Surgical Site Infection Surveillance Research Articles

Comparison of an algorithm, and coding data, with traditional surveillance to identify surgical site infections in Australia: a retrospective multi-centred cohort study

Surveillance of healthcare associated infections (HAIs) in Australia is disparate, resource intensive, unsustainable and provides limited information. Traditional HAI surveillance is time intensive and agreement levels between clinicians has been shown to be variable. The aim was to compare two methods, a semi-automated algorithm, and coding data, against traditional surgical site infections (SSI) surveillance methods. This retrospective multi-centre cohort study included all patients undergoing a hip (HPRO) or knee (KPRO) joint replacements and coronary artery bypass graft (CBGB) surgery over 2 years at 2 large metropolitan hospitals. Routine SSI data were obtained via the infection prevention team, a previously developed algorithm was applied to all patient records, and the ICD-10-AM data were searched for those categorised as having a SSI. Overall, 1447, 1416 and 1026 patients who underwent HPRO, KPRO and CBGB respectively were included. The highest Se values were generated by the algorithm: HPRO D/O 0.87(95%CI:0.66-0.96), CBGB 0.86(95%CI:0.64-0.96) and HPRO all SSI 0.77(95%CI:0.57-89), the lowest Se was Code CBGB D/O 0.03(95%CI:0.00-0.21). The highest PPV values were generated by the algorithm: HPRO D/O 0.97(95%CI:0.77-0.99), CBGB D/O 0.97(95%CI:0.76-0.99) and the Code HPRO D/O 0.9(95%CI:0.66-0.99). Both the algorithm and coding data resulted in a substantial reduction in the number of medical records required to review. The application of algorithms to enhance SSI surveillance demonstrates high accuracy in identifying patient records that require review by infection prevention teams to determine the presence of an SSI. Coding data alone should not be used to identify SSI's.

An explainable long short-term memory network for surgical site infection identification

BackgroundCurrently, surgical site infection surveillance relies on labor-intensive manual chart review. Recently suggested solutions involve machine learning to identify surgical site infections directly from the medical record. Deep learning is a form of machine learning that has historically performed better than traditional methods while being harder to interpret. We propose a deep learning model, a long short-term memory network, for the identification of surgical site infection from the medical record with an attention layer for explainability. MethodsWe retrieved structured data and clinical notes from the University of Utah Health System’s electronic health care record for operative events randomly selected for manual chart review from January 2016 to June 2021. Surgical site infection occurring within 30 days of surgery was determined according to the National Surgical Quality Improvement Program definition. We trained the long short-term memory model along with traditional machine learning models for comparison. We calculated several performance metrics from a holdout test set and performed additional analyses to understand the performance of the long short-term memory, including an explainability analysis. ResultsSurgical site infection was present in 4.7% of the total 9,185 operative events. The area under the receiver operating characteristic curve and sensitivity of the long short-term memory was higher (area under the receiver operating characteristic curve: 0.954, sensitivity: 0.920) compared to the top traditional model (area under the receiver operating characteristic curve: 0.937, sensitivity: 0.736). The top 5 features of the long short-term memory included 2 procedure codes and 3 laboratory values. ConclusionSurgical site infection surveillance is vital for the reduction of surgical site infection rates. Our explainable long short-term memory achieved a comparable area under the receiver operating characteristic curve and greater sensitivity when compared to traditional machine learning methods. With explainable deep learning, automated surgical site infection surveillance could replace burdensome manual chart review processes.

Surveillance of infections of surgical sites and lower respiratory tracts should be combined: experiences from the German surveillance module for operated patients (OP-KISS), 2018 to 2022.

