Accuracy Of Administrative Data Research Articles

Validity of Italian administrative healthcare data in describing the real-world utilization of infusive antineoplastic drugs: the study case of rituximab use in patients treated at the University Hospital of Siena for onco-haematological indications.

Italian administrative healthcare databases are frequently used for studies on real-world drug utilization. However, there is currently a lack of evidence on the accuracy of administrative data in describing the use of infusive antineoplastics. In this study, we used rituximab as a case study to investigate the validity of the regional administrative healthcare database of Tuscany (RAD) in describing the utilization of infusive antineoplastics. We identified patients aged 18 years or older who had received ≥1 rituximab administration between 2011 and 2014 in the onco-haematology ward of the University Hospital of Siena. We retrieved this information from the Hospital Pharmacy Database (HPD-UHS) and linked the person-level information to RAD. Patients who had received ≥1dispensing of rituximab, single administration episodes, and patients treated for non-Hodgkin Lymphoma (nHL) or Chronic Lymphocytic Leukemia (CLL) were identified in RAD and validated using HPD-UHS as the reference standard. We identified the indications of use using algorithms based on diagnostic codes (ICD9CM codes, nHL=200*, 202*; CLL=204.1). We tested 22 algorithms of different complexity for each indication of use and calculated sensitivity and positive predictive value (PPV), with 95% confidence intervals (95%CI), as measures of validity. According to HPD-UHS, 307 patients received rituximab for nHL (N=174), CLL (N=21), or other unspecified indications (N=112) in the onco-haematology ward of the University Hospital of Siena. We identified 295 rituximab users in RAD (sensitivity=96.1%), but PPV could not be assessed due to missing information in RAD on dispensing hospital wards. We identified individual rituximab administration episodes with sensitivity=78.6% [95%CI: 76.4-80.6] and PPV=87.6% [95%CI: 86.1-89.2]. Sensitivity of algorithms tested for identifying nHL and CLL ranged from 87.7% to 91.9% for nHL and from 52.4% to 82.7% for CLL. PPV ranged from 64.7% to 66.1% for nHL and from 32.4% to 37.5% for CLL. Our findings suggest that RAD is a very sensitive source of information for identifying patients who received rituximab for onco-haematological indications. Single administration episodes were identified with good-to-high accuracy. Patients receiving rituximab for nHL were identified with high sensitivity and acceptable PPV, while the validity for CLL was suboptimal.

Measuring and explaining the changing nature of work - The Linked Personnel Panel enriched with administrative employment data (LPP-ADIAB)

In a rapidly changing world of work, policy-makers have new questions for evidence-based policy advice that go beyond what administrative data can provide. For example, the focus is on flexible forms of work or the permeability of the boundary between work and leisure time. Survey data can provide a good basis for answering these questions. In many respects, however, they do not come close to the representativeness and measuring accuracy of administrative data.
 Here the combination of survey data with administrative data can help to get the best out of two systems. On the one hand, objective information about the population is available and on the other hand, the survey data contain much more detailed information on a variety of topics.
 With the Linked Personnel Panel (LPP), the administrative data are used both for drawing a sample that is as undistorted as possible and for adjusting weighting factors. The LPP is a linked operational and employee data set that collects information on personnel instruments, corporate culture and work quality. Supplemented with the administrative data as LPP-ADIAB, we have advised policy makers on the relevance of home office for companies and employees. We know from the data of the employees whether their activity is suitable for home office. In the first step, these data are linked with the job information of the same employees from the administrative data. In the second step, we can transfer these values to the employees in Germany and estimate an undepleted homeoffice potential.
 With the help of the LPP-ADIAB, we were able to show the limited effect of a right to homeoffice and thus provided a comprehensive picture in a discussion led by a small group of supporters.

