Establishing semi-automated infection surveillance in obstetrics and gynaecology

BACKGROUND

Surveillance is an acknowledged method to decrease nosocomial infections, such as surgical site infections (SSIs). Electronic healthcare records create the opportunity for automated surveillance. While approaches for different types of surgeries and indicators already exist, there are very few for obstetrics and gynaecology.

AIM

To analyse the sensitivity and workload reduction of semi-automated surveillance in obstetrics and gynaecology.

METHODS

In this retrospective, single-centre study at a 1438-bed tertiary care hospital in Germany, semi-automated SSI surveillance using the indicators 'antibiotic prescription', 'microbiological data' and 'administrative data' (diagnosis codes, readmission, post-hospitalization care) was compared with manual analysis and categorization of all patient files. Breast surgeries (BSs) conducted in 2018 and caesarean sections (CSs) that met the inclusion criteria between May 2013 and December 2019 were included. Indicators were analysed for sensitivity, number of analysed procedures needed to identify one case, and potential workload reduction in detecting SSIs in comparison with the control group.

FINDINGS

The reference standard showed nine SSIs in 416 BSs (2.2%). Sensitivities for the indicators 'antibiotic prescription', 'diagnosis code', 'microbiological sample taken', and the combination 'diagnosis code or microbiological sample' were 100%, 88.9%, 66.7% and 100%, respectively. The reference standard showed 54 SSIs in 3438 CSs (1.6%). Sensitivities for the indicators 'collection of microbiological samples', 'diagnosis codes', 'readmission/post-hospitalization care', and the combination of all indicators were 38.9%, 27.8%, 85.2% and 94.4%, respectively.

CONCLUSIONS

Semi-automated surveillance systems may reduce workload by maintaining high sensitivity depending on the type of surgery, local circumstances and thorough digitalization.

Using artificial intelligence to reduce orthopedic surgical site infection surveillance workload: Algorithm design, validation, and implementation in 4 Spanish hospitals

BACKGROUND

Surgical site infection (SSI) surveillance is a labor-intensive endeavor. We present the design and validation of an algorithm for SSI detection after hip replacement surgery, and a report of its successful implementation in 4 public hospitals in Madrid, Spain.

METHODS

We designed a multivariable algorithm, AI-HPRO, using natural language processing (NLP) and extreme gradient boosting to screen for SSI in patients undergoing hip replacement surgery. The development and validation cohorts included data from 19,661 health care episodes from 4 hospitals in Madrid, Spain.

RESULTS

Positive microbiological cultures, the text variable "infection", and prescription of clindamycin were strong markers of SSI. Statistical analysis of the final model indicated high sensitivity (99.18%) and specificity (91.01%) with an F1-score of 0.32, AUC of 0.989, accuracy of 91.27%, and negative predictive value of 99.98%.

DISCUSSION

Implementation of the AI-HPRO algorithm reduced the surveillance time from 975 person/hours to 63.5 person/hours and permitted an 88.95% reduction in the total volume of clinical records to be reviewed manually. The model presents a higher negative predictive value (99.98%) than algorithms relying on NLP alone (94%) or NLP and logistic regression (97%).

CONCLUSIONS

This is the first report of an algorithm combining NLP and extreme gradient-boosting to permit accurate, real-time orthopedic SSI surveillance.

Tranexamic acid is not associated with decreased infection risk after primary shoulder arthroplasty: a cohort study of 9276 patients

BACKGROUND

The use of tranexamic acid (TXA) is associated with less blood loss and reduced transfusion risk after shoulder arthroplasty surgery. Recent studies have shown lower odds of postoperative infection after hip or knee arthroplasty with its use. The purpose of this study was to determine whether TXA use reduces the risk of infection after primary elective shoulder arthroplasty.

METHODS

A retrospective cohort study was conducted using data from a US integrated health care system's shoulder arthroplasty registry. Patients aged 18 years who underwent primary elective anatomic total shoulder arthroplasty for osteoarthritis or primary reverse shoulder arthroplasty for rotator cuff arthropathy were included (2013-2020). We compared patients who received preoperative intravenous TXA to those who did not receive TXA by assessing the risk for revision due to deep infection within 5 years' follow-up using multivariable Cox proportional hazard regression. Interaction between TXA and diabetes status was analyzed separately.

RESULTS

The study sample included 9276 shoulder arthroplasties performed by 153 surgeons at 43 hospitals. The mean age was 70.0 years and 48% were male. The 5-year probability of revision for deep infection was 0.8% and 0.7% for patients with and without TXA, respectively. We failed to observe a difference in infection risk after adjustment for confounders and surgeon differences (hazard ratio [HR] 1.00, 95% confidence interval [CI] 0.56-1.80, P = .998). Further, no differences were observed in patients with (HR 1.64, 95% CI 0.42-6.44, P = .481) or without diabetes (HR 0.79, 95% CI 0.40-1.55, P = .488).

CONCLUSION

In a multicenter cohort of more than 9000 primary shoulder arthroplasty procedures, the use of preoperative TXA was not associated with a decrease in the 5-year probability of revision for deep infection.

Can we automate spine fusion surgical site infection data capture?

PURPOSE

Spine fusion surgical site infection (SSI) rate is reported to national quality databases and used as a benchmark for orthopedic departments and hospital systems. However, accurate data require resource-heavy administrative review and even this has shown to vary. We aimed to create a passive electronic medical record (EMR) algorithm to automatically capture spine fusion SSI and compared its accuracy against the administrative chart review and self-reported morbidity and mortality (M&M) rates.

METHODS

We retrospectively reviewed a single institution's spine fusion records for 7 years for all 90-day post-operative SSIs. We used Centers for Disease Control and Prevention (CDC) SSI definition coupled with intention to treat as an infection by orthopedics/infectious disease service as the gold standard. We compared our gold standard to administrative hand-checked SSI data, anonymously reported departmental M&M, and a passive EMR algorithm (ICD-9 or -10 post-operative SSI diagnosis code entered, or all four of: positive culture, antibiotic prescription between 3-90 days post-op, re-operation/re-admission, and a qualifying diagnosis).

RESULTS

Nine hundred and fourteen spine fusions were included, with a 2.8% SSI rate (0.9% superficial and 2.0% deep). Passive EMR algorithm was the most sensitive at 89% (vs 76% administrative review, 73% M&M); all were highly specific at 99-100%. M&M was 100% positively predictive, administrative review 95%, and EMR 79%.

CONCLUSION

Our passive EMR algorithm was more sensitive to pediatric spine fusion 90-day SSI than self-reported M&M and hand-checked administrative chart review. Although EMR may over-report, it can be used by others to narrow the initial sample for review, reduce resource burden involved with administrative spine SSI review, and provide a quality check for M&M self-reporting.

LEVEL OF EVIDENCE

III.

Temporal Trends in Deep Surgical Site Infections After Six Orthopaedic Procedures Over a 12-year Period Within a US-based Healthcare System

INTRODUCTION

Centers of excellence and bundled payment models have driven perioperative optimization and surgical site infection (SSI) prevention with decolonization protocols and antibiotic prophylaxis strategies. We sought to evaluate time trends in the incidence of deep SSI and its causative organisms after six orthopaedic procedures in a US-based integrated healthcare system.

METHODS

We conducted a population-level time-trend study using data from Kaiser Permanente's orthopaedic registries. All patients who underwent primary anterior cruciate ligament reconstruction (ACLR), total knee arthroplasty (TKA), elective total hip arthroplasty (THA), hip fracture repair, shoulder arthroplasty, and spine surgery were identified (2009 to 2020). The annual incidence of 90-day deep SSI was identified according to the National Healthcare Safety Network/Centers for Disease Control and Prevention guidelines with manual chart validation for identified infections. Poisson regression was used to evaluate annual trends in SSI incidence with surgical year as the exposure of interest. Annual trends in overall incidence and organism-specific incidence were considered.

