Central line-associated infections as defined by the Centers for Medicare and Medicaid Services' Hospital-acquired condition versus standard infection control surveillance: why hospital compare seems conflicted

Validating administrative data to identify complex surgical site infections following cardiac implantable electronic device implantation: a comparison of traditional methods and machine learning

Background

Cardiac implantable electronic device (CIED) surgical site infections (SSIs) have been outpacing the increases in implantation of these devices. While traditional surveillance of these SSIs by infection prevention and control would likely be the most accurate, this is not practical in many centers where resources are constrained. Therefore, we explored the validity of administrative data at identifying these SSIs.

Methods

We used a cohort of all patients with CIED implantation in Calgary, Alberta where traditional surveillance was done for infections from Jan 1, 2013 to December 31, 2019. We used this infection subgroup as our “gold standard” and then utilized various combinations of administrative data to determine which best optimized the sensitivity and specificity at identifying infection. We evaluated six approaches to identifying CIED infection using administrative data, which included four algorithms using International Classification of Diseases codes and/or Canadian Classification of Health Intervention codes, and two machine learning models. A secondary objective of our study was to assess if machine learning techniques with training of logistic regression models would outperform our pre-selected codes.

Results

We determined that all of the pre-selected algorithms performed well at identifying CIED infections but the machine learning model was able to produce the optimal method of identification with an area under the receiver operating characteristic curve (AUC) of 96.8%. The best performing pre-selected algorithm yielded an AUC of 94.6%.

Conclusions

Our findings suggest that administrative data can be used to effectively identify CIED infections. While machine learning performed the most optimally, in centers with limited analytic capabilities a simpler algorithm of pre-selected codes also has excellent yield. This can be valuable for centers without traditional surveillance to follow trends in SSIs over time and identify when rates of infection are increasing. This can lead to enhanced interventions for prevention of SSIs.

Unit-Level Changes in Central Line-Associated Bloodstream Infection Before and After Implementation of the Affordable Care Act and Mandatory Reporting Legislation

BACKGROUND

Central line-associated bloodstream infection (CLABSI) prevention efforts have increased over the past decade because of implications of the Affordable Care Act and mandatory reporting laws. These legislative measures allow for reduced reimbursement to hospitals with high level of CLABSIs and other health care-associated infections.

OBJECTIVE

The aim of this study was to explore the impact of legislation and mandatory reporting on CLABSI rates and reporting.

METHODS

The study team performed a retrospective review of medical intensive care unit patients in January 2008, 2012, and 2015 to examine changes in CLABSI reporting by 2 methods (International Classification of Diseases [ICD] by providers and Centers for Disease Control by infection prevention [IP]), as well as changes in central line use over time. Data were summarized and compared. Percent agreement and κ statistics were calculated for ICD- and IP-coded CLABSIs.

RESULTS

Among 465 intensive care unit patients, most were white (89.9%), males (52.0%), aged 58.7 ± 17.1 years. Only 3 new CLABSIs were reported during the study period: 2 by ICD and IP in 2008, 1 by ICD in 2012, and 0 by either method in 2015. The percent agreement (99.6%) and κ (0.799) represent excellent agreement. Central line usage was similar for each time period.

DISCUSSION

The number of CLABSIs decreased over time; however, the findings were limited, and a larger sample over a longer period is needed to draw conclusions about the influence of legislative changes. One discrepancy was observed between the 2 reporting methods, which is consistent with other studies. More research is needed to understand the complexity of provider coding practices and changes in central line use (eg, duration, type, location) over time.

Validity of administrative data in identifying complex surgical site infections from a population-based cohort after primary hip and knee arthroplasty in Alberta, Canada

BACKGROUND

Surgical site infections (SSIs) are a substantial burden to healthcare systems in North America. Administrative data is one method though which these may be identified, but the accuracy of using such data is uncertain.

METHODS

We followed a population-based cohort of patients who received primary hip/knee arthroplasty in Alberta, Canada, for whom a comprehensive Infection Prevention and Control (IPC) prospective surveillance methodology was used to track SSIs. Patients were also followed using International Classification of Diseases, Tenth Revision (ICD-10) codes. We assessed the sensitivity/specificity and positive/negative predictive values of ICD-10 codes compared to IPC surveillance.

RESULTS

Between April 1, 2012, and March 31, 2015, 24,512 people received hip/knee arthroplasty. Of these, 258 (1.05%) had a complex SSI found by IPC surveillance. Sensitivity and specificity of ICD-10 codes in identifying complex SSIs after hip/knee arthroplasty were 85.3% (95% confidence interval [CI] 80.3%-89.4%) and 99.5% (95% CI 99.4%-99.6%), respectively. Positive and negative predictive values were 63.6% (95% CI 58.3%-68.7%) and 99.8% (95% CI 99.8%-99.9%), respectively.

DISCUSSION

Administrative data have reasonable testing characteristics for identifying complex SSIs after arthroplasty. For centers without prospective surveillance programs, this could be useful in identifying hospitals with frequent complex SSIs after arthroplasty.

CONCLUSIONS

A comprehensive IPC surveillance program is superior at detecting SSIs after arthroplasty.