Abstract 191: Identification of Adults With Congenital Heart Disease From Administrative Data

Introduction: With improved management, there are now more adults with congenital heart disease (ACHD) than children. However, long-term survival with moderate or severe ACHD remains limited, and there is relatively sparse literature on the intensity or quality of end-of-life care for these patients. The goal of this analysis is to examine the accuracy of administrative data for identifying patients who died with ACHD to facilitate study of care provided near the end of life. Methods: We created a list of ICD-9 and ICD-10 codes representing ACHD of moderate or great complexity. We performed a search for these codes in the electronic health record (EHR) of adults who received care 2010-2016 within our healthcare system. We used state death records to identify which of these patients died during the same timeframe. Manual EHR review was completed to evaluate performance of this search strategy. Identified patients were also compared to a list of patients seen in our ACHD clinic and known to have died during 2010-2016. Results: Using ICD data, 121 patients were identified, of which 66 actually had the moderate or greater complexity ACHD conditions by EHR review (positive predictive value, 0.55; 95% confidence interval 0.45, 0.63). EHR review confirmed 12 patients with Eisenmenger Syndrome, for which there is no specific ICD code. “Cyanosis+other” did not identify any of these, “VSD+other” (ventricular septal defect) identified 6, and there were 6 whose only ACHD code was VSD. Of the remaining 55 patients, 24 had ACHD not on the targeted list, largely due to coding error. In addition, despite being coded as having ACHD, 31 patients had no identified ACHD on EHR review. These misidentifications were attributed to coding error for 15 patients. Another 11 patients (35%) had acquired VSD due to myocardial infarction or endocarditis, for which there is no separate ICD code. Codes with the highest degree of error, incorrect more than 50% of the time, were those for congenitally corrected transposition, endocardial cushion defect, and hypoplastic left heart syndrome. The list of known deceased clinic patients included 21 with ACHD of interest. Only 1 of these was not identified by the ICD search, yielding a sensitivity for our list of ICD codes in this small sample of 0.95 (0.77,0.99). Conclusion: Use of administrative data to identify patients with ACHD of moderate or great complexity who have died had good sensitivity but suboptimal positive predictive value. Strategies to improve accuracy can be employed. Excluding patients who have codes for myocardial infarction or endocarditis in addition to VSD and using “VSD+other” as an additional proxy for Eisenmenger Syndrome are two examples. Administrative data is not ideal for identification of patients with ACHD of moderate or great complexity who have died, and manual EHR review is necessary to confirm these diagnoses.

Abstract TP187: Low Sensitivity and High Specificity of Administrative Data for Transient Ischemic Attacks in the Emergency Department

Introduction: Accurate surveillance of TIA is important for monitoring disease burden and evaluating temporal trends. Passive surveillance is a time and cost-effective method to identify TIA using administrative data. Although TIA is primarily managed in the emergency department (ED) without admission to hospital, prior administrative data validation studies have mainly evaluated inpatient databases. We determined the validity of the ICD-10 codes to identify TIA in an ED administrative database. Methods: The study population was obtained from two ongoing studies on the diagnosis of TIA and minor stroke versus stroke mimic. Stroke mimics were actively recruited. Patients enrolled between December 1st 2013 and October 30th 2015 with an ED visit were included in the current study. ED discharge diagnoses were obtained from the National Ambulatory Care Reporting System database. We determined the sensitivity, specificity, and positive predictive value (PPV) of the ICD-10 TIA codes by using two reference standards: 1) the ED chart abstraction and 2) the 90-day final diagnosis, both adjudicated by stroke neurologists. Different case definition algorithms were tested. Results: We included 417 patients. ED adjudication showed 163 (39.1%) TIA, 155 (37.2%) ischemic stroke, and 99 (23.7%) stroke mimics. The most restrictive algorithm, defined as a TIA code in the main position had the lowest sensitivity (36.8%), but highest specificity (92.5%) and PPV (76.0%). The most inclusive algorithm, defined as a TIA code in any positions with and without query prefix had the highest sensitivity (63.8%), but lowest specificity (81.5%) and PPV (68.9%). Comparing the final 90-day diagnosis with coding showed similar results. Conclusions: TIA can be identified with high specificity, but low sensitivity from ED discharge diagnoses. By including patients with stroke mimics, we determined both the false positive and negative rates, allowing for the calculation of sensitivity and specificity. We used two reference standards to verify the accuracy of administrative data. Future studies are necessary to understand the reasons for the low sensitivity of administrative data for TIA and whether the miscoded patients are systematically different from the accurately coded ones.