RESULTS

The final study sample was composed of 465,797 primary orthopaedic procedures. Over the 12-year study period, a decreasing trend in deep SSI was observed for ACLR and hip fracture repair. Although there was variation in incidence rates for specific operative years for TKA, elective THA, shoulder arthroplasty, and spine surgery, no consistent decreasing trends over time were found. Decreasing rates of Staphylococcus aureus infections over time after hip fracture repair, shoulder arthroplasty, and spine surgery and decreasing trends in antibiotic resistance after elective THA and spine surgery were also observed. Increasing trends of polymicrobial infections were observed after TKA and Cutibacterium acnes after elective THA.

CONCLUSIONS

The overall incidence of deep SSI after six orthopaedic procedures was rare. Decreasing SSI rates were observed for ACLR and hip fracture repair within our US-based healthcare system. Polymicrobial infections after TKA and Cutibacterium acnes after elective THA warrant closer surveillance.

LEVEL OF EVIDENCE

IV.

Differing Microorganism Profile in Early and Late Prosthetic Joint Infections Following Primary Total Knee Arthroplasty - Implications for Empiric Antibiotic Treatment

BACKGROUND

Prosthetic joint infection (PJI) is the leading cause of revision following total knee arthroplasty (TKA). Prior to microorganism identification, the choice of the correct empiric antibiotics is critical to treatment success. This study aims to 1) compare the microorganism and resistance profile in early and late PJIs; 2) recommend appropriate empiric antibiotics.

METHODS

A multicentre retrospective review was performed over a 15-year period. First episode PJIs were classified by both the Tsukayama Classification and Auckland Classification. For each PJI case, the causative organism and antibiotic sensitivity were recorded.

RESULTS

Of eligible patients, 232 culture-positive PJI cases were included. Using either classification system, early PJIs (<4 weeks or <1 year since primary) were significantly more likely to be resistant and polymicrobial. The predominant organisms were coagulase-negative Staphylococci in early PJIs while Staphylococcus aureus was the most common in late PJIs. The distribution of gram-negative cases was higher in early Class-A than late Class-C PJIs (25% versus 6%, P = .004). Vancomycin provided significantly superior coverage when compared to Flucloxacillin for early infections, and addition of a gram-negative agent achieved coverage over 90% using both classification systems.

CONCLUSION

Based on the microbiological pattern in Tsukayama criteria, Vancomycin with the consideration of Gram-negative agent should be considered for Class-A infections given the high proportion of resistant and polymicrobial cases. For Class-C infections, Cephazolin or Flucloxacillin is likely sufficient. We recommend antibiotics to be withheld in Class-B infections until cultures and sensitivities are known.

Utilizing technology to increase efficiency of infection prevention data collection: Our experience using electronic medical records for symptom surveillance

Utilizing technology can increase efficiency for infection prevention data collection. The purpose of this article is to explain our experience using electronic medical record (EMR) reports for symptom surveillance to increase understanding of how technology enhances infection prevention abilities. The new EMR report method took on average just over 8 minutes a week to run compared to the traditional manual method which took 7 hours a week indicating a major time decrease for data collection.

Surgical Site Infections

Surgical site infections (SSIs) are among the most common and most costly health care-associated infections, leading to adverse patient outcomes and death. Wound contamination occurs with each incision, but proven strategies exist to decrease the risk of SSI. In particular, improved adherence to evidence-based preventive measures related to appropriate antimicrobial prophylaxis can decrease the rate of SSI. Aggressive surgical debridement and effective antimicrobial therapy are needed to optimize the treatment of SSI.

Large increase of vertebral osteomyelitis in France: a 2010-2019 cross-sectional study

Vertebral osteomyelitis (VO) represents 4-10% of bone and joint infections. In Western countries, its incidence seems to increase, simultaneously with an increasing number of comorbidities among an ageing population. This study aimed to assess the evolution of VO epidemiology in France over the 2010-2019 decade. A nationwide cross-sectional study was conducted using the French hospital discharge data collected through the French diagnosis-related groups 'Programme de Médicalisation des Systèmes d'Information'. VOs were detected with a previously validated case definition using International Classification of Diseases 10 (ICD-10) codes, implemented with the French current procedural terminology codes. The study population included all patients hospitalised in France during the 2010-2019 decade, aged 15 years old and more. Patient and hospital stay characteristics and their evolutions were described. During the study period, 42 105 patients were hospitalised for VO in France involving 60 878 hospital stays. The mean VO incidence was 7.8/100 000 over the study period, increasing from 6.1/100 000 in 2010 to 11.3/100 000 in 2019. The mean age was 64.8 years old and the sex ratio was 1.56. There were 31 341 (74.4%) patients with at least one comorbidity and 3059 (7.3%) deceased during their hospital stay. Even if rare, device-associated VOs (4450 hospital stays, 7.3%) highly increased over the period. The reliability of the method, based upon an exhaustive database and a validated case definition, provided an effective tool to compare data over time in real-life conditions to regularly update the epidemiology of VO.

Diagnostic accuracy of electronic surveillance tool for catheter-associated urinary tract infections in tertiary care hospitals: A meta-analysis

BACKGROUND

Automated systems have been developed to reduce labor-intensive manual recordings during nosocomial infection surveillance. The diagnostic accuracies of these systems have differed in various settings.

METHODS

We designed this meta-analysis to evaluate the diagnostic accuracy of an electronic surveillance tool for catheter-associated urinary tract infections (CAUTIs) in tertiary care hospitals. We systematically searched databases such as Medline, Scopus, Cochrane library and Embase (from inception until November 2019) for relevant studies. We assessed the quality of trials using the diagnostic accuracy studies-2 tool, and performed a meta-analysis to obtain a pooled sensitivity and specificity for electronic surveillance. We included 6 studies with 16,492 patients in the analysis.

RESULTS

We found a pooled sensitivity of electronic diagnostic surveillance for CAUTIs of 97.5% (95% confidence interval [CI], 67.6-99.9%) and a pooled specificity of 92.6% (95% CI, 55.2-99.2%). The diagnostic odds ratio was 494 (95% CI, 89-2747). The positive likelihood ratio was 13.1 (95% CI, 1.63-105.8) and the negative likelihood ratio 0.02 (95% CI, 0.001-0.40). A bivariate box plot indicated the possibility of heterogeneity between the included studies.

CONCLUSION

Our review suggests that electronic surveillance is useful for diagnosing CAUTIs among hospitalized patients in tertiary care hospitals due to its high sensitivity and specificity.

Role of real-world digital data for orthopedic implant automated surveillance: a systematic review

INTRODUCTION

Data collection automation through the reuse of real-world digital data from clinical data warehouses (CDW) could represent a great opportunity to improve medical device monitoring. For instance, this approach is starting to be used for the design of automated decision support systems for joint replacement monitoring. However, a number of obstacles remains, such as data quality and interoperability through the use of common and regularly updated terminologies, and the use of a Unique Device Identifier (UDI).

AREAS COVERED

To present the existing models of automated surveillance of orthopedic devices, a systematic review of initiatives using real-world digital health data to monitor joint replacement surgery was performed following the PRISMA 2020 guidelines. The main objective was to identify the data sources, the target populations, the population size, the device location, and the main results of studies on such initiatives.

EXPERT OPINION

Analysis of the identified studies showed that real-world digital data offer many opportunities for improving the automation of monitoring in orthopedics. The contribution of real-world data, especially through natural language processing, UDI use in CDW and the integration of device databases, is needed for automated and more robust health surveillance.