The role of the built environment and private rooms for reducing central line-associated bloodstream infections

Private hospital rooms are believed to offer some protective effect against hospital-acquired infections, including central line-associated bloodstream infections. Yet a recent meta-analysis found the evidence-base to be lacking from a policy perspective. We sought to determine whether private rooms were associated with a lower risk of central-line infections. We examined the discharge records of more than one million inpatients from 335 Texas hospitals to determine patients that stayed in private rooms. Patients who stayed in bay rooms had 64 percent more central line infections than patients who stayed in private rooms. Even after adjusting for relevant covariates, patients assigned to bay rooms had a 21 percent greater relative risk of a central line infection (p = 0.005), compared with patients assigned to private rooms. At the hospital level, a 10% increase in private rooms was associated with an 8.6% decrease in central line infections (p<0.001), regardless of individual patients' room assignment. This study demonstrates and validates the use of private rooms as a structural measure and independent predictor of hospital quality.

Centers for medicare and medicaid services hospital-acquired conditions policy for central line-associated bloodstream infection (CLABSI) and catheter-associated urinary tract infection (CAUTI) shows minimal impact on hospital reimbursement

OBJECTIVE

In 2008, the Centers for Medicare and Medicaid Services (CMS) stopped reimbursing for hospital-acquired conditions (HACs) not present on admission (POA). We sought to understand why this policy did not impact central line-associated bloodstream infection (CLABSI) and catheter-associated urinary tract infection (CAUTI) trends.

DESIGN

Retrospective cohort study.

SETTING

Acute-care hospitals in the United States.ParticipantsFee-for-service Medicare patients discharged January 1, 2007, through December 31, 2011.

METHODS

Using inpatient Medicare claims data, we analyzed billing practices before and after the HAC policy was implemented, including the use and POA designation of codes for CLABSI or CAUTI. For the 3-year period following policy implementation, we determined the impact on diagnosis-related groups (DRG) determining reimbursement as well as hospital characteristics associated with the reimbursement impact.

RESULTS

During the study period, 65,205,607 Medicare fee-for-service hospitalizations occurred at 3,291 acute-care, nonfederal US hospitals. Based on coding, CLABSI and CAUTI affected 0.23% and 0.06% of these hospitalizations, respectively. In addition, following the HAC policy, 82% of the CLABSI codes and 91% of the CAUTI codes were marked POA, which represented a large increase in the use of this designation. Finally, for the small numbers of CLABSI and CAUTI coded as not POA, financial impacts were detected on only 0.4% of the hospitalizations with a CLABSI code and 5.7% with a CAUTI code.

CONCLUSIONS

Part of the reason the HAC policy did not have its intended impact is that billing codes for CLABSI and CAUTI were rarely used, were commonly listed as POA in the postpolicy period, and infrequently impacted hospital reimbursement.

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.

Low Completeness of Bacteraemia Registration in the Danish National Patient Registry

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

Health care-associated infections among critically ill children in the US, 2007-2012

BACKGROUND

Health care-associated infections (HAIs) are harmful and costly and can result in substantial morbidity for hospitalized children; however, little is known about national trends in HAIs in neonatal and pediatric populations. Our objective was to determine the incidence of HAIs among a large sample of hospitals in the United States caring for critically ill children from 2007 to 2012.

METHODS

In this cohort study, we included NICUs and PICUs located in hospitals reporting data to the Centers for Disease Control and Prevention's National Healthcare Safety Network for central line-associated bloodstream infections (CLABSIs), ventilator-associated pneumonias, and catheter-associated urinary tract infections. We used a time-series design to evaluate changes in HAI rates.

RESULTS

A total of 173 US hospitals provided data from NICUs, and 64 provided data from PICUs. From 2007 to 2012, rates of CLABSIs decreased in NICUs from 4.9 to 1.5 per 1000 central-line days (incidence rate ratio (IRR) per quarter = 0.96, 95% confidence interval 0.94-0.97) and in PICUs from 4.7 to 1.0 per 1000 central-line days (IRR per quarter = 0.96 [0.94-0.98]). Rates of ventilator-associated pneumonias decreased in NICUs from 1.6 to 0.6 per 1000 ventilator days (IRR per quarter = 0.97 [0.93-0.99]) and PICUs from 1.9 to 0.7 per 1000 ventilator-days (IRR per quarter = 0.95 [0.92-0.98]). Rates of catheter-associated urinary tract infections did not change significantly in PICUs.

CONCLUSIONS

Between 2007 and 2012 there were substantial reductions in HAIs among hospitalized neonates and children.

Validity of ICD-9-CM codes for the identification of complications related to central venous catheterization

Two complications of central venous catheterization (CVC), iatrogenic pneumothorax and central line-associated bloodstream infection (CLABSI), have dedicated International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes. Despite increasing use of ICD-9-CM codes for research and pay-for-performance purposes, their validity for detecting complications of CVC has not been established. Complications of CVCs placed between July 2010 and December 2011 were identified by ICD-9-CM codes in discharge records from a single hospital and compared with those revealed by medical record abstraction. The ICD-9-CM code for iatrogenic pneumothorax had a sensitivity of 66.7%, specificity of 100%, positive predictive value (PPV) of 100%, and negative predictive value (NPV) of 99.5%. The ICD-9-CM codes for CLABSI had a sensitivity of 33.3%, specificity of 99.0%, PPV of 28.6%, and NPV of 99.2%. The low sensitivity and variable PPV of ICD-9-CM codes for detection of complications of CVC raise concerns about their use for research or pay-for-performance purposes.