BackgroundSurveillance of lower respiratory tract infections (LRTI) of operated patients conventionally focuses on intubated patients in intensive care units (ICU). Post-operative immobilisation increases the risk of LRTI not associated with ventilators. Operated patients, however, have thus far not been a primary target for LRTI surveillance.AimWe aimed to describe the applied LRTI surveillance method in the German surveillance module for operated patients (OP-KISS) and to report data between 2018 and 2022.MethodsSurveillance of LRTI can be performed voluntarily in addition to surgical site infection (SSI) surveillance in OP-KISS. We calculated LRTI rates per 100 operations for all procedures combined, as well as for individual surgical groups and procedures. Additionally, a combined post-operative infection rate (SSI and LRTI) was calculated.ResultsSurveillance of LRTI was performed in 4% of all participating OP-KISS departments and for 2% (23,239 of 1,332,438) of all procedures in the OP-KISS database. The pooled LRTI rate was 0.9 per 100 operations, with marked differences between different types of surgery (3.6 for lobectomies, 0.1 for traumatology and orthopaedics). The share of LRTI among all post-operative infections was highly variable. For lobectomies, the LRTI rate was higher than the SSI rate (3.6 vs 1.5 per 100 operations).ConclusionSurveillance of post-operative LRTI is not yet widely adopted by German hospitals. Based on the data in this study, lobectomies represent a prime target for post-operative LRTI surveillance.

Effectiveness of HealthTracker for post-caesarean section surgical site infection surveillance: An intervention study.

This intervention study aimed to implement and evaluate the effectiveness of HealthTracker, a post-discharge surgical site infection surveillance system. Participants were 730 women birthing by caesarean section at a large hospital over a 6-month period. Data were downloaded from clinical data systems and HealthTracker. Receiver operating characteristics were used to assess HealthTracker. Over a 6-month period, 382 women completed HealthTracker, with 83 scoring ≥6, indicating signs and symptoms of surgical site infection. Of this 83, 58 sought advice from health professionals, 29 returned to hospital, and 45 received antibiotics. A total of 20 infections from a total population of 730 were confirmed, with 14 out of 382 respondents confirmed via HealthTracker. Receiver operating characteristics identified HealthTracker as an excellent indicator of surgical site infection. HealthTracker is a feasible mHealth option for monitoring post-discharge surgical site infection post-caesarean section. In addition, by providing alerts, advising women to monitor their symptoms and seek treatment if necessary, HealthTracker has the potential to enhance self-efficacy for surgical wound monitoring at home.

Impact of penicillin allergy labels on surgical site infections in a large UK cohort of gastrointestinal surgery patients.

Studies in the USA, Canada and France have reported higher surgical site infection (SSI) risk in patients with a penicillin allergy label (PAL). Here, we investigate the association between PALs and SSI in the UK, a country with distinct epidemiology of infecting pathogens and range of antimicrobial regimens in routine use. Electronic health records and national SSI surveillance data were collated for a retrospective cohort of gastrointestinal surgery patients at Cambridge University Hospitals NHS Foundation Trust from 1 January 2015 to 31 December 2021. Univariable and multivariable logistic regression were used to examine the effects of PALs and the use of non-β-lactam-based prophylaxis on likelihood of SSI, 30 day post-operative mortality, 7 day post-operative acute kidney injury and 60 day post-operative infection/colonization with antimicrobial-resistant bacteria or Clostridioides difficile. Our data comprised 3644 patients and 4085 operations; 461 were undertaken in the presence of PALs (11.3%). SSI was detected after 435/4085 (10.7%) operations. Neither the presence of PALs, nor the use of non-β-lactam-based prophylaxis were found to be associated with SSI: adjusted OR (aOR) 0.90 (95% CI 0.65-1.25) and 1.20 (0.88-1.62), respectively. PALs were independently associated with increased odds of newly identified MRSA infection/colonization in the 60 days after surgery: aOR 2.71 (95% CI 1.13-6.49). Negative association was observed for newly identified infection/colonization with third-generation cephalosporin-resistant Gram-negative bacteria: aOR 0.38 (95% CI 0.16-0.89). No evidence was found for an association between PALs and the likelihood of SSI in this large UK cohort, suggesting significant international variation in the impact of PALs on surgical patients.

Cohort study to characterise surgical site infections after open surgery in the UK’s National Health Service

ObjectiveTo characterise surgical site infections (SSIs) after open surgery in the UK’s National Health Service.DesignRetrospective cohort analysis of electronic records of patients from Clinical Practice Research Datalink, linked with Hospital...

Healthcare professionals' knowledge and attitudes towards surgical site infection and surveillance: A quasi-experimental study.