Success Rates of Debridement, Antibiotics, and Implant Retention in 230 Infected Total Knee Arthroplasties: Implications for Classification of Periprosthetic Joint Infection

BACKGROUND

Prosthetic joint infection (PJI) is the most common cause of failure following total knee arthroplasty (TKA). This study aimed to determine the success of debridement, antibiotics, and implant retention (DAIR) in a large cohort of TKA PJIs and assess the utility of current classification systems in predicting DAIR outcomes in early postoperative, late hematogenous, and chronic PJIs.

METHODS

In a multicenter review over 15 years, 230 patients underwent DAIR for first episode PJI following primary TKA. Patient demographics, disease and surgical factors, treatment regime, and outcomes were identified. Univariate and multivariate survival analyses were performed to identify factors associated with successful DAIR. Continuous variables with predictive value were further analyzed using receiver operating characteristic curves. The ability to predict DAIR outcomes of multiple classification systems was also assessed.

RESULTS

Patients were followed for an average of 6.9 years. The overall success rate of DAIR was 53.9%. On receiver operating characteristic analysis, 3 months (area under the curve = 0.63) and 1-year age (area under the curve = 0.66) of implant cut-offs was similarly predictive of outcomes. On multivariate survival analysis, DAIR was successful in 64% of "early" PJIs (implant <1 year) vs 38% of "late hematogenous" PJIs (implant >1 year; odds ratio [OR] 1.78, P = .01). For late PJIs (implant >1 year), Staphylococcus aureus (OR 4.70, P < .001) and gram-negative infections (OR 2.56, P = .031) were risk factors for DAIR failure.

CONCLUSION

DAIR has a high failure rate in all PJIs occurring more than a year post primary TKA, particularly when caused by S aureus or gram-negative bacteria. The age of implant is an important predictor of DAIR outcomes.

Commercially Prepared Antibiotic-Loaded Bone Cement and Infection Risk Following Cemented Primary Total Knee Arthroplasty

BACKGROUND

The efficacy of commercially available antibiotic-loaded bone cement (ABC) in preventing infection in total knee arthroplasty (TKA) is unclear. We sought to determine the effectiveness of commercially available ABC in reducing the risk of infection following TKA, both overall and among 3 subgroups of patients with a higher risk of infection (diabetes, body mass index ≥35 kg/m, and American Society of Anesthesiologists classification ≥3), and to evaluate the association between the use of ABC and the risks of aseptic revision and revision for aseptic loosening.

METHODS

The Kaiser Permanente Total Joint Replacement Registry was utilized to evaluate 87,018 primary cemented TKAs performed from 2008 to 2016. The primary outcome was time to infection (90-day deep infection or septic revision). Reduced infection risk with ABC relative to regular cement was tested with use of propensity-score-weighted Cox proportional-hazards models with superiority and noninferiority testing. All analyses were replicated for each of the 3 high-risk subgroups. For the secondary revision outcomes, propensity-score-weighted Cox proportional-hazards models were utilized.

RESULTS

Regular cement was found to be noninferior to ABC with respect to risk infection (hazard ratio [HR], 1.14; 95% confidence interval [CI], 0.93 to 1.40) and cost across all TKA patients. However, a lower risk of infection was observed with ABC among TKA patients with diabetes (HR, 0.72; 95% CI, 0.52 to 0.99). There was no evidence of a difference in risk of revision for ABC compared with regular cement.

CONCLUSIONS

We found that the additional cost associated with the use of commercially prepared ABC in primary TKA was not justified in all patients; however, the risk of reduction was lower among patients with diabetes who received ABC. Further study is warranted to identify the efficacy of ABC among other high-risk populations.

LEVEL OF EVIDENCE

Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.

Comparative study on mid- and long-term clinical effects of medial pivot prosthesis and posterior-stabilized prosthesis after total knee arthroplasty

Objective

The purpose of this study was to explore the mid-and long-term clinical effects of Chinese patients with medial pivot (MP) prosthesis and posterior-stabilized (PS) prosthesis after total knee arthroplasty (TKA), to provide a reference for the recommendation of clinical prostheses.

Methods

A retrospective analysis of 802 patients who received TKA was performed from June 2010 to December 2013. A total of 432 patients received a MP prosthesis (MP group) and 375 patients received a PS prosthesis (PS group). Postoperative range of motion (ROM), clinical scores including the knee scoring system (KSS), the Western Ontario and McMaster Universities osteoarthritis index (WOMAC), the forgotten joint score (FJS), and postoperative complications were compared between the two groups.

Results

A total of 527 patients were followed up, including 290 in the MP group and 237 in the PS group. Both groups achieved satisfactory results in terms of KSS score, WOMAC score, and postoperative ROM, which were significantly improved compared with those before surgery, but the difference between the groups was not statistically significant (P > 0.05). The FJS scores of the MP group and the PS group were satisfactory and no significant difference was observed (P = 0.426). Postoperative complications occurred in 5 and 11 patients in the MP group and PS group, respectively.

Conclusion

The clinical results of TKA with MP or PS in Chinese patients at mid- and long-term are encouraging, and no significant differences were observed between the two types of prostheses. Studies have also shown that both prostheses are safe for Chinese patients.

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 an algorithm for semiautomated surveillance to detect deep surgical site infections after primary total hip or knee arthroplasty-A multicenter study

OBJECTIVE

Surveillance of healthcare-associated infections is often performed by manual chart review. Semiautomated surveillance may substantially reduce workload and subjective data interpretation. We assessed the validity of a previously published algorithm for semiautomated surveillance of deep surgical site infections (SSIs) after total hip arthroplasty (THA) or total knee arthroplasty (TKA) in Dutch hospitals. In addition, we explored the ability of a hospital to automatically select the patients under surveillance.

DESIGN

Multicenter retrospective cohort study.

METHODS

Hospitals identified patients who underwent THA or TKA either by procedure codes or by conventional surveillance. For these patients, routine care data regarding microbiology results, antibiotics, (re)admissions, and surgeries within 120 days following THA or TKA were extracted from electronic health records. Patient selection was compared with conventional surveillance and patients were retrospectively classified as low or high probability of having developed deep SSI by the algorithm. Sensitivity, positive predictive value (PPV), and workload reduction were calculated and compared to conventional surveillance.

RESULTS

Of 9,554 extracted THA and TKA surgeries, 1,175 (12.3%) were revisions, and 8,378 primary surgeries remained for algorithm validation (95 deep SSIs, 1.1%). Sensitivity ranged from 93.6% to 100% and PPV ranged from 55.8% to 72.2%. Workload was reduced by ≥98%. Also, 2 SSIs (2.1%) missed by the algorithm were explained by flaws in data selection.

CONCLUSIONS

This algorithm reliably detects patients with a high probability of having developed deep SSI after THA or TKA in Dutch hospitals. Our results provide essential information for successful implementation of semiautomated surveillance for deep SSIs after THA or TKA.

Perioperative Clindamycin Use in Penicillin Allergic Patients Is Associated With a Higher Risk of Infection After Shoulder Arthroplasty

BACKGROUND

This study determines whether infection rates differ between prophylactic antibiotic use for patients with or without penicillin allergy before shoulder arthroplasty surgery.

METHODS

Seven thousand one hundred forty primary shoulder arthroplasties operated between 2005 and 2016 were identified. We compared deep surgical site infection risk of patients who received perioperative vancomycin alone (6.2%, N = 444) or clindamycin alone (7.1%, N = 508) for penicillin allergy versus patients who received cefazolin alone without penicillin allergy (86.7%, N = 6,188).

RESULTS

Seventy deep infections (1.2% 5-year cumulative incidence) were observed. The most common organism was Cutibacterium acnes (39.4%, N = 27). Compared with patients treated with cefazolin, infection risk was not different for those treated with vancomycin (hazard ratio = 1.17, 95% confidence interval 0.42 to 3.30, P = 0.8), but a higher risk of infection was identified for those treated with clindamycin alone (hazard ratio = 3.45, 95% confidence interval 1.84 to 6.47, P < 0.001).

