High Outliers Research Articles

American College of Surgeons National Surgical Quality Improvement Program Pediatric: A beta phase report

PurposeThe American College of Surgeons (ACS) National Surgical Quality Improvement Program Pediatric (NSQIP-P) expanded to beta phase testing with the enrollment of 29 institutions. Data collection and analysis were aimed at program refinement and development of risk-adjusted models for inter-institutional comparisons. MethodsData from the first full year of beta-phase NSQIP-P were analyzed. Patient accrual used ACS-NSQIP methodology tailored to pediatric specialties. Preliminary risk adjusted modeling for all pediatric and neonatal operations and pediatric (excluding neonatal) abdominal operations was performed for all cause morbidity (other than death) and surgical site infections (SSI) using hierarchical logistic regression methodology and eight predictor variables. Results were expressed as odds ratios with 95% confidence intervals. ResultsDuring calendar year 2010, 29 institutions enrolled 37,141 patients. 1644 total CPT codes were entered, of which 456 accounted for 90% of the cases. 450 codes were entered only once (1.2% of cases). For all cases, overall mortality was 0.25%, overall morbidity 7.9%, and the SSI rate 1.8%. For neonatal cases, mortality was 2.39%, morbidity 18.7%, and the SSI rate 3%. For the all operations model, risk-adjusted morbidity institutional odds ratios ranged 0.48–2.63, with 9/29 hospitals categorized as low outliers and 9/29 high outliers, while risk-adjusted SSI institutional odds ratios ranged 0.36–2.04, with 2/29 hospitals low outliers and 7/29 high outliers. ConclusionThis report represents the first risk-adjusted hospital-level comparison of surgical outcomes in infants and children using NSQIP-P data. Programmatic and analytic modifications will improve the impact of this program as it moves into full implementation. These results indicate that NSQIP-P has the potential to serve as a model for determining risk-adjusted outcomes in the neonatal and pediatric population with the goal of developing quality improvement initiatives for the surgical care of children.

Cost Analysis of Integrative Inpatient Treatment Based on DRG Data: The Example of Anthroposophic Medicine

Background. Much work has been done to evaluate the outcome of integrative inpatient treatment but scarcely the costs. This paper evaluates the costs for inpatient treatment in three anthroposophic hospitals (AHs). Material and Methods. Cost and performance data from a total of 23,180 cases were analyzed and compared to national reference data. Subgroup analysis was performed between the cases with and without anthroposophic medical complex (AMC) treatment. Results. Costs and length of stay in the cases without AMC displayed no relevant differences compared to the national reference data. In contrast the inlier cases with AMC caused an average of € 1,394 more costs. However costs per diem were not higher than those in the national reference data. Hence, the delivery of AMC was associated with a prolonged length of stay. 46.6% of the cases with AMC were high outliers. Only 10.6% of the inlier cases with AMC were discharged before reaching the mean length of stay of each DRG. Discussion. Treatment in an AH is not generally associated with an increased use of resources. However, the provision of AMC leads to a prolonged length of stay and cannot be adequately reimbursed by the current G-DRG system. Due to the heterogeneity of the patient population, an additional payment should be negotiated individually.

Colon and Rectal Surgery Is a “High Outlier” Specialty: A Case Study Using Postoperative Urinary Tract Infection

The incidence of major complications after colorectal resection remains high. Rates may even exceed those predicted by risk-adjustment models commonly in use. Colon and rectal resections account for the greatest share of all postoperative morbidity and mortality in the National Surgical Quality Improvement Program general surgery cohort, and a disproportionate share of the costs associated with postoperative complications. Infectious complications are particularly overrepresented, as surgical site and urinary tract infections are 2-3 times more common after colon and rectal surgery than other general surgery operations. Thus, surgeons with a specialty interest in colorectal surgery may be flagged as “high outliers” unless more is done to account for procedure-specific risk.