SSI is one of the most prevalent healthcare-associated infections and is associated with extended hospital stays, increased need for reoperation and higher hospital readmission rates. Implementing systematic SSI surveillance can reduce these adverse outcomes. Implementing a surveillance system into a hospital is a complex intervention requiring that staff involved in a patient's perioperative journey have the knowledge of SSI prevention, the data required for surveillance, an understanding of how data informs quality improvement initiatives and their role in surveillance. The aim of this study was to evaluate the impact of a complex intervention on the knowledge and attitudes of healthcare professionals towards surgical site infection (SSI), SSI prevention and surveillance in a university hospital setting. The study used a quasi-experimental pre-test-post-test design. The impact of a complex intervention was evaluated by measuring healthcare professionals' (n = 74) knowledge of and attitudes towards SSI and surveillance. Normalisation process theory (NPT) guided the study and the development of the intervention. There was a statistically significant increase in scores on the knowledge of SSI and prevention from pre-intervention to post-test. The knowledge of risk factors scores at post-test was significantly higher than that at pre-intervention. Overall attitudes to SSI prevention and surveillance were good both pre-intervention and post-test but there was a significant change in the attitude of participants. The findings reveal an overall positive impact of the complex intervention on the knowledge and attitude of healthcare professionals relating to SSI, SSI prevention and surveillance; however, the extent of the change varied across items measured.

1469. Development and Implementation of a Predictive Model for a Semi-automated Surgical Site Infection Surveillance System

Abstract Background Automated surveillance systems for healthcare-associated infections have been implemented worldwide, but monitoring surgical site infections (SSI) is still challenging. Surveillance of SSI requires a review of multiple clinical aspects and medical records, which is time-consuming and prone to human error. We aimed to develop a predictive model for SSI and implement a semi-automated surveillance system. Methods The study included 213 major surgeries randomly selected from 25 different types at a tertiary care hospital in Seoul, South Korea, between June 2019 and May 2022. Infection preventionists diagnosed SSI based on US National Healthcare Safety Network criteria. After identifying variables significantly associated with SSI through univariate analysis, we developed several predictive models. We then conducted multivariable logistic regression analysis, calculated coefficients and p-values, and evaluated model performance using Area Under the ROC Curve (AUC) analysis, as well as sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Results Of 213 surgeries, 13 (6.1%) were exploratory laparotomy, while 11 (5.2%) were cesarean section and ventricular shunting each. 74 cases (34.7%) were diagnosed with SSI. The study found that a predictive model that included use of antimicrobial agent, positive culture test, infection-related imaging findings (e.g., CT, MRI, ultrasonography), record of wound dressing, and pus or dehiscence of wound records, identified patients at high risk for SSI with an AUC of 0.853 (95% CI: 0.795-0.911). We weighted these variables and found that with a cutoff of 4 or higher out of a total of 8 points demonstrated sensitivity of 75.7% (95% CI: 65.9-85.5), specificity of 82.7% (95% CI: 76.5-89.0), PPV of 86.5% (95% CI: 80.7-92.3), and NPV of 80.3% (95% CI: 74.9-85.6). Based on this model, a semi-automated surveillance system was developed. Conclusion The predictive model and semi-automated surveillance system of SSI have the potential to improve the efficiency and accuracy. Further optimization of the predictive model by validating it for different types of surgeries and incorporating additional infectious variables and weights for each variable to increase its accuracy and applicability is needed. Disclosures All Authors: No reported disclosures

Association between the introduction of a national targeted intervention program and the incidence of surgical site infections in Swiss acute care hospitals