CONCLUSION

A higher risk of postoperative infection is found after prophylactic use of intravenous clindamycin antibiotic after shoulder arthroplasty. Vancomycin is preferred over clindamycin for patients with penicillin allergy.

LEVEL OF EVIDENCE

III, retrospective cohort study.

Screw and Sheath Tibial Fixation Associated With a Higher Likelihood of Deep Infection After Hamstring Graft Anterior Cruciate Ligament Reconstruction

BACKGROUND

Hamstring autograft anterior cruciate ligament reconstructions (ACLRs) have exhibited higher infection rates compared with bone-patellar tendon-bone (BPTB) autograft. The reason for this observed difference is unclear, warranting investigation.

PURPOSE

To evaluate the association between tibial fixation, either with or without a sheath and screw construct, and the risk of deep infection after hamstring autograft ACLR, using BPTB autograft as a reference group for comparison.

STUDY DESIGN

Cohort study; Level of evidence, 3.

METHODS

Using the Kaiser Permanente ACLR Registry, we identified all primary isolated, unilateral, single-bundle ACLRs with a BPTB or hamstring autograft (January 1, 2008, to September 30, 2016). The exposure groups included the following: (1) BPTB ACLR, (2) hamstring ACLR using a screw and sheath construct for tibial fixation (HS with screw and sheath), and (3) hamstring ACLR using a method other than a screw and sheath construct for tibial fixation (HS without screw and sheath). We used logistic regression to evaluate the likelihood of 90-day postoperative deep infection using BPTB autograft as the reference group and adjusting for age, sex, and body mass index. The number needed to be exposed (NNE) was calculated.

RESULTS

Of 15,671 ACLRs, 6745 (43.0%) used a BPTB graft, 2852 (18.2%) used HS with screw and sheath tibial fixation, and 6074 (38.8%) used HS without screw and sheath tibial fixation. There were 38 (0.2%) 90-day deep infections: 11 (0.2%) for BPTB, 14 (0.5%) for HS with screw and sheath, and 13 (0.2%) for HS without screw and sheath. Staphylococcus aureus for the BPTB group and Staphylococcus epidermidis in both hamstring groups were the most common infecting organisms. HS with screw and sheath had a higher likelihood of 90-day deep infection compared with BPTB ACLR (odds ratio [OR], 2.87; 95% CI, 1.29-6.38). We failed to observe a difference for HS without screw and sheath compared with BPTB ACLR (OR, 1.23; 95% CI, 0.54-2.77). The NNE was 330 and 2701 for HS with and HS without screw and sheath, respectively.

CONCLUSION

Although the overall infection rate after ACLR is low, the higher likelihood of infections when sheath and screw combined are used for tibial fixation of a hamstring autograft ACLR should be a consideration when this procedure is performed.

Natural Language Processing for the Identification of Surgical Site Infections in Orthopaedics

BACKGROUND

The identification of surgical site infections for infection surveillance in hospitals depends on the manual abstraction of medical records and, for research purposes, depends mainly on the use of administrative or claims data. The objective of this study was to determine whether automating the abstraction process with natural language processing (NLP)-based models that analyze the free-text notes of the medical record can identify surgical site infections with predictive abilities that match the manual abstraction process and that surpass surgical site infection identification from administrative data.

METHODS

We used surgical site infection surveillance data compiled by the infection prevention team to identify surgical site infections among patients undergoing orthopaedic surgical procedures at a tertiary care academic medical center from 2011 to 2017. We compiled a list of keywords suggestive of surgical site infections, and we used NLP to identify occurrences of these keywords and their grammatical variants in the free-text notes of the medical record. The key outcome was a binary indicator of whether a surgical site infection occurred. We estimated 7 incremental multivariable logistic regression models using a combination of administrative and NLP-derived variables. We split the analytic cohort into training (80%) and testing data sets (20%), and we used a tenfold cross-validation approach. The main analytic cohort included 172 surgical site infection cases and 200 controls that were repeatedly and randomly selected from a pool of 1,407 controls.

RESULTS

For Model 1 (variables from administrative data only), the sensitivity was 68% and the positive predictive value was 70%; for Model 4 (with NLP 5-grams [distinct sequences of 5 contiguous words] from the medical record), the sensitivity was 97% and the positive predictive value was 97%; and for Model 7 (a combination of Models 1 and 4), the sensitivity was 97% and the positive predictive value was 97%. Thus, NLP-based models identified 97% of surgical site infections identified by manual abstraction with high precision and 43% more surgical site infections compared with models that used administrative data only.

CONCLUSIONS

Models that used NLP keywords achieved predictive abilities that were comparable with the manual abstraction process and were superior to models that used administrative data only. NLP has the potential to automate and aid accurate surgical site infection identification and, thus, play an important role in their prevention.

CLINICAL RELEVANCE

This study examines NLP's potential to automate the identification of surgical site infections. This automation can potentially aid the prevention and early identification of these surgical complications, thereby reducing their adverse clinical and economic impact.

Failed Debridement and Implant Retention Does Not Compromise the Success of Subsequent Staged Revision in Infected Total Knee Arthroplasty

BACKGROUND

Periprosthetic joint infection (PJI) is the leading cause of early revisions after total knee arthroplasty. Debridement, antibiotics, and implant retention (DAIR) procedures are often the initial treatment for PJI. However, there is concern that failed DAIR undermines the future success of revision procedures. This study aims to investigate the impact of DAIR on the success of subsequent staged revisions for PJI.

METHODS

A multicenter retrospective review was performed over a 15-year period. Treatment success was defined as implant retention without the use of long-term suppressive antibiotics. This was compared between patients who underwent a staged revision as the first procedure for PJI (staged-only) and patients who failed DAIR before staged revision (F-DAIR). Competing risk survival analysis was performed to compare the 2 groups and considered for patient demographics, American Society of Anesthesiologists score, organism type, body mass index, age of prosthesis, and duration of symptoms.

RESULTS

Of 291 eligible patients, 63 underwent staged revision and 228 underwent DAIR as the first procedure for PJI. Of the 228 DAIR patients, 75 failed DAIR and underwent subsequent staged revision (F-DAIR). At mean follow-up of 6.2 years, the success rate was 72% in the F-DAIR group and 81% in the staged-only group. On survival analysis, there was no significant difference in subdistribution hazard ratio comparing the probability of failure (implant retention) in the 2 treatments groups (subdistribution hazard ratio = 0.72; 95% confidence interval 0.32-1.61; P = .42).

CONCLUSION

This study suggested that a previously failed DAIR does not compromise the success rate of a subsequent staged revision.

Walk a Mile in the Leadership's Shoes: Why Focus on Quality Improvement?

Quality Improvement (QI) throughout health care in the United States continues to be of growing importance to both patients and providers. Leaders in health care including physicians, nurses, hospital administrators, and payors are all responsible for ensuring the continuation and growth of QI initiatives. This article will discuss various ways that healthcare leaders, with specific regard to orthopedic surgery, have utilized QI measures to provide better, more efficient, care to patients.

Validation of semiautomated surgical site infection surveillance using electronic screening algorithms in 38 surgery categories

OBJECTIVE

To verify the validity of a semiautomated surgical site infection (SSI) surveillance system using electronic screening algorithms in 38 categories of surgery.

DESIGN

A cohort study for validation of semiautomated SSI surveillance system using screening algorithms.

SETTING

A 1,989-bed tertiary-care referral center in Seoul, Republic of Korea.