The effect of dead-on-arrival and emergency department death classification on risk-adjusted performance in the American College of Surgeons Trauma Quality Improvement Program

The American College of Surgeons' Trauma Quality Improvement Program is focused on identifying variations in outcomes across trauma centers for the purposes of performance improvement. In previous analyses, patients who died in the emergency department were excluded. We investigated the effect of inclusion and exclusion of emergency department (ED) deaths (dead on arrival [DOA] and died in ED [DIE]) on analyses of overall risk-adjusted trauma center performance. Data for patients admitted to 65 Trauma Quality Improvement Program hospitals during the 2009 calendar year was used. A logistic regression model was developed to estimate risk-adjusted mortality. Trauma centers were then ranked based on their observed-to-expected (O/E) mortality ratio with 90% confidence intervals (CIs) and classified by outlier status: low outliers/high performers had a 90% CI for O/E mortality ratio of less than 1, and high outliers/low performers had a 90% CI for O/E mortality ratio of greater than 1. Changes in outlier status, rank, and quartile were examined with and without DOA and DIE patients included in the analyses to discern the impact of such exclusions on overall risk-adjusted center-specific performance. Thirty-one trauma centers (48%) reported no DOA patients in 2009, while 6 centers (9%) reported more than 10. Of 224 patients, 14 (6.2%) had a documented time of death of more than 30 minutes after ED arrival despite being recorded as DOA. Forty-one trauma centers (63%) changed rank by three positions or less. Ten trauma centers changed their quartile ranking by a single quartile, but no centers were found to change quartile rank more than one quartile. Changes in outlier status occurred for 6 trauma centers (9%). The relative frequency of patients classified as DOA varies greatly between trauma centers. Misclassification of patients as DOA occurs. Inclusion of ED deaths in risk-adjusted analysis of mortality results in a small but insignificant change in predicting the outcome results of a trauma center. This change is less than the rate of finding a center to be a high or low outlier by chance alone using the 90% CI. Inclusion of DOA and DIE patients in risk-adjusted analysis of mortality is appropriate and eliminates the bias introduced by exclusion of ED deaths owing to misuse of the DOA classification. Prognostic/epidemiologic study, level III.

높은 이상점 비율을 갖는 고감도 가이거모드 영상 라이다 데이터로부터 이상점 검출

라이다 센서로 취득된 점군에는 실제 물리적인 표면에 존재하지 않는 이상점이 포함되어 있다. 이러한 이상점들은 활용을 위한 후속처리를 하기 전에 반드시 제거되어야 한다. 특히 민감도가 아주 높은 가이거 모드 검출기를 이용하는 라이다로 취득한 데이터는 높은 비율의 이상점을 포함하고 있다. 이로 인해 기존의 알고리즘은 이러한 데이터로부터 성공적으로 이상점을 검출하는데 어려움이 있었다. 이에 본 연구는 가이거 모드 영상 라이다로 획득된 높은 이상점 비율을 갖는 점군에서 이상점을 제거하는 방법을 제안한다. 제안된 방법은 의미 있는 표적의 표면은 검출기상에서 두 개 이상의 이웃픽셀에 검출되며, 이러한 이웃픽셀들로부터 출력되는 거리값은 유사하다는 점을 이용한다. 개발된 제거 기법은 시뮬레이션으로 생성된 다양한 점밀도와 이상점 비율의 모의 데이터에 적용하여 임계값과 데이터 특성에 따른 성능을 분석하였다. 대부분의 경우에 약 99% 이상의 이상점 검출성능이 나타났으며, 데이터 특성에 강인하고 임계값에 크게 민감하지 않는 검출성능을 확인하였다. 제안된 방법은 향후 가이거 모드 라이다 데이터의 온라인 실시간 처리 또는 후처리에 효과적으로 활용될 것으로 판단된다. Point clouds acquired by a LIDAR(Light Detection And Ranging, also LADAR) system often contain erroneous points called outliers seeming not to be on physical surfaces, which should be carefully detected and eliminated before further processing for applications. Particularly in case of LIDAR systems employing with a Gieger-mode array detector (GmFPA) of high sensitivity, the outlier ratio is significantly high, which makes existing algorithms often fail to detect the outliers from such a data set. In this paper, we propose a method to discriminate outliers from a point cloud with high outlier ratio acquired by a GmFPA LIDAR system. The underlying assumption of this method is that a meaningful targe surface occupy at least two adjacent pixels and the ranges from these pixels are similar. We applied the proposed method to simulated LIDAR data of different point density and outlier ratio and analyzed the performance according to different thresholds and data properties. Consequently, we found that the outlier detection probabilities are about 99% in most cases. We also confirmed that the proposed method is robust to data properties and less sensitive to the thresholds. The method will be effectively utilized for on-line realtime processing and post-processing of GmFPA LIDAR data.