Accuracy of Administrative Data in Comparison to Administered Antimicrobial Doses for Hospitalized Children

Accuracy of Administrative Data in Comparison to Administered Antimicrobial Doses for Hospitalized Children

Comparing Administrative and Clinical Data for Central Line Associated Blood Stream Infections in Pediatric Intensive Care Unit and Pediatric Cardiothoracic Intensive Care Unit

Central line associated bloodstream infections (CLABSIs) are a frequent source of health complication for patients of all ages, including for patients in the pediatric intensive care unit (PICU) and Pediatric Cardiothoracic Intensive Care Unit (PCTU). Many hospitals, including the University of Michigan Health System, currently use the International Classification of Disease (ICD) coding system when coding for CLABSI.The purpose of this study was to determine the accuracy of coding for CLABSI infections with ICD-9CM codes in PICU and PCTU patients. A retrospective chart review was conducted for 75 PICU and PCTU patients with 90 events of hospital acquired central line infections at the University of Michigan Health System (from 2007-2011). The different variables examined in the chart review included the type of central line the patient had, the duration of the stay of the line, the type of organism infecting the patient, and the treatment the patient received. A review was conducted to assess if patients had received the proper ICD-9CM code for their hospital acquired infection. In addition, each patient chart was searched using Electronic Medical Record Search Engine to determine if any phrases that commonly referred to hospital acquired CLABSIs were present in their charts. Our review found that in most CLABSI cases the hospital’s administrative data diagnosis using ICD-9CM coding systems did not code for the CLABSI. Our results indicate a low sensitivity of 32% in the PICU and an even lower sensitivity of 12% in the PCTU. Using these results, we can conclude that the ICD-9CM coding system cannot be used for accurately defining hospital acquired CLABSIs in administrative data. With the new use of the ICD-10CM coding system, further research is needed to assess the effects of the ICD-10CM coding system on the accuracy of administrative data.

Operative intervention rates for acute diverticulitis: a multicentre state-wide study.

Acute colonic diverticulitis is placing an increasing strain on our health care resources. Measurement of the problem is difficult at a regional level, yet essential to improve and optimize treatment of this condition. Therefore, we aimed to use Australian state-level administrative data to determine the current practice and outcomes in major metropolitan hospitals. Coding algorithms designed to increase the yield and accuracy of administrative data were used to find emergency admissions from the Victorian Admitted Episodes Dataset. Eight tertiary referral centres with specialist colorectal services from 2009 to 2013 were studied. Key metrics including the operative intervention rate were measured. There were 2829 emergency admissions for acute diverticulitis across 4 years in eight hospitals, with 724 being complicated. The emergency operative intervention rate was 10.4%, with a third of admissions for complicated diverticulitis having an operation. Hartmann's procedure was the most commonly performed emergency operation, accounting for 72% of resections. Patient characteristics were consistent across the hospitals, including a median length of stay of 3 and 6 days for uncomplicated and complicated diverticulitis, respectively. Hartmann's procedure is currently the most common emergency operation for acute complicated diverticulitis in Victorian metropolitan hospitals. Our practice and outcomes can be measured meaningfully using administrative data.

PRACTITIONER APPLICATION: Thirty-Day Readmission Rates as a Measure of Quality: Causes of Readmission After Orthopedic Surgeries and Accuracy of Administrative Data

PRACTITIONER APPLICATION: Thirty-Day Readmission Rates as a Measure of Quality: Causes of Readmission After Orthopedic Surgeries and Accuracy of Administrative Data

Thirty-Day Readmission Rates as a Measure of Quality: Causes of Readmission After Orthopedic Surgeries and Accuracy of Administrative Data