METHODS

A dataset of 40,516 surgical procedures in 38 categories stored in the conventional SSI surveillance registry at the Samsung Medical Center between January 2013 and December 2014 was used as the reference standard. In the semiautomated surveillance system, electronic screening algorithms flagged cases meeting at least 1 of 3 criteria: antibiotic prescription, microbial culture, and infectious disease consultation. Flagged cases were audited by infection preventionists. Analyses of sensitivity, specificity, and positive predictive value (PPV) were conducted for the semiautomated surveillance system, and its effect on reducing the workload for chart review was evaluated.

RESULTS

A total of 575 SSI events (1·42%) were identified by conventional SSI surveillance. The sensitivity of the semiautomated SSI surveillance was 96·7%, and the PPV of the screening algorithms alone was 4·1%. Semiautomated SSI surveillance reduced the chart review workload of the infection preventionists from 1,283 to 482 person hours per year (a 62·4% decrease).

CONCLUSIONS

Compared to conventional surveillance, semiautomated surveillance using electronic screening algorithms followed by chart review of selected cases can provide high-validity surveillance results and can significantly reduce the workload of infection preventionists.

Validity and Reliability of Administrative Coded Data for the Identification of Hospital-Acquired Infections: An Updated Systematic Review with Meta-Analysis and Meta-Regression Analysis

OBJECTIVE

To conduct an updated assessment of the validity and reliability of administrative coded data (ACD) in identifying hospital-acquired infections (HAIs).

METHODS

We systematically searched three libraries for studies on ACD detecting HAIs compared to manual chart review. Meta-analyses were conducted for prosthetic and nonprosthetic surgical site infections (SSIs), Clostridium difficile infections (CDIs), ventilator-associated pneumonias/events (VAPs/VAEs) and non-VAPs/VAEs, catheter-associated urinary tract infections (CAUTIs), and central venous catheter-related bloodstream infections (CLABSIs). A random-effects meta-regression model was constructed.

RESULTS

Of 1,906 references found, we retrieved 38 documents, of which 33 provided meta-analyzable data (N = 567,826 patients). ACD identified HAI incidence with high specificity (≥93 percent), prosthetic SSIs with high sensitivity (95 percent), and both CDIs and nonprosthetic SSIs with moderate sensitivity (65 percent). ACD exhibited substantial agreement with traditional surveillance methods for CDI (κ = 0.70) and provided strong diagnostic odds ratios (DORs) for the identification of CDIs (DOR = 772.07) and SSIs (DOR = 78.20). ACD performance in identifying nosocomial pneumonia depended on the ICD coding system (DORICD-10/ICD-9-CM  = 0.05; p = .036). Algorithmic coding improved ACD's sensitivity for SSIs up to 22 percent. Overall, high heterogeneity was observed, without significant publication bias.

CONCLUSIONS

Administrative coded data may not be sufficiently accurate or reliable for the majority of HAIs. Still, subgrouping and algorithmic coding as tools for improving ACD validity deserve further investigation, specifically for prosthetic SSIs. Analyzing a potential lower discriminative ability of ICD-10 coding system is also a pending issue.

Key features of bone and joint infections following the implementation of reference centers in France

OBJECTIVES

French reference centers for bone and joint infections (BJI) were implemented from 2009 onwards to improve the management of complex BJIs. This study compared BJI burden before and after the implementation of these reference centers.

PATIENTS AND METHODS

BJI hospital stays were selected from the 2008 and 2013 national hospital discharge database using a validated algorithm, adding the new complex BJI code created in 2011. Epidemiology and economic burden were assessed.

RESULTS

BJI prevalence increased in 2013 (70 vs. 54/100,000 in 2008). Characteristics of BJI remained similar between 2008 and 2013: septic arthritis (50%), increasing prevalence with age and sex, case fatality 5%, mean length of stay 17.5 days, rehospitalization 20%. However, device-associated BJIs increased (34 vs. 26%) as well as costs (€421 million vs. €259 in 2008). Similar device-associated BJI characteristics between 2008 and 2013 were: septic arthritis (70%), case fatality (3%), but with more hospitalizations in reference centers (34 vs. 30%) and a higher cost per stay. Among the 7% of coded complex BJIs, the mean length of stay was 22.2 days and mean cost was €11,960.

CONCLUSIONS

BJI prevalence highly increased in France. Complex BJI prevalence assessment is complicated by the absence of clinical consensus and probable undercoding. A validation of clinical case definition of complex BJI is required.

Risk of Infection After Allograft Anterior Cruciate Ligament Reconstruction: Are Nonprocessed Allografts More Likely to Get Infected? A Cohort Study of Over 10,000 Allografts

BACKGROUND

Allograft tissue is frequently used in anterior cruciate ligament reconstruction (ACLR). It is often irradiated and/or chemically processed to decrease the risk of disease transmission, but some tissue is aseptically harvested without further processing. Irradiated and chemically processed allograft tissue appears to have a higher risk of revision, but whether this processing decreases the risk of infection is not clear.

PURPOSE

To determine the incidence of deep surgical site infection after ACLR with allograft in a large community-based sample and to evaluate the association of allograft processing and the risk of deep infection.

STUDY DESIGN

Cohort study; Level of evidence, 3.

METHODS

The authors conducted a cohort study using the Kaiser Permanente Anterior Cruciate Ligament Reconstruction Registry. Primary isolated unilateral ACLR with allograft were identified from February 1, 2005 to September 30, 2015. Ninety-day postoperative deep infections were identified via an electronic screening algorithm and then validated through chart review. Logistic regression was used to evaluate the likelihood of 90-day postoperative deep infection per allograft processing method: processed (graft treated chemically and/or irradiated) or nonprocessed (graft not irradiated or chemically processed).

RESULTS

Of 10,190 allograft cases, 8425 (82.7%) received a processed allograft, and 1765 (17.3%) received a nonprocessed allograft. There were 15 (0.15%) deep infections during the study period: 4 (26.7%) coagulase-negative Staphylococcus, 4 (26.7%) methicillin-sensitive Staphylococcus aureus, 1 (6.7%) Peptostreptococcus micros, and 6 (40.0%) with no growth. There was no difference in the likelihood for 90-day deep infection for processed versus nonprocessed allografts (odds ratio = 1.36, 95% CI = 0.31-6.04).

CONCLUSION

The overall incidence of deep infection after ACLR with allograft tissue was very low (0.15%), suggesting that the methods currently employed by tissue banks to minimize the risk of infection are effective. In this cohort, no difference in the likelihood of infection between processed and nonprocessed allografts could be identified.

Comparative Effectiveness and Safety of Drug Prophylaxis for Prevention of Venous Thromboembolism After Total Knee Arthroplasty

BACKGROUND

Rates of venous thromboembolism in contemporary studies of primary total knee arthroplasty (TKA) have been reported to be as high as 3.5%. Although drug prophylaxis is effective, the best option among these regimens is not well established. The purpose of this study was to evaluate the comparative effectiveness and safety of aspirin, low-molecular-weight heparin, synthetic pentasaccharide factor Xa inhibitors, and vitamin K antagonist.

METHODS

Data were from a US total joint replacement registry, with 30,499 patients receiving unilateral TKA from May 16, 2006, to December 31, 2013. Patients received either aspirin (324-325 mg daily), enoxaparin (40-60 mg daily), fondaparinux (2.5 mg daily), or warfarin (all doses) and were followed up 90 days postoperatively on several outcomes: deep vein thrombosis, pulmonary embolism, major bleeding, wound complications, infection, and death.

RESULTS

There was no evidence that fondaparinux, enoxaparin, or warfarin were superior to aspirin in the prevention of pulmonary embolism, deep vein thrombosis, or venous thromboembolism or that aspirin was safer than these alternatives. However, enoxaparin was found to be as safe as aspirin with respect to bleeding, and fondaparinux was as safe as aspirin for risk of wound complications.