Landscape heterogeneity and local adaptation define the spatial genetic structure of Pacific salmon in a pristine environment

Identifying the spatial distribution of genetic variation across the landscape is an essential step in informing species conservation. Comparison of closely related and geographically overlapping species can be particularly useful in cases where landscape may similarly influence genetic structure. Congruent patterns among species highlight the importance that landscape heterogeneity plays in determining genetic structure whereas contrasting patterns emphasize differences in species-specific ecology and life-history or the importance of species-specific adaptation to local environments. We examined the interacting roles of demography and adaptation in determining spatial genetic structure in two closely related and geographically overlapping species in a pristine environment. Using single nucleotide polymorphism (SNP) loci exhibiting both neutral and putative adaptive variation, we evaluated the genetic structure of sockeye salmon in the Copper River, Alaska; these data were compared to existing data for Chinook salmon from the same region. Overall, both species exhibited patterns of isolation by distance; the spatial distribution of populations largely determined the distribution of genetic variation across the landscape. Further, both species exhibited largely congruent patterns of within- and among-population genetic diversity, highlighting the role that landscape heterogeneity and historical processes play in determining spatial genetic structure. Potential adaptive differences among geographically proximate sockeye salmon populations were observed when high FST outlier SNPs were evaluated in a landscape genetics context. Results were evaluated in the context of conservation efforts with an emphasis on reproductive isolation, historical processes, and local adaptation.

Outlier Detection Algorithm Basing on Similarity Measurement Relation

Outlier detection is an important field of data mining, which is widely used in credit card fraud detection, network intrusion detection ,etc. A kind of high dimensional data similarity metric function and the concept of class density are given in the paper, basing on the combination of hierarchical clustering and similarity, as well as outlier detection algorithm about similarity measurement is presented after the redefinition of high dimension density outliers is put. The algorithm has some value for outliers detection of high dimensional data set in view of experimental result.

Mitigating High Latency Outliers for Cloud-Based Telecommunication Services

Telecommunication applications are distinguished by their stringent requirements for availability and completion times. A highly available, low-latency, distributed data store is therefore a critical component of cloud-based realizations of telecommunication services. We present a systematic experimental evaluation of state-of-the-art database systems as components of telecommunication applications. We show that while their average latencies are well within the required time scales, the distribution of latencies exhibits a long tail of unacceptably large outliers which may significantly impair meeting the performance requirements of telecommunication applications. To address the observed phenomenon of high latency outliers, we present a new solution that is implemented in a Bell Labs system code named Flurry. Flurry is based on using the first response from a replica rather than waiting for all or a quorum of responses from replicas. To handle incorrect responses arising from message losses, Flurry uses a novel checking algorithm based on vector clocks to determine the correctness of a replica's response. We present experimental evaluation results which show that Flurry significantly reduces both the average response time and the probability of unacceptable response times to values that would allow meeting the availability and completion time thresholds required for telecommunication services. © 2012 Alcatel-Lucent.