BackgroundIn Switzerland, the national surgical site infection (SSI) surveillance program showed a modest decrease in SSI rates for different procedures over the last decade. The study aimed to determine whether a multimodal, targeted intervention program in addition to existing SSI surveillance is associated with decreased SSI rates in the participating hospitals.MethodsProspective multicenter pre- and postintervention study conducted in eight Swiss acute care hospitals between 2013 and 2020. All consecutive patients > 18 years undergoing cardiac, colon, or hip/knee replacement surgery were included. The follow-up period was 30 days and one year for implant-related surgery. Patients with at least one follow-up were included. The intervention was to optimize three elements of preoperative management: (i) hair removal; (ii) skin disinfection; and (iii) perioperative antimicrobial prophylaxis. We compared SSI incidence rates (main outcome measure) pre- and postintervention (three years each) adjusted for potential confounders. Poisson generalized linear mixed models fitted to quarter-yearly confirmed SSIs and adjusted for baseline differences between hospitals and procedures. Adherence was routinely monitored through on-site visits.ResultsA total of 10 151 patients were included, with a similar median age pre- and postintervention (69.6 and IQR 60.9, 76.8 years, vs 69.5 and IQR 60.4, 76.8 years, respectively; P = 0.55) and similar proportions of females (44.8% vs. 46.1%, respectively; P = 0.227). Preintervention, 309 SSIs occurred in 5 489 patients (5.6%), compared to 226 infections in 4 662 cases (4.8%, P = 0.09) postintervention. The adjusted incidence rate ratio (aIRR) for overall SSI after intervention implementation was 0.81 (95% CI, 0.68 to 0.96, P = 0.02). For cardiac surgery (n = 2 927), the aIRR of SSI was 0.48 (95% CI, 0.32 to 0.72, P < 0.001). For hip/knee replacement surgery (n = 4 522), the aIRR was 0.88 (95% CI, 0.52 to 1.48, P = 0.63), and for colon surgery (n = 2 702), the aIRR was 0.92 (95% CI, 0.75 to 1.14, P = 0.49).ConclusionsThe SSI intervention bundle was associated with a statistically significant decrease in SSI cases. A significant association was observed for cardiac surgery. Adding a specific intervention program can add value compared to routine surveillance only. Further prevention modules might be necessary for colon and orthopedic surgery.

The augmented value of using clinical notes in semi-automated surveillance of deep surgical site infections after colorectal surgery

BackgroundIn patients who underwent colorectal surgery, an existing semi-automated surveillance algorithm based on structured data achieves high sensitivity in detecting deep surgical site infections (SSI), however, generates a significant number of false positives. The inclusion of unstructured, clinical narratives to the algorithm may decrease the number of patients requiring manual chart review. The aim of this study was to investigate the performance of this semi-automated surveillance algorithm augmented with a natural language processing (NLP) component to improve positive predictive value (PPV) and thus workload reduction (WR).MethodsRetrospective, observational cohort study in patients who underwent colorectal surgery from January 1, 2015, through September 30, 2020. NLP was used to detect keyword counts in clinical notes. Several NLP-algorithms were developed with different count input types and classifiers, and added as component to the original semi-automated algorithm. Traditional manual surveillance was compared with the NLP-augmented surveillance algorithms and sensitivity, specificity, PPV and WR were calculated.ResultsFrom the NLP-augmented models, the decision tree models with discretized counts or binary counts had the best performance (sensitivity 95.1% (95%CI 83.5–99.4%), WR 60.9%) and improved PPV and WR by only 2.6% and 3.6%, respectively, compared to the original algorithm.ConclusionsThe addition of an NLP component to the existing algorithm had modest effect on WR (decrease of 1.4–12.5%), at the cost of sensitivity. For future implementation it will be a trade-off between optimal case-finding techniques versus practical considerations such as acceptability and availability of resources.

PRESS survey: PREvention of surgical site infection-a global pan-specialty survey of practice protocol.

Surgical site infections (SSI) complicate up to 40% of surgical procedures, leading to increased patient morbidity and mortality. Previous research identified disparities in SSI prevention guidelines and clinical practices across different institutions. The study aims to identify variations in SSI prevention practices within and between specialties and financial systems and provide a representation of existing SSI preventative measures to help improve the standardization of SSI prevention practices. This collaborative cross-sectional survey will be aimed at pan-surgical specialties internationally. The study has been designed and will be reported in line with the CROSS and CHERRIES standards. An international study steering committee will design and internally validate the survey in multiple consensus-based rounds. This will be based on SSI prevention measures outlined in the CDC (2017), WHO (2018), NICE (2019), Wounds UK (2020) and the International Surgical Wound Complications Advisory Panel (ISWCAP) guidelines. The questionnaire will include demographics, SSI surveillance, preoperative, peri-operative and postoperative SSI prevention. Data will be collected on participants' surgical specialty, operative grade, of practice and financial healthcare system of practice. The online survey will be designed and disseminated using QualtricsXM Platform™ through national and international surgical colleges and societies, in addition to social media and snowballing. Data collection will be open for 3 months with reminders, and raking will be used to ascertain the sample. Responses will be analyzed, and the chi-square test used to evaluate the impact of SSI prevention variables on responses. Current SSI prevention practice in UK Vascular surgery varies considerably, with little consensus on many measures. Given the inconsistency in guidelines on how to prevent SSIs, there is a need for standardization. This survey will investigate the disparity in SSI preventative measures between different surgical fields and countries.