The rate of unplanned 30-day readmissions to the hospital after discharge is being used as a marker to compare the quality of care across hospitals and to set reimbursement levels for care. While the readmission rate can be reported using administrative data, the accuracy of these data is variable, and defining which readmissions are unplanned and preventable is often difficult. The purpose of this study was to review readmissions to a single orthopedic hospital to identify the causes for readmission and, in particular, which readmissions are planned versus unplanned. Using that hospital's administrative database of patient records from 2007 to 2009, we identified all patients who were readmitted to the hospital within 30 days of a previous hospitalization for a procedure. Readmissions were broadly categorized as planned (a staged or rescheduled procedure or a direct transfer) or unplanned. Unplanned readmissions were defined as either surgical or nonsurgical complications (medical conditions not directly related to the procedure). Almost 30 percent of readmissions were planned. Of the unplanned readmissions, close to 60 percent were triggered by an infection or a concern for an infection. Nonsurgical complications accounted for 18.2 percent of unplanned readmissions. This study highlights the importance of careful data collection and abstraction when calculating early readmission rates. Preventing surgical site infection and better coordinating care between orthopedic surgeons and primary care and medical subspecialty physicians may significantly reduce readmission rates.

Nonpayment for Urinary Tract Infections Based on Faulty Data

Using claims data to support public quality reporting and nonpayment for hospital-acquired conditions may be flawed, at least when it comes to catheter-associated urinary tract infections. Such infections were most likely being underreported by hospital coders, and as a result, the “accuracy of reporting from the data set is suspect,” according to Dr. Jennifer A. Meddings of the University of Michigan, Ann Arbor, and her colleagues. Since the data on catheter-associated urinary tract infections (CAUTIs) are thought to be faulty, quality comparisons and incentive or disincentive payments based on it are also in doubt, they added. The researchers studied state discharge data from 96 acute care hospitals in Michigan, comparing adults discharged in 2007 with those discharged in 2009 (Ann. Intern. Med. 2012;157:305-12). There were 767,531 discharges in 2007 and 781,343 in 2009 in the state. Previous studies have found 4.5 hospital-acquired infections per 100 hospitalizations, at least a third in the urinary tract. And yet the authors said that only 0.3% of the Michigan discharges involved a hospital-acquired CAUTI. Only 25 of the 781,343 (0.003%) hospitalizations in 2009 experienced a nonpayment for a CAUTI. (The authors estimated a $132,675 loss for each of those cases.) CAUTIs are being underreported for several reasons, Dr. Meddings said. A large proportion of UTIs are catheter-associated, but a review of medical records shows that they are generally not well-documented by physicians. Nurses tend to be more accurate in reporting the infections and that they are catheter-related, but those notes aren't used by hospital coders for billing. Also, if a coder suspects a UTI occurred after admission, he or she is supposed to contact the health care provider, but that doesn't happen very often. Finally, the Centers for Medicare & Medicaid Services does not require coders to list all hospital-acquired conditions in claims data. CMS has been using claims data since 2008 to determine whether it should withhold payment for certain hospital-acquired conditions and as the basis for public reporting on the Hospital Compare website. There has been a long history of concern about the value and accuracy of administrative data. In this study, it is especially concerning because CAUTIs are thought to be the most common health care infection, said Dr. Sean M. Berenholtz, associate professor of anesthesia, critical care medicine, and health policy and management at Johns Hopkins University, Baltimore. There is not a lot of morbidity and mortality associated with CAUTIs, but because they are the most common, the fact that they are being underrepresented on coding is concerning, he said. It's still not clear how to increase the accuracy of diagnosis in coding or how to get hospitals to report such complications more often. The incentives and disincentives for poor performance have good meaning and intention, but the system is so complex that it isn't translating into improvements in quality, Dr. Berenholtz said. It's been almost 5 years since CMS has stopped paying for various complications, but there has not been much evidence to show that the nonpayment has had an impact on quality of care, he said. The study argues for the need to ensure that accurate data are used for financial incentives or disincentives, he added. The study is probably generalizable, Dr. Berenholtz said. It has not been replicated yet, he added, but there is no obvious reason that Michigan would be intentionally underreporting infections. He called the researchers exceedingly talented and well-established. The study was funded by the Blue Cross Blue Shield of Michigan Foundation. The authors reported that they had no conflict of interest.