Factors influencing hospital high length of stay outliers

BackgroundThe study of length of stay (LOS) outliers is important for the management and financing of hospitals. Our aim was to study variables associated with high LOS outliers and their evolution over time.MethodsWe used hospital administrative data from inpatient episodes in public acute care hospitals in the Portuguese National Health Service (NHS), with discharges between years 2000 and 2009, together with some hospital characteristics. The dependent variable, LOS outliers, was calculated for each diagnosis related group (DRG) using a trim point defined for each year by the geometric mean plus two standard deviations. Hospitals were classified on the basis of administrative, economic and teaching characteristics. We also studied the influence of comorbidities and readmissions. Logistic regression models, including a multivariable logistic regression, were used in the analysis. All the logistic regressions were fitted using generalized estimating equations (GEE).ResultsIn near nine million inpatient episodes analysed we found a proportion of 3.9% high LOS outliers, accounting for 19.2% of total inpatient days. The number of hospital patient discharges increased between years 2000 and 2005 and slightly decreased after that. The proportion of outliers ranged between the lowest value of 3.6% (in years 2001 and 2002) and the highest value of 4.3% in 2009. Teaching hospitals with over 1,000 beds have significantly more outliers than other hospitals, even after adjustment to readmissions and several patient characteristics.ConclusionsIn the last years both average LOS and high LOS outliers are increasing in Portuguese NHS hospitals. As high LOS outliers represent an important proportion in the total inpatient days, this should be seen as an important alert for the management of hospitals and for national health policies. As expected, age, type of admission, and hospital type were significantly associated with high LOS outliers. The proportion of high outliers does not seem to be related to their financial coverage; they should be studied in order to highlight areas for further investigation. The increasing complexity of both hospitals and patients may be the single most important determinant of high LOS outliers and must therefore be taken into account by health managers when considering hospital costs.

NSQIP in Otolaryngology: Our 4‐Year Experience

Objective: To present a 4-year experience with the American College of Surgeons (ACS) National Surgical Quality and Improvement Program (NSQIP) in Otolaryngology, Head and Neck Surgery, to highlight successes, and to demonstrate areas that provide opportunities for quality improvement. Method: The Henry Ford Health System (HFHS) database was utilized to analyze NSQIP outcomes and quality measures obtained for the Department of Otolaryngology–Head and Neck Surgery. This had been an ongoing project in our department over the last 4 years. Results: Overall mortality (0.34%) continues to be equal to the national average (0.34%). The overall wound infection rate (23.38%) has improved but is still slightly higher than the national average (17.4%). Initially, the wound infection rate was a significant high outlier (16.67%). The incidence of infection has improved to 0.00% (3rd Q, 2011). Individual categorical gains have been made in unplanned intubations, pneumonia rates, and off the ventilator rates within 48 hours. Sepsis and septic shock rates continue to be benchmarked compared to the national average. Patient length of stay with wound infection have improved and resulted in savings for the institution. Conclusion: Institutional participation in NSQIP has resulted in demonstrable improvements in patient care outcomes and key quality measures for patients treated at the Department of Otolaryngology–Head and Neck Surgery.

Comparison of Hospital Performance in Trauma vs Emergency and Elective General Surgery

As emergency general surgery (EMGS) and trauma care are increasingly being provided by the same personnel with overlapping resources, we postulated that the quality of care provided to EMGS and trauma patients would be similar. We also evaluated the relationship between trauma and elective general surgery (ELGS) care, believing that performance would be similar across these services as it reflects institutional culture. Retrospective cohort study comparing hospital performance in trauma and EMGS care and in trauma and ELGS care. Regression models for mortality and serious morbidity were constructed for trauma, EMGS, and ELGS hospitals contributing to both the National Trauma Data Bank (2007) and American College of Surgeons National Surgical Quality Improvement Program (2005-2008). Forty-six hospitals. Correlations of observed to expected ratios were examined. Outlier status (hospitals with CIs of observed to expected ratios excluding 1.0) was compared using weighted . There was no significant relationship between trauma and EMGS mortality (r=-0.01, P=.94; =-0.10, P=.61) or between trauma and ELGS mortality (r=0.23, P=.12; =0.07, P=.62). There was no significant relationship between trauma and EMGS morbidity (r=0.21, P=.17; =0.04, P=.63) or between trauma and ELGS morbidity (r=0.16, P=.30; =0.11, P=.37). No hospitals were consistently low or high outliers across all 3 groups. Trauma performance improvement programs are well established compared with those for EMGS. Although EMGS patients use similar structures and processes as trauma patients, there is a lack of correlation between the quality of care provided to trauma and EMGS patients; EMGS should be incorporated into trauma performance improvement programs.