Barriers and facilitators for surgical site infection surveillance for adult cardiac surgery in a high-income setting: an in-depth exploration

Surgical site infection (SSI) surveillance aims to facilitate a reduction in SSIs through identifying infection rates, benchmarking, triggering clinical review and instituting infection control measures. Participation in surveillance is, however, variable suggesting opportunities to improve wider adoption. To gain an in-depth understanding of the barriers and facilitators for SSI surveillance in a high-income European setting. Key informant interviews with 16 surveillance staff, infection prevention staff, nurses and surgeons from nine cardiac hospitals in England. Data were analysed thematically. SSI surveillance was reported to be resource-intensive. Barriers to surveillance included challenges associated with data collection: data being located in numerous places, multiple SSI data reporting schemes, difficulty in finding denominator data, lack of interface between computerised systems, 'labour intensive' or 'antiquated' methods to collect data (e.g. using postal systems for patient questionnaires). Additional reported concerns included: relevance of definitions, perceived variability in data reporting, lack of surgeon engagement, unsupportive managers, low priority of SSIs among staff, and a 'blame culture' around high SSI rates. Facilitators were increased resources, better use of digital technologies (e.g. remote digital wound monitoring), integrating surveillance within routine clinical work, having champions, mandating surveillance, ensuring a closer relationship between surveillance and improved patient outcomes, increasing the focus on post-discharge surveillance, and integration with primary care data. Using novel interviews with 'front-line' staff, identified opportunities for improving participation in SSI surveillance. Translating these findings into action will increase surveillance activity and bring patient safety benefits to a larger pool of surgical patients.

Association of antimicrobial perioperative prophylaxis with cefuroxime plus metronidazole or amoxicillin/clavulanic acid and surgical site infections in colorectal surgery

ObjectiveTo compare intravenous (IV) amoxicillin/clavulanic acid (A/CA) to IV cefuroxime plus metronidazole (C + M) for preventing surgical site infections (SSI) in colorectal surgery.BackgroundGiven their spectra that include most Enterobacterales and anaerobes, C + M is commonly recommended as prophylaxis of SSI in colorectal surgery. A/CA offers good coverage of Enterobacterales and anaerobes as well, but, in contrast to C + M, it also includes Enterococcus faecalis which is also isolated from patients with SSI and could trigger anastomotic leakage.MethodsData from a Swiss SSI surveillance program were used to compare SSI rates after class II (clean contaminated) colorectal surgery between patients who received C + M and those who received A/CA. We employed multivariable logistic regression to adjust for potential confounders, along with propensity score matching to adjust for group imbalance.ResultsFrom 2009 to 2018, 27,922 patients from 127 hospitals were included. SSI was diagnosed in 3132 (11.2%): 278/1835 (15.1%) in those who received A/CA and 2854/26,087 (10.9%) in those who received C + M (p < 0.001). The crude OR for SSI in the A/CA group as compared to C + M was 1.45 [CI 95% 1.21–1.75]. The adjusted OR was 1.49 [1.24–1.78]. This finding persisted in a 1:1 propensity score matched cohort of 1835 patients pairs with an OR of 1.60 [1.28–2.00]. Other factors independently associated with SSI were an ASA score > 2, a longer duration of operation, and a reoperation for a non-infectious complication. Protective factors were female sex, older age, antibiotic prophylaxis received 60 to 30 min before surgery, elective operation, and endoscopic approach.ConclusionsDespite its activity against enterococci, A/CA was less effective than C + M for preventing SSI, suggesting that it should not be a first choice antibiotic prophylaxis for colorectal surgery.

National validation of the Centers for Medicare & Medicaid Services strategy for identifying potential surgical-site infections following colon surgery and abdominal hysterectomy.