CONCLUSION

Among TKA patients, we did not find evidence for decreased effectiveness or increased safety with use of aspirin, but enoxaparin had comparable safety to aspirin for bleeding and fondaparinux had comparable safety to aspirin for wound complications.

Impact of nosocomial infections surveillance on nosocomial infection rates: A systematic review

BACKGROUND

According to previously studies, nosocomial infections (NIs) surveillance could effectively reduce infection rates. As NIs surveillance systems have been implemented in some hospitals for several years, their impact on NIs need to be explored. Therefore, the purpose of this review is to evaluate the tendency of NI rates during the surveillance period and the impact of surveillance on NI rates.

METHODS

A systematic literature search of the PubMed database to identify papers that evaluated effect of surveillance on NIs, all kinds of NIs occurred during hospitalization or discharged were included. Exclude articles investigated the surveillance combined with other infection control measures.

RESULTS

Twenty-five articles were included. NI rates had different levels of reduction during surveillance period, the reduction were not limited by state, department, surveillance system, and NI type. Continuous surveillance had a positive impact on NI, OR/RR were ranged from 0.43 to 0.95.

CONCLUSION

Participation in NI surveillance is associated with reducing infection rates, though RCTs need to further prove the effective role of surveillance. Hospitals may consider to perform NIs surveillance systems according to its own conditions.

A real-time surgical site infections surveillance mode to monitor surgery classification-specific, hospital-wide surgical site infections in a Chinese tertiary hospital

We introduced a real-time surgical site infections surveillance mode (SSISM) to monitor hospital-wide surgical site infections (SSIs) based on the ICD-9-CM Volume 3 operational codes and the ICD-10 disease codes. Compared with the gold standard, the SSISM confirmed 71.9% (82/114) of SSIs from 3,048 operations with a 60-fold time-savings. The SSISM could obtain the SSI rates for each type of surgery or disease among hospital-wide inpatients in a tertiary hospital with 3,800 beds.

Efficacy of Prophylactic Antibiotics in Simple Knee Arthroscopy

PURPOSE

To determine the association between the use of preoperative antibiotics and the risk of postoperative infection after simple knee arthroscopy.

METHODS

The electronic medical records of a large integrated health care organization were used to identify patients who underwent simple knee arthroscopy between 2007 and 2012. Patient demographics, potential infection risk factors, and antibiotic administration data were extracted. Simple knee arthroscopy included debridement, meniscectomy, meniscus repair, synovectomy, microfracture, and lateral release. Complex knee arthroscopy, septic knees, and cases involving fractures were excluded. Deep infection was defined as a positive synovial fluid culture or signs and symptoms of infection and gross pus in the knee. Superficial infection was defined as clinical signs of infection localized to a portal site and treatment with an antibiotic.

RESULTS

Of 40,810 simple knee arthroscopies, 32,836 (80.5%) received preoperative antibiotics and 7,974 (19.5%) did not. There were 25 deep infections in the antibiotic group (0.08%) and 11 in the no-antibiotics group (0.14%) (risk ratio = 0.55, 95% confidence interval: 0.27 to 1.12, P = .10). There were 134 superficial infections in the antibiotic group (0.41%) and 32 in the no-antibiotics group (0.40%) (risk ratio = 1.01, 95% confidence interval: 0.29 to 1.49, P = .93).

CONCLUSIONS

In our large sample of patients who underwent simple knee arthroscopy, there was no association between preoperative antibiotic use and postoperative deep or superficial infection rates at the 95% confidence level (P = .05). There was an association between preoperative antibiotic use and a decreased deep infection rate at the P = .10 level.

LEVEL OF EVIDENCE

Level IV, case series.

New Zealand Joint Registry data underestimates the rate of prosthetic joint infection

Background and purpose - Recent studies have revealed deficiencies in the accuracy of data from joint registries when reoperations for prosthetic joint infections (PJIs) are reported, particularly when no components are changed. We compared the accuracy of data from the New Zealand Joint Registry (NZJR) to a multicenter audit of hospital records to establish the rate of capture for PJI reoperations. Methods - 4,009 cases undergoing total knee or hip arthroplasty performed at 3 tertiary referral hospitals over a 3-year period were audited using multiple hospital datasets and the NZJR. The number of reoperations for PJI that were performed within 2 years of the primary arthroplasty was obtained using both methods and the data were compared. Results - The NZJR reported a 2-year reoperation rate for PJI of 0.67%, as compared to 1.1% from the audit of hospital records, giving the NZJR a sensitivity of 63%. Only 4 of 11 debridement-in-situ-only procedures and 7 of 12 modular exchange procedures were captured in the NZJR. Interpretation - The national joint registry underestimated the rate of reoperation for PJI by one third. Strategies for improving the accuracy of data might include revising and clarifying the registry forms to include all reoperations for PJI and frequent validation of the registry data against other databases.

Quality Assessment of Hospital Discharge Database for Routine Surveillance of Hip and Knee Arthroplasty–Related Infections

Objective.

Surgical site infection (SSI) surveillance represents a key method of nosocomial infection control programs worldwide. However, most SSI surveillance systems are considered to be poorly cost effective regarding human and economic resources required for data collection and patient follow up. This study aims to assess the efficacy of using hospital discharge databases (HDDs) as a routine surveillance system for detecting hip or knee arthroplasty–related infections (HKAIs).

Methods.

A case-control study was conducted among patients hospitalized in the Centre region of France between 2008 and 2010. HKAI cases were extracted from the HDD with various algorithms based on the International Classification of Diseases, Tenth Revision, and procedure codes. The control subjects were patients with hip or knee arthroplasty (HKA) without infection selected at random from the HDD during the study period. The gold standard was medical chart review. Sensitivity (Se), specificity (Spe), positive predictive value (PPV), and negative predictive value (NPV) were calculated to evaluate the efficacy of the surveillance system.

Results.

Among 18,265 hospital stays for HKA, corresponding to 17,388 patients, medical reports were checked for 1,010 hospital stays (989 patients). We identified 530 cases in total (incidence rate, 1% [95% confidence interval (CI), 0.4%–1.6%), and 333 cases were detected by routine surveillance. As compared with 480 controls, Se was 98%, Spe was 71%, PPV was 63%, and NPV was 99%. Using a more specific case definition, based on a sample of 681 hospital stays, Se was 97%, Spe was 95%, PPV was 87%, and NPV was 98%.

Conclusions.

This study demonstrates the potential of HDD as a tool for routine SSI surveillance after low-risk surgery, under conditions of having an appropriate algorithm for selecting infections.

Infect Control Hosp Epidemiol 2014;35(6):646–651

Impact of an Automated Surveillance to Detect Surgical-Site Infections in Patients Undergoing Total Hip and Knee Arthroplasty in Brazil

In this retrospective study, we compared automated surveillance with conventional surveillance to detect surgical site infection after primary total hip or knee arthroplasty. Automated surveillance demonstrated better efficacy than routine surveillance in SSI diagnosis, sensitivity, and predictive negative value in hip and knee arthroplasty. Infect Control Hosp Epidemiol 2016;37:991-993.

Accuracy of administrative data for surveillance of healthcare-associated infections: a systematic review

OBJECTIVE

Measuring the incidence of healthcare-associated infections (HAI) is of increasing importance in current healthcare delivery systems. Administrative data algorithms, including (combinations of) diagnosis codes, are commonly used to determine the occurrence of HAI, either to support within-hospital surveillance programmes or as free-standing quality indicators. We conducted a systematic review evaluating the diagnostic accuracy of administrative data for the detection of HAI.

METHODS

Systematic search of Medline, Embase, CINAHL and Cochrane for relevant studies (1995-2013). Methodological quality assessment was performed using QUADAS-2 criteria; diagnostic accuracy estimates were stratified by HAI type and key study characteristics.