Hospital performance reports based on severity adjusted mortality rates in patients with cirrhosis depend on the method of risk adjustment

Background. Hospital outcome report cards are used to judge provider performance, including for liver transplantation. We aimed to determine the impact of the choice of risk adjustment method on hospital rankings based on mortality rates in cirrhotic patients.Material and methods. We identified 68,426 cirrho-tic patients hospitalized in the Nationwide Inpatient Sample database. Four risk adjustment methods (the Charlson/Deyo and Elixhauser algorithms, Disease Staging, and All Patient Refined Diagnosis Related Groups) were used in logistic regression models for mortality. Observed to expected (O/E) death rates were calculated for each method and hospital. Statistical outliers with higher or lower than expected mortality were identified and rankings compared across methods.Results. Unadjusted mortality rates for the 553 hospitals ranged from 1.4 to 30% (overall, 10.6%). For 163 hospitals (29.5%), observed mortality differed significantly from expected when judged by one or more, but not all four, risk adjustment methods (25.9% higher than expected mortality and 3.6% lower than expected mortality). Only 28% of poor performers and 10% of superior performers were consistently ranked as such by all methods. Agreement between methods as to whether hospitals were flagged as outliers was moderate (kappa 0.51-0.59), except the Charlson/Deyo and Elixhauser algorithms which demonstrated excellent agreement (kappa 0.75).Conclusions. Hospital performance reports for patients with cirrhosis require sensitivity to the method of risk adjustment. Depending upon the method, up to 30% of hospitals may be flagged as outliers by one, but not all methods. These discrepancies could have important implications for centers erroneously labeled as high mortality outliers.

Single measures of performance do not reflect overall institutional quality in colorectal cancer surgery

ObjectiveTo evaluate overall performance of English colorectal cancer surgical units identified as outliers for a single quality measure—30 day inhospital mortality.Design144 542 patients that underwent primary major colorectal cancer resection...

Organizational Characteristics of High‐ and Low‐Performing Anticoagulation Clinics in the Veterans Health Administration

Anticoagulation clinics (ACCs) can improve anticoagulation control and prevent adverse events. However, ACCs vary widely in their performance on anticoagulation control. Our objective was to compare the organization and management of top-performing with that of bottom-performing ACCs. Three high outlier and three low outlier ACCs in the Veterans Health Administration (VA). Site visits with qualitative data collection and analysis. We conducted semi-structured interviews with ACC staff regarding work flow, staffing, organization, and quality assurance efforts. We also observed ACC operations and collected documents, such as the clinic protocol. We used grounded thematic analysis to examine site-level factors associated with high and low outlier status. High outlier sites were characterized by (1) adequate (pharmacist) staffing and effective use of (nonpharmacist) support personnel; (2) innovation to standardize clinical practice around evidence-based guidelines; (3) the presence of a quality champion for the ACC; (4) higher staff qualifications; (5) a climate of ongoing group learning; and (6) internal efforts to measure performance. Although high outliers had all of these features, no low outlier had more than two of them. The top-performing ACCs in the VA system shared six relatively recognizable characteristics. Efforts to improve performance should focus on these domains.

Do cancer-specific variables improve risk-adjusted hospital quality comparisons?