National validation of claims-based surveillance for surgical-site infections (SSIs) following colon surgery and abdominal hysterectomy. Retrospective cohort study. US hospitals selected for data validation by Centers for Medicare & Medicaid Services (CMS). The study included 550 hospitals performing colon surgery and 458 hospitals performing abdominal hysterectomy in federal fiscal year 2013. We requested 1,200 medical records from hospitals selected for validation as part of the CMS Hospital Inpatient Quality Reporting program. For colon surgery, we sampled 60% with a billing code suggestive of SSI during their index admission and/or readmission within 30 days and 40% who were readmitted without one of these codes. For abdominal hysterectomy, we included all patients with an SSI code during their index admission, all patients readmitted within 30 days, and a sample of those with a prolonged surgical admission (length of stay > 7 days). We calculated sensitivity and positive predictive value for the different groups. We identified 142 colon-surgery SSIs (46 superficial SSIs and 96 deep and organ-space SSIs) and 127 abdominal-hysterectomy SSIs (58 superficial SSIs and 69 deep and organ-space SSIs). Extrapolating to the full CMS data validation cohort, we estimated an SSI rate of 8.3% for colon surgery and 3.0% for abdominal hysterectomy. Our colon-surgery surveillance codes identified 93% of SSIs, with 1 SSI identified for every 2.6 patients reviewed. Our abdominal-hysterectomy surveillance codes identified 73% of SSIs, with 1 SSI identified for every 1.6 patients reviewed. Using claims to target record review for SSI validation performed well in a national sample.

Mortality and Case Fatality Rates Associated With Surgical Site Infections: A Retrospective Surveillance Study.

Background: Surgical site infections (SSIs) have been associated with increases in terms of costs, hospital stay, morbidity, and mortality. We aimed to assess trends in SSIs monitored through 10 years of surveillance activities in our region, and to describe mortality attributable to SSIs in the two most frequently monitored surgical procedures: colorectal surgery and hip arthroplasty. Methods: A retrospective cohort study was conducted among the 42 hospitals participating in the surveillance network of our region in northern Italy. All colorectal and hip arthroplasty procedures performed between January 1st, 2010, and December 31st, 2019, and monitored through the surveillance system were included in the study. Surgical site infection rates, overall mortality, case fatality rates (CFR), and mortality attributable to SSIs were evaluated overall and by year of participation in the surveillance program. Results: In total, 11,417 colon surgery and 20,804 hip arthroplasty procedures were included. Among colon surgery procedures, SSI rates decreased from 9.21% in 2010 to 5.7% in 2019. A significant decreasing trend was found for overall mortality (p = 0.008), which progressively decreased from 4.96% in 2010 to 2.96% in 2019. Among hip arthroplasty procedures, no significant trend emerged for SSI and mortality rates. Considering the 10-year period, the CFR was 6.62% and 3.7% for SSIs after colon surgery and hip arthroplasty procedures, respectively. Conclusions: The impact of SSIs on the clinical outcomes of patients undergoing surgery highlights the importance of SSI surveillance.

Development of machine learning models for the detection of surgical site infections following total hip and knee arthroplasty: a multicenter cohort study

BackgroundPopulation based surveillance of surgical site infections (SSIs) requires precise case-finding strategies. We sought to develop and validate machine learning models to automate the process of complex (deep incisional/organ space) SSIs case detection.MethodsThis retrospective cohort study included adult patients (age ≥ 18 years) admitted to Calgary, Canada acute care hospitals who underwent primary total elective hip (THA) or knee (TKA) arthroplasty between Jan 1st, 2013 and Aug 31st, 2020. True SSI conditions were judged by the Alberta Health Services Infection Prevention and Control (IPC) program staff. Using the IPC cases as labels, we developed and validated nine XGBoost models to identify deep incisional SSIs, organ space SSIs and complex SSIs using administrative data, electronic medical records (EMR) free text data, and both. The performance of machine learning models was assessed by sensitivity, specificity, positive predictive value, negative predictive value, F1 score, the area under the receiver operating characteristic curve (ROC AUC) and the area under the precision–recall curve (PR AUC). In addition, a bootstrap 95% confidence interval (95% CI) was calculated.ResultsThere were 22,059 unique patients with 27,360 hospital admissions resulting in 88,351 days of hospital stay. This included 16,561 (60.5%) TKA and 10,799 (39.5%) THA procedures. There were 235 ascertained SSIs. Of them, 77 (32.8%) were superficial incisional SSIs, 57 (24.3%) were deep incisional SSIs, and 101 (42.9%) were organ space SSIs. The incidence rates were 0.37 for superficial incisional SSIs, 0.21 for deep incisional SSIs, 0.37 for organ space and 0.58 for complex SSIs per 100 surgical procedures, respectively. The optimal XGBoost models using administrative data and text data combined achieved a ROC AUC of 0.906 (95% CI 0.835–0.978), PR AUC of 0.637 (95% CI 0.528–0.746), and F1 score of 0.79 (0.67–0.90).ConclusionsOur findings suggest machine learning models derived from administrative data and EMR text data achieved high performance and can be used to automate the detection of complex SSIs.