RESULTS

57 studies were included, the majority aiming to detect surgical site or bloodstream infections. Study designs were very diverse regarding the specification of their administrative data algorithm (code selections, follow-up) and definitions of HAI presence. One-third of studies had important methodological limitations including differential or incomplete HAI ascertainment or lack of blinding of assessors. Observed sensitivity and positive predictive values of administrative data algorithms for HAI detection were very heterogeneous and generally modest at best, both for within-hospital algorithms and for formal quality indicators; accuracy was particularly poor for the identification of device-associated HAI such as central line associated bloodstream infections. The large heterogeneity in study designs across the included studies precluded formal calculation of summary diagnostic accuracy estimates in most instances.

CONCLUSIONS

Administrative data had limited and highly variable accuracy for the detection of HAI, and their judicious use for internal surveillance efforts and external quality assessment is recommended. If hospitals and policymakers choose to rely on administrative data for HAI surveillance, continued improvements to existing algorithms and their robust validation are imperative.

Effectiveness of Using Mobile Phone Image Capture for Collecting Secondary Data: A Case Study on Immunization History Data Among Children in Remote Areas of Thailand

Background

Entering data onto paper-based forms, then digitizing them, is a traditional data-management method that might result in poor data quality, especially when the secondary data are incomplete, illegible, or missing. Transcription errors from source documents to case report forms (CRFs) are common, and subsequently the errors pass from the CRFs to the electronic database.

Objective

This study aimed to demonstrate the usefulness and to evaluate the effectiveness of mobile phone camera applications in capturing health-related data, aiming for data quality and completeness as compared to current routine practices exercised by government officials.

Methods

In this study, the concept of “data entry via phone image capture” (DEPIC) was introduced and developed to capture data directly from source documents. This case study was based on immunization history data recorded in a mother and child health (MCH) logbook. The MCH logbooks (kept by parents) were updated whenever parents brought their children to health care facilities for immunization. Traditionally, health providers are supposed to key in duplicate information of the immunization history of each child; both on the MCH logbook, which is returned to the parents, and on the individual immunization history card, which is kept at the health care unit to be subsequently entered into the electronic health care information system (HCIS). In this study, DEPIC utilized the photographic functionality of mobile phones to capture images of all immunization-history records on logbook pages and to transcribe these records directly into the database using a data-entry screen corresponding to logbook data records. DEPIC data were then compared with HCIS data-points for quality, completeness, and consistency.

Results

As a proof-of-concept, DEPIC captured immunization history records of 363 ethnic children living in remote areas from their MCH logbooks. Comparison of the 2 databases, DEPIC versus HCIS, revealed differences in the percentage of completeness and consistency of immunization history records. Comparing the records of each logbook in the DEPIC and HCIS databases, 17.3% (63/363) of children had complete immunization history records in the DEPIC database, but no complete records were reported in the HCIS database. Regarding the individual’s actual vaccination dates, comparison of records taken from MCH logbook and those in the HCIS found that 24.2% (88/363) of the children’s records were absolutely inconsistent. In addition, statistics derived from the DEPIC records showed a higher immunization coverage and much more compliance to immunization schedule by age group when compared to records derived from the HCIS database.

Conclusions

DEPIC, or the concept of collecting data via image capture directly from their primary sources, has proven to be a useful data collection method in terms of completeness and consistency. In this study, DEPIC was implemented in data collection of a single survey. The DEPIC concept, however, can be easily applied in other types of survey research, for example, collecting data on changes or trends based on image evidence over time. With its image evidence and audit trail features, DEPIC has the potential for being used even in clinical studies since it could generate improved data integrity and more reliable statistics for use in both health care and research settings.

Surgical Site Infection After Primary Hip and Knee Arthroplasty: A Cohort Study Using a Hospital Database

BACKGROUND

Hip or knee arthroplasty infection (HKAI) leads to heavy medical consequences even if rare.

OBJECTIVE

To assess the routine use of a hospital discharge detection algorithm of prosthetic joint infection as a novel additional tool for surveillance.

METHODS

A historic 5-year cohort study was built using a hospital database of people undergoing a first hip or knee arthroplasty in 1 French region (2.5 million inhabitants, 39 private and public hospitals): 32,678 patients with arthroplasty code plus corresponding prosthetic material code were tagged. HKAI occurrence was then tracked in the follow-up on the basis of a previously validated algorithm using International Statistical Classification of Disease, Tenth Revision, codes as well as the surgical procedures coded. HKAI density incidence was estimated during the follow-up (up to 4 years after surgery); risk factors were analyzed using Cox regression.

RESULTS

A total of 604 HKAI patients were identified: 1-year HKAI incidence was1.31%, and density incidence was 2.2/100 person-years in hip and 2.5/100 person-years in knee. HKAI occurred within the first 30 days after surgery for 30% but more than 1 year after replacement for 29%. Patients aged 75 years or older, male, or having liver diseases, alcohol abuse, or ulcer sore had higher risk of infection. The inpatient case fatality in HKAI patients was 11.4%.

CONCLUSIONS

The hospital database method used to measure occurrence and risk factors of prosthetic joint infection helped to survey HKAI and could optimize healthcare delivery.

Infect Control Hosp Epidemiol 2015;36(10):1198–1207

Data elements and validation methods used for electronic surveillance of health care-associated infections: a systematic review

BACKGROUND

We describe the primary data sources, data elements, and validation methods currently used in electronic surveillance systems (ESS) for identification and surveillance of health care-associated infections (HAIs), and compares these data elements and validation methods with recommended standards.

METHODS

Using Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, a PubMed and manual search was conducted to identify research articles describing ESS for identification and surveillance of HAIs published January 1, 2009-August 31, 2014. Selected articles were evaluated to determine what data elements and validation methods were included.

RESULTS

Among the 509 articles identified in the original literature search, 30 met the inclusion criteria. Whereas the majority of studies (83%) used recommended data sources and validated the numerator (80%), only 10% of studies performed external and internal validation. In addition, there was variation in the ESS data formats used.

CONCLUSIONS

Our findings suggest that the majority of ESS for HAI surveillance use standard definitions, but the lack of widespread internal data, denominator, and external validation in these systems reduces the reliability of their findings. Additionally, advanced programming skills are required to create, implement, and maintain these systems and to reduce the variability in data formats.

A Community-Based Hip Fracture Registry: Population, Methods, and Outcomes

Cases of hip fracture recorded from 1/2009 to 12/2011 were ascertained using the Kaiser Permanente Hip Fracture Registry. The registry collects information on patient, procedure, surgeon, facility, and surgical outcomes. The population (N = 12,562) was predominantly white, women, and older (≥ 75 years), and 32% had at least 5 comorbidities. The average length of follow-up was 1.1 years. Hemiarthroplasty was the most common procedure (33.1%). Most fractures were treated by medium-volume surgeons at high-volume facilities. The 90-day readmission rate was 22.1%, and the mortality rate was 12.3%.

Two-Day Length of Stay is Not Inferior to 3 Days in Total Knee Arthroplasty with Regards to 30-Day Readmissions

The impact of a shortened length of stay (LOS) following total knee arthroplasty (TKA) on the risk of readmission is not well documented despite recent trends towards shorter hospitalization. We retrospectively compared the adjusted risk of 30-day readmission following TKA between patients with 2-, 3- and 4-day LOS using current postoperative care protocols. A total of 23,655 consecutive primary, unilateral TKAs operated between 01/01/2009 and 12/31/2011 were studied retrospectively using non-inferiority testing. The main outcome was 30-day readmission. Two-day LOS decreased the odds of readmission by a factor of 0.96, with an upper bound one-sided 95% confidence interval of 1.10. After adjusting for other variables, LOS of 2 days is not inferior to 3 days with respect to the risk of 30-day readmission.

