410; 95% CI, 1.413-4.111), diabetes (OR, 1.741; 95% CI, 1.178-2.573), pulmonary disease (OR, 1.935; 95% CI, 1.178-3.180), atrial fibrillation (OR, 1.854; 95% CI, 1.096-3.138), critical preoperative state (OR, 2.196; 95% CI, 1.209-3.891), and bilateral internal mammary artery grafting (OR, 2.088; 95% CI, 1.422-3.066) were predictors of DSWI (derivation data set). An additive risk score was calculated by assigning 1 point to each of these independent risk factors for DSWI. In the validation data set, the rate of DSWI increased along with the E-CABG DSWI scores (score of 0, 1.0%; score of 1, 1.8%; score of 2, 2.2%; score of 3, 6.9%; score ≥4: 12.1%; P < .0001). Net reclassification improvement, integrated discrimination improvement, and decision curve analysis showed that the E-CABG DSWI score performed better than other risk scores.

CONCLUSIONS

DSWI is associated with poor outcome after CABG, and its risk can be stratified using the E-CABG DSWI score.

TRIAL REGISTRATION

clinicaltrials.gov identifier: NCT02319083.

Multi-centre prospective internal and external evaluation of the Brompton Harefield Infection Score (BHIS)

BACKGROUND

Previously, we reported that the Brompton Harefield Infection Score (BHIS) accurately predicts surgical site infection (SSI) after coronary artery bypass grafting (CABG). The BHIS was developed using two-centre data and stratifies SSI risk into three groups based on female gender, diabetes or HbA1c > 7.5%, body mass index ≥ 35, left ventricular ejection fraction < 45% and emergency surgery. The purpose of this study was to prospectively evaluate BHIS internally as well as externally.

METHODS

Multi-centre prospective evaluation involving three tertiary centres took place between October 2012 and November 2015. SSI was classified using the Public Health England protocol. Receiver operating characteristic (ROC) curves assessed predictive accuracy.

RESULTS

Across the four hospital sites, 168 of 4308 (3.9%) CABG patients had a SSI. Categorising the hospitals by BHIS score revealed that 65% of all patients were low risk (BHIS 0-1), 26% were medium risk (BHIS 2-3) and 8% were high risk (BHIS ≥ 4). The area under the ROC curve was in the range of 0.702-0.785. Overall area under the ROC curve was 0.709.

CONCLUSIONS

BHIS provides a novel, internally and externally evaluated score for a patient's risk of SSI after CABG. It enables clinicians to focus on strategies to prospectively identify high-risk patients and improve outcomes.

Development and Validation of a Stratification Tool for Predicting Risk of Deep Sternal Wound Infection after Coronary Artery Bypass Grafting at a Brazilian Hospital

Objective

Deep sternal wound infection following coronary artery bypass grafting is a serious complication associated with significant morbidity and mortality. Despite the substantial impact of deep sternal wound infection, there is a lack of specific risk stratification tools to predict this complication after coronary artery bypass grafting. This study was undertaken to develop a specific prognostic scoring system for the development of deep sternal wound infection that could risk-stratify patients undergoing coronary artery bypass grafting and be applied right after the surgical procedure.

Methods

Between March 2007 and August 2016, continuous, prospective surveillance data on deep sternal wound infection and a set of 27 variables of 1500 patients were collected. Using binary logistic regression analysis, we identified independent predictors of deep sternal wound infection. Initially we developed a predictive model in a subset of 500 patients. Dataset was expanded to other 1000 consecutive cases and a final model and risk score were derived. Calibration of the scores was performed using the Hosmer-Lemeshow test.

Results

The model had area under Receiver Operating Characteristic (ROC) curve of 0.729 (0.821 for preliminary dataset). Baseline risk score incorporated independent predictors of deep sternal wound infection: obesity (P=0.046; OR 2.58; 95% CI 1.11-6.68), diabetes (P=0.046; OR 2.61; 95% CI 1.12-6.63), smoking (P=0.008; OR 2.10; 95% CI 1.12-4.67), pedicled internal thoracic artery (P=0.012; OR 5.11; 95% CI 1.42-18.40), and on-pump coronary artery bypass grafting (P=0.042; OR 2.20; 95% CI 1.13-5.81). A risk stratification system was, then, developed.

Conclusion

This tool effectively predicts deep sternal wound infection risk at our center and may help with risk stratification in relation to public reporting and targeted prevention strategies in patients undergoing coronary artery bypass grafting.

Incidence of surgical site infections cannot be derived reliably from point prevalence survey data in Dutch hospitals

Thorough studies on whether point prevalence surveys of healthcare-associated infections (HAIs) can be used to reliably estimate incidence of surgical site infections (SSIs) are scarce. We examined this topic using surveillance data of 58 hospitals that participated in two Dutch national surveillances; HAI prevalence and SSI incidence surveillance, respectively. First, we simulated daily prevalences of SSIs from incidence data. Subsequently, Rhame & Sudderth's formula was used to estimate SSI incidence from prevalence. Finally, we developed random-effects models to predict SSI incidence from prevalence and other relevant variables. The prevalences simulated from incidence data indicated that daily prevalence varied greatly. Incidences calculated with Rhame & Sudderth's formula often had values below zero, due to the large number of SSIs occurring post-discharge. Excluding these SSIs, still resulted in poor correlation between calculated and observed incidence. The two models best predicting total incidence and incidence during initial hospital stay both performed poorly (proportion of explained variance of 0·25 and 0·10, respectively). In conclusion, incidence of SSIs cannot be reliably estimated from point prevalence data in Dutch hospitals by any of the applied methods. We therefore conclude that prevalence surveys are not a useful measure to give reliable insight into incidence of SSIs.

Classification of hospital acquired complications using temporal clinical information from a large electronic health record

Hospital acquired complications (HACs) are serious problems affecting modern day healthcare institutions. It is estimated that HACs result in an approximately 10% increase in total inpatient hospital costs across US hospitals. With US hospital spending totaling nearly $900 billion per annum, the damages caused by HACs are no small matter. Early detection and prevention of HACs could greatly reduce strains on the US healthcare system and improve patient morbidity & mortality rates. Here, we describe a machine-learning model for predicting the occurrence of HACs within five distinct categories using temporal clinical data. Using our approach, we find that at least $10 billion of excessive hospital costs could be saved in the US alone, with the institution of effective preventive measures. In addition, we also identify several keystone features that demonstrate high predictive power for HACs over different time periods following patient admission. The classifiers and features analyzed in this study show high promise of being able to be used for accurate prediction of HACs in clinical settings, and furthermore provide novel insights into the contribution of various clinical factors to the risk of developing HACs as a function of healthcare system exposure.