Evaluation of a client questionnaire at diagnosing surgical site infections in an active surveillance system.

To report sensitivity, specificity, predictive values and accuracy of a client questionnaire at diagnosing surgical site infections (SSIs) and describe the impact of active surveillance on SSI detection. Prospective, cohort study. Dogs and cats undergoing soft tissue or orthopedic surgery over a 12-month period at a referral hospital. Clients were emailed a questionnaire 30 days postoperatively, or 90 days where an implant was used. Three algorithms were developed to diagnose SSIs using one or both of two criteria: (1) presence of any wound healing problems; (2) wound dehiscence or antibiotic prescription, and either purulent discharge or two or more clinical signs (redness, pain, heat, swelling, discharge). Algorithmic diagnoses were compared to gold standard diagnoses made by veterinarians. Of 754 surgical procedures, 309 responses were completed with 173 corresponding gold standard diagnoses. The most accurate algorithm determined "SSI" or "No SSI" from 90.2% of responses with 95.5% (92.4-98.6) accuracy, 82.6% (77-88.3) sensitivity, 97.7% (95.5-100) specificity, 86.4% (81.2-91.5) positive predictive value, and 97% (94.5-99.6) negative predictive value. "No SSI" was diagnosed in responses not meeting criterion 1, and "SSI" in responses meeting criteria 1 and 2. "Inconclusive" responses, comprising 9.8% of responses, met criterion 1 but not 2. Overall SSI rate was 62/754 (8.2%) and 12/62 (19.4%) SSIs were detected by active surveillance only. Use of this client questionnaire accurately diagnosed SSIs; active surveillance increased SSI detection. Surveillance of SSIs should be active and can be simplified by using a client questionnaire and algorithmic diagnoses, allowing automated distribution, data collection and analysis.

Digital tools for post-discharge surveillance of surgical site infection.

Conduct a scoping review on the development and use of digital tools for post-discharge surgical site infection surveillance. Scoping review. Science Direct, PubMed, Embase, Literatura Latino-Americana e do Caribe em Ciências da Saúde and Cumulative Index to Nursing and Allied Health Literature were searched from 2013 to May 2022. Six intellectual property registries were reviewed from 2013 to 2022. The review followed the Joanna Briggs Institute model, and included intellectual property records (applications, prototypes and software) and scientific articles published in any language on the development and/or testing of digital tools for post-discharge surveillance of surgical site infection among surgical patients aged 18 and over. One intellectual property record and 13 scientific articles were identified, covering 10 digital tools. The intellectual property record was developed and registered by a China educational institution in 2018. The majority of manuscripts were prospective cohort studies and randomized clinical trials, published between 2016 and 2022, and more than half were conducted in the United States. The population included adult patients undergoing cardiac, thoracic, vascular, abdominal, arthroplasty and caesarean surgery. The main functionalities of the digital tools were the previously prepared questionnaire, the attachment of a wound image, the integrated Web system and the evaluation of data by the health team, with post-discharge surgical site infection surveillance time between 14 and 30 days after surgery. Digital tools show promise for the surveillance of surgical site infection, collaborating with the early detection of wound infection. Mobile technology was favourable for detecting surgical site infections, reducing unnecessary visits to the health service, and increasing patient satisfaction. Technological advances in the health area open new perspectives for post-discharge surveillance of surgical site infection. WHAT IS ALREADY KNOWN?: There is underreporting of surgical site infections due to difficulties related to traditional methods of post-discharge surveillance. The use of digital tools within surgical site infection surveillance is increasing. Benefits of using digital tools within surgical site infection surveillance have been reported. WHAT HAS THIS STUDY ADDED TO OUR KNOWLEDGE?: This scoping review is one of the first to analyse the development and use of digital tools for post-discharge surveillance of surgical site infection in different countries. The main functionalities of digital tools are: structured questionnaires; attachment of wound images; integrated web systems; and evaluation of data by professionals. The use of mobile technology is favourable for detecting surgical site infections with a reduction in costs from face-to-face consultations and increased patient satisfaction. WHERE AND ON WHOM WILL THE RESEARCH HAVE AN IMPACT?: Healthcare providers can successfully use digital tools for surgical site infection post-discharge surveillance. Remote monitoring can reduce unnecessary patient visits to healthcare facilities. Policy makers can study how to implement digital platforms for remote patient monitoring. PRISMA statement for Scoping Reviews (PRISMA-ScR). No patient or public contribution. The study protocol was registered in the OSF (https://doi.org/10.17605/OSF.IO/BA8D6).