585 Background: For patients undergoing surgery for cancer, it has been suggested that risk-adjustment with cancer-specific variables is needed when evaluating short-term outcomes. Our objectives were to assess the influence of cancer-related variables on postoperative complications and hospital quality comparisons. Methods: Patients from ACS NSQIP and NCDB who underwent colorectal resection for cancer were linked (2006-2008) to create a dataset containing robust information on comorbidities, complications, and oncologic variables. Three hierarchical models were developed predicting the NSQIP outcome 30-day mortality or any serious morbidity using variables from (1) NSQIP only, (2) NCDB only, and (3) a combined model using NSQIP and NCDB. Models were compared with fit statistics and hospital outlier agreement. Results: From 146 NSQIP hospitals, 11401 patients underwent a colorectal resection for cancer, of which, 1954 (17%) experienced a mortality or serious morbidity event. The first five variables selected in the NCDB-only model were Charlson comorbidity score, neoadjuvant therapy use, T stage, primary payer, and M stage (c-statistic, 0.64; AIC, 9886). The first five variables selected in the NSQIP-only model were ASA class, preop sepsis, albumin, surgical procedure, and COPD (c-statistic, 0.66; AIC, 9787). In the combined model, neoadjuvant therapy use was the only cancer-specific variable selected in the top five. The remaining variables were ASA class, preop sepsis, albumin, and wound class (c-statistic, 0.67; AIC, 9455). At the hospital-level, the NCDB-only model identified three high outliers (worse than expected) and one low outlier (better than expected). Both the NSQIP-only and combined models identified the same four high and two low outlying hospitals (kappa: 1.0), which agreed marginally with the NCDB-only model (kappa: 0.59). Conclusions: Addition of cancer-specific variables to NSQIP models slightly improved model fit; however, hospital outcome comparisons were identical. For patients with colorectal cancer undergoing resection, cancer-related factors have limited predictive ability for short-term outcomes and did not influence hospital quality comparisons.

What can we learn from international comparisons of costs by DRG?

The objective of this study was to compare costs data by diagnosis related group (DRG) between Belgium and Switzerland. Our hypotheses were that differences between countries can probably be explained by methodological differences in cost calculations, by differences in medical practices and by differences in cost structures within the two countries. Classifications of DRG used in the two countries differ (AP-DRGs version 1.7 in Switzerland and APR-DRGs version 15.0 in Belgium). The first step of this study was to transform Belgian summaries into Swiss AP-DRGs. Belgian and Swiss data were calculated with a clinical costing methodology (full costing). Belgian and Swiss costs were converted into US$ PPP (purchasing power parity) in order to neutralize differences in purchasing power between countries. The results of this study showed higher costs in Switzerland despite standardization of cost data according to PPP. The difference is not explained by the case-mix index because this was similar for inliers between the two countries. The length of stay (LOS) was also quite similar for inliers between the two countries. The case-mix index was, however, higher for high outliers in Belgium, as reflected in a higher LOS for these patients. Higher costs in Switzerland are thus probably explained mainly by the higher number of agency staff by service in this country or because of differences in medical practices. It is possible to make international comparisons but only if there is standardization of the case-mix between countries and only if comparable accountancy methodologies are used. Harmonization of DRGs groups, nomenclature and accountancy is thus required.

Anti-type II collagen antibodies are associated with early radiographic destruction in rheumatoid arthritis

IntroductionWe have previously reported that high levels of antibodies specific for native human type II collagen (anti-CII) at the time of RA diagnosis were associated with concurrent but not later signs of inflammation. This was associated with CII/anti-CII immune complex (IC)-induced production of pro-inflammatory cytokines in vitro. In contrast, anti-cyclic citrullinated peptide antibodies (anti-CCP) were associated both with late inflammation and late radiological destruction in the same RA cohort. We therefore hypothesized that anti-CII are also associated with early erosions.MethodsTwo-hundred-and-fifty-six patients from an early RA cohort were included. Baseline levels of anti-CII, anti-CCP and anti-mutated citrullinated vimentin were analyzed with ELISA, and rheumatoid factor levels were determined by nephelometry. Radiographs of hands and feet at baseline, after one and after two years were quantified using the 32-joints Larsen erosion score.ResultsLevels of anti-CII were bimodally distributed in the RA cohort, with a small (3.1%, 8/256) group of very high outliers with a median level 87 times higher than the median for the healthy control group. Using a cut-off discriminating the outlier group that was associated with anti-CII IC-induced production of proinflammatory cytokines in vitro, baseline anti-CII antibodies were significantly (p = 0.0486) associated with increased radiographic damage at the time of diagnosis. Anti-CII-positive patient had also significantly increased HAQ score (p = 0.0303), CRP (p = 0.0026) and ESR (p = 0.0396) at the time of diagnosis but not during follow-up. The median age among anti-CII-positive subjects was 12 years higher than among the anti-CII-negative patients.ConclusionIn contrary to anti-CCP, anti-CII-positive patients with RA have increased joint destruction and HAQ score at baseline. Anti-CII thus characterizes an early inflammatory/destructive phenotype, in contrast to the late appearance of an inflammatory/destructive phenotype in anti-CCP positive RA patients. The anti-CII phenotype might account for part of the elderly acute onset RA phenotype with rather good prognosis.