Surgical site infections after hip arthroplasty in Norway, 2005-2011: influence of duration and intensity of postdischarge surveillance

BACKGROUND

Most surgical site infections (SSIs) after hip arthroplasty are detected after a patient is discharged from hospital, making postdischarge surveillance (PDS) an important component in surveillance systems. We investigated how long it was necessary to monitor hip arthroplasty patients for SSIs after hospital discharge and if passive PDS through readmissions could replace active PDS by patient questionnaire in detecting SSIs.

METHODS

We used data from the Norwegian surveillance system from 2005-2011, which has active 1-year PDS, to investigate proportions of SSIs found at different time intervals after surgery and whether these SSIs could have been detected through passive PDS by investigating the proportion of patients with SSIs that were readmitted.

RESULTS

We found that 79% of all SSIs and 82% of deep SSIs were detected after hospital discharge. 95% of deep SSIs were detected within 90 days after surgery. 14% of the deep SSIs were detected beyond 30 days after surgery, and all of these patients were readmitted because of their SSI and thus could have been detected by passive PDS.

CONCLUSIONS

Our data suggest that most deep SSIs are detected within 90 days and that passive PDS beyond 30 days after surgery may replace active PDS without reducing sensitivity.

No decreased infection rate when using antibiotic-impregnated cement in primary total joint arthroplasty

There has been much debate and controversy about the routine use of antibiotic-impregnated bone cement in primary total joint arthroplasty. The purpose of this study was to undertake a meta-analysis to determine whether the use of antibiotic-impregnated bone cement would reduce the incidence of infection after primary total joint arthroplasty. Of 313 citations identified for screening, 6 trials reporting 26,791 patients were eligible for data extraction and meta-analysis. The authors found no statistically significant difference between antibiotic-impregnated bone cement and plain bone cement in terms of the incidence of infection. The results indicated that the use of antibiotic-impregnated bone cement in primary total joint arthroplasty did not lead to a decrease in the rate of infection.

Validation of reoperations due to infection in the Swedish Hip Arthroplasty Register

Background

Complete or almost complete recording of reoperations is essential to enable a correct interpretation of data in arthroplasty registers. The completeness of recordings due to infection is unknown in the Swedish Hip Arthroplasty Register (SHAR). We therefore used a combination of data from the Swedish Prescribed Drug Register (SPDR) and studies of medical records to validate the data of reoperations due to infection in the SHAR.

Methods

All patients registered for a primary Total Hip Replacement (THR) in the SHAR between July 1, 2005 and December 31, 2008 were selected for the study (45,531 patients with 49,219 THRs) and were matched with the SPDR. All patients with a minimum of 4 weeks of continuous outpatient antibiotic treatment within 2 years after their primary THR (1,989 patients, with 2,219 THRs) were selected for a medical records review to find the THRs reoperated due to infection.

Results

599 (1.3%) of the THRs had been reoperated within 2 years after the index operation and in 47.4% of these the prosthesis had been revised or extracted. 400 of the THRs were registered for a reoperation in the SHAR resulting in a completeness of 67%.

Conclusions

The completeness of registration due to early infection after THR questions whether the SHAR reoperation data can be used in order to evaluate changes in postoperative infection rates.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2474-15-384) contains supplementary material, which is available to authorized users.

Identifying colon and open reduction of fracture surgical site infections using a partially automated electronic algorithm

BACKGROUND

Algorithms leveraging electronic data may reduce manual review burden for surgical site infection (SSI) surveillance with little to no reduction in sensitivity. We developed an algorithm to identify colon and open reduction of fracture (FX) SSIs to reduce manual chart review.

METHODS

A retrospective cohort of colon and FX procedures and associated SSIs was constructed. Potential SSIs were identified by positive microbiologic cultures or administrative data for diagnosis or treatment of wound infection. Sensitivity and specificity of the algorithm were assessed. The number of charts needing review to identify 1 SSI, and the potential time-savings from the algorithm, were calculated.

RESULTS

Four hundred seventy-three colon (SSI rate = 7%) and 1081 FX (SSI rate = 3%) procedures were identified. The algorithm was 91% and 97% sensitive and 76% and 93% specific for colon and FX procedures, respectively. Overall, chart review would have been reduced by 24.3 hours per 100 procedures, decreasing the number of charts to review to identify 1 SSI from 23.9 for manual review to 3.9 with the algorithm.

CONCLUSIONS

The algorithm identified SSIs with excellent sensitivity and specificity, resulting in substantial reductions in manual chart review. This algorithm could be tailored and applied to other hospitals.

The effect of chronic kidney disease on total hip arthroplasty

Patients with chronic kidney disease (CKD) undergoing total hip arthroplasty (THA) were evaluated for risk of revision, surgical site infection (SSI), thromboembolic events, mortality and readmission. 20,720 primary TKA cases were included (smaller sample for readmission evaluation, N = 9322). The prevalence of CKD among THA patients was 6.1% (N = 1269). After adjustment for age, gender, race, general health, and diabetes, CKD patients were at 1.4 (95% confidence interval 1.1-1.8) increased risk of readmission within 90 days. The adjusted risks for revision (overall, aseptic, and septic), SSI (deep and superficial), deep vein thrombosis, pulmonary embolism, and mortality (30-day, 90-day, ever) were not significantly different between patients with CKD and those without CKD. However, increased risk for 90-day readmission underscores that CKD patients are a fundamentally different population of patients.

Is it possible to identify cases of coronary artery bypass graft postoperative surgical site infection accurately from claims data?

Background

Claims data has usually been used in recent studies to identify cases of healthcare-associated infection. However, several studies have indicated that the ICD-9-CM codes might be inappropriate for identifying such cases from claims data; therefore, several researchers developed alternative identification models to correctly identify more cases from claims data. The purpose of this study was to investigate three common approaches to develop alternative models for the identification of cases of coronary artery bypass graft (CABG) surgical site infection, and to compare the performance between these models and the ICD-9-CM model.

Methods

The 2005–2008 National Health Insurance claims data and healthcare-associated infection surveillance data from two medical centers were used in this study for model development and model verification. In addition to the use of ICD-9-CM codes, this study also used classification algorithms, a multivariable regression model, and a decision tree model in the development of alternative identification models. In the classification algorithms, we defined three levels (strict, moderate, and loose) of the criteria in terms of their strictness. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were used to evaluate the performance of each model.

Results

The ICD-9-CM-based model showed good specificity and negative predictive value, but sensitivity and positive predictive value were poor. Performances of the other models were varied, except for negative predictive value. Among the models, the performance of the decision tree model was excellent, especially in terms of positive predictive value.

Conclusion

The accuracy of identification of cases of CABG surgical site infection is an important issue in claims data. Use of the decision tree model to identify such cases can improve the accuracy of patient-level outcome research. This model should be considered when performing future research using claims data.

The impact of pre-operative weight loss on incidence of surgical site infection and readmission rates after total joint arthroplasty

This study characterized a cohort of obese total hip arthroplasty (THA) and total knee arthroplasty (TKA) patients (1/1/2008-12/31/2010) and evaluated whether a clinically significant amount of pre-operative weight loss (5% decrease in body weight) is associated with a decreased risk of surgical site infections (SSI) and readmissions post-surgery. 10,718 TKAs and 4066 THAs were identified. During the one year pre-TKA 7.6% of patients gained weight, 12.4% lost weight, and 79.9% remained the same. In the one year pre-THA, 6.3% of patients gained weight, 18.0% lost weight, and 75.7% remained the same. In TKAs and THAs, after adjusting for covariates, the risk of SSI and readmission was not significantly different in the patients who gained or lost weight pre-operatively compared to those who remained the same.