Compliance with surgical site infection prevention measures in hospitals

Objective: To assess compliance with actions for surgical site infection prevention and control, adopted in clinical practice in accordance with the recommendations proposed by the World Health Organization: auditing the moment of antibiotic administration, trichotomy with an electric clipper/disposable blade, material sterility confirmation, and surveillance and dissemination of surgical site infection rates. Methods: This is a cross-sectional observational study, carried out in 30 hospitals in the state of Minas Gerais, from February 2018 to April 2019, based on an audit of a surgical procedure, situational diagnosis and interview with Hospital Infection Control Services. Results In 93.3% of Hospital Infection Control Services, protocols for prophylactic antibiotic use and compliance audits were reported, 69% reported trichotomy with an electric clipper. All carried out surgical site infection surveillance, however, only 63.3% carried out rate disclosure. In the situational diagnosis, trichotomy was performed in 76.7% inside the operating room with an electric clipper (56.7%). In the audit of the moment of antimicrobial administration between 30-60 minutes before surgical incision, compliance was identified in 63.3%; 93.3% of services confirmed material sterility through process indicators. Conclusion: Differences were found between recommendations from the Hospital Infection Control Service and compliance with good practices during situational diagnosis and auditing of surgical procedures, reinforcing the need for training and audits aimed at effectively complying with professionals’ practices regarding such measures.

Comorbidities directly extracted from the hospital database for adjusting SSI risk in the new national semiautomated surveillance system in France: The SPICMI network.

To evaluate the performance of a comorbidity-based risk-adjustment model for surgical-site infection (SSI) reporting and benchmarking using a panel of variables extracted from the hospital discharge database (HDD), including comorbidities, compared to other models that use variables from different data sources. The French national surveillance program for SSI (SPICMI) has collected data from voluntary hospitals in the first 6 months of 2020 and 2021, for 16 selected surgery procedures, using a semiautomated algorithm for detection. Four risk-adjustment models were selected with logistic regression analysis, combining the different patterns of variables: National Nosocomial Infections Surveillance System (NNIS) risk-index components, individual operative data, and 6 individual comorbidities according to International Classification of Disease, Tenth Revision (ICD-10) diagnosis: obesity, diabetes, malnutrition, hypertension, cancer, or immunosuppression. Areas under the curve (AUCs) were calculated and compared. Overall, 294 SSI were detected among 11,975 procedures included. All 6 comorbidities were related to SSI in the univariate analysis. The AUC of the selected model including comorbidities (0.675; 95% confidence interval [CI], 0.642-0.707), was significantly higher than the AUC of the model without comorbidities (0.641; 95% CI, 0.609-0.672; P = .016) or the AUC using the NNIS-index components (0.598; 95% CI, 0.564-0.630; P < .001). The HDD-based model AUC (0.659; 95% CI, 0.625-0.692) did not differ significantly from the selected model without comorbidities (P = .23). Including HDD-based comorbidities as patient case-mix variables instead of NNIS risk index factors could be an effective approach for risk-adjustment of automated SSI surveillance more widely accessible to hospitals.