Visual interactive evolutionary algorithm for high dimensional outlier detection and data clustering problems

Usual visualisation techniques for multidimensional datasets, such as parallel coordinates and scatterplot matrices, do not scale well to high numbers of dimensions. A common approach to solve this problem is dimensionality selection. Existing dimensionality selection techniques usually select pertinent dimension subsets that are significant to the user without loose of information. We present concrete cooperation between automatic algorithms, interactive algorithms and visualisation tools: the evolutionary algorithm is used to obtain optimal dimension subsets which represent the original dataset without loosing information for unsupervised mode (clustering or outlier detection). The last effective cooperation is a visualisation tool used to present the user interactive evolutionary algorithm results and let him actively participate in evolutionary algorithm searching with more efficiency resulting in a faster evolutionary algorithm convergence. We have implemented our approach and applied it to real dataset to confirm this approach is effective for supporting the user in the exploration of high dimensional datasets.

A Colorectal “Care Bundle” to Reduce Surgical Site Infections in Colorectal Surgeries: A Single-Center Experience

Kaiser Sunnyside Medical Center has participated in the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) since January 2006. Data on general and colorectal surgical site infections (SSIs) demonstrated a need for improvement in SSI rates. To evaluate application of a "care bundle" for patients undergoing colorectal operations, with the goal of reducing overall SSI rates. We prospectively implemented multiple interventions, with retrospective analysis of data using the NSQIP database. The overall, superficial, deep, and organ/space SSI rates were compared before and after implementation of this colorectal care bundle. Between January 2006 and December 2009, there were 430 colorectal cases in our NSQIP report with 91 infections, an overall rate of 21.16%. Between January 2010, when the colorectal care bundle was implemented, and June 2011, there were 195 cases and 13 infections, a 6.67% overall rate. The absolute decrease of 14.49% is significant (p < 0.0001). The rate of superficial SSI decreased from 15.12% to 3.59% (p < 0.0001). The rates for deep and organ/space SSI also showed a decrease; however, this was not statistically significant. The NSQIP observed-to-expected ratio for colorectal SSI decreased from a range of 1.27 to 1.83 before implementation to 0.54 after implementation (fiscal year 2010). Our institution was a NSQIP high outlier in general surgery SSIs and had a high proportion of these cases represented in colorectal cases. By instituting a care bundle composed of core and adjunct strategies, we significantly decreased our rate of colorectal SSIs.

Comparison of different approaches to evaluate External Quality Assessment Data

In EQA programs, Z-scores are used to evaluate laboratory performance. They should indicate poorly performing laboratories, regardless of the presence of outliers. For this, two different types of approaches exist. The first type are “outlier-based” approaches, which first exclude outlying values, calculate the average and standard deviation on the remaining data and obtain Z-scores for all values (e.g., Grubbs and Dixon). The second type includes the “robust” approaches (e.g., Tukey and Qn or the algorithm recommended by ISO). The different approaches were assessed by randomly generated samples from the Normal and Student t distributions. Part of the sample data were contaminated with outliers. The number of false and true outliers was recorded and subsequently, Positive and Negative Predictive Values were derived. Also, the sampling mean and variability were calculated for location and scale estimators. The various approaches performed similarly for sample sizes above 10 and when outliers were at good distance from the centre. For smaller sample sizes and closer outliers, however, the approaches performed quite differently. Tukey's method was characterised by a high true and a high false outlier rate, while the ISO and Qn approaches demonstrated weak performance. Grubbs test yielded overall the best results.