Accuracy of Administrative Data in Identifying Ulcerative Colitis Patients Presenting with Acute Flare and Undergoing Colectomy: 2010 Presidential Poster

Accuracy of Administrative Data in Identifying Ulcerative Colitis Patients Presenting with Acute Flare and Undergoing Colectomy: 2010 Presidential Poster

Accuracy of Administrative Data (AD) Verses Expert Review (ER) To Identify and Classify Selected Care-Associated Adverse Outcomes

BACKGROUND/OBJECTIVES: In November of 2005, Florida became the first state to publicly report performance outcome measures on hospitals, including the rates of care-associated post-operative sepsis (POS) and decubitus ulcer (DU). Those rates are determined by the state authority having jurisdiction (AHJ) using Agency for Healthcare Research and Quality (AHRQ) data that provides 30 measures reflecting hospital volumes, mortality rates, utilization rates and patient safety. This study compares the POS and DU findings for one of our system hospitals as initially reported by the AHJ with our own ER.

Accuracy of administrative data in trauma: splenic injuries as an example.

Accurate data are needed to evaluate clinical outcomes, therapeutic modalities, and quality of care in trauma. Administrative data, usually used for billing, have been used to evaluate performance and assess therapy in other medical specialties. This study was performed to determine whether administrative databases are accurate in the recording of information about trauma patients with splenic injuries. Patients who had blunt splenic injuries were identified using a state trauma registry. The medical records of those patients were reviewed. The data collected by chart review were compared with data in the statewide administrative database of patients who had splenic injuries at the same four Level I and II trauma centers in the same 5-year period. Age, sex, admission date, and hospital were matched to assure comparison of the identical cohort. chi2 analysis was used to compare dichotomous data and Student's t test continuous data. The administrative database identified 641 and the trauma registry identified 529 patients with a diagnosis of splenic injury. A total of 401 patients were found in both databases. Of these, 120 (22.7%) patients were not recorded in the administrative database. Injury Severity Score was underreported by the administrative database (25.74 +/- 14.7 vs. 19.52 +/- 11, p < 0.0001). The administrative database underreported orthopedic, chest, and head injuries (317 vs. 215, 325 vs. 228, and 234 vs. 155, respectively; all p < 0.0001). Use of abdominal computed tomographic scan and diagnostic peritoneal lavage were also underreported (260 vs. 56 and 104 vs.17, both p < 0.0001). The number of operations on the spleen and number of orthopedic procedures were underreported (259 vs. 225, p < 0.014 and 147 vs. 94, p < 0.0001). Complications were markedly underreported by the administrative database (200 vs. 47, p < 0.0001) This study shows that administrative data lack accuracy in the recording of associated injuries, injury severity, diagnostics, procedures, and outcomes data in patients with splenic injuries. Whether these data should be used to evaluate treatment modalities or quality of care in trauma is questionable.

Accuracy of administrative data in trauma

Accuracy of administrative data in trauma

Accuracy of administrative data in trauma

Background: Accurate data are needed to evaluate outcomes, therapeutics, and quality of care. This study assesses the accuracy of administrative databases in recording information about trauma patients. Methods: Patients with thoracic aorta injury were identified with a state trauma registry, and the medical records were reviewed. Data collected were compared to administrative data on patients with thoracic aorta injuries, at the same hospitals in the same time period. Results: Fifteen patients (16.3%) with thoracic aorta injury were not recorded in the administrative database, and 23 patients (18.7%) were misdiagnosed. Ninety-one patients were found in both data sources. The administrative database significantly ( P < .05) underrecorded abdominal injuries (50 vs 35), orthopedic injuries (117 vs 75), and chest injuries (77 vs 48). The number of aortograms (78 vs 8), type of operative procedures (use of graft; 70 vs 30), use of bypass (35 vs 16), and complications (77 vs 33) were underreported ( P < .05). The Injury Severity Score was underestimated by the administrative database (38.65 ± 12.41 vs 25.66 ± 9.53; P < .05). Conclusions: Administrative data lack accuracy in the recording of associated injury, injury severity, diagnostic, and procedural data. Whether these data should be used to evaluate treatment or quality of care in trauma is questionable. (Surgery 1999:126:191-7.)