Validating administrative data in stroke research.

Venous thromboembolism (VTE) is a common medical condition manifested as deep vein thrombosis (DVT) or pulmonary embolism (PE). Few data exist on the total economic burden of DVT and PE. To (1) quantify the economic burden of DVT and PE in direct medical costs and utilization and (2) determine the rates of hospital readmission for DVT and PE. Hospital claims containing DVT or PE as a primary or secondary discharge diagnosis during the period February 1998 through June 2004 were identified by retrospective analysis using the Integrated Health Care Information Services (IHCIS) National Managed Care Database. For the cost analysis, we included patients that had been enrolled in a health care plan for a minimum of 30 days prior to and 365 days following the DVT or PE hospitalization. For the readmission analysis, patients were required to have a minimum length of stay of 3 days and a pre-enrollment of 365 days. We quantified the cost burden to the health plan by examining annual DVT- and PE-related payments made by the health plan to providers for inpatient and outpatient care. Of 5 million plus discharges in the database with dates of service between February 1, 1998, and June 30, 2004, 32,193 (0.64%) had DVT or PE as a primary discharge diagnosis, and 26,159 (0.52%) had DVT or PE as a secondary discharge diagnosis. After application of the inclusion and exclusion criteria, there were 5,348 patients with a primary discharge diagnosis of DVT and 4,593 patients with a secondary discharge diagnosis of DVT. For PE, 2,984 patients had a primary discharge diagnosis, and 1,119 had a secondary discharge diagnosis. The hospital readmission rates within 1 year for the combined diagnoses (DVT or PE) were 5.3% for primary and 14.3% for secondary diagnoses; 44.3% of the PE readmissions occurred within the first 30 days. Within 90 days, 50.7% of DVT readmissions and 58.6% of PE readmissions occurred. Regarding cost for a primary diagnosis, the average total annual provider payments made by a health plan were $10,804 for DVT and $16,644 for PE. For secondary diagnoses, the average total annual costs were $7,594 for DVT and $13,018 for PE. The mean hospital cost per readmission for a recurrent DVT ($11,862) was higher than the mean cost for the initial hospitalization ($9,805, P=0.006), but the mean cost per PE readmission ($14,722) was similar to the mean cost for the initial hospitalization ($14,146, P =0.38). The economic burden of DVT and PE in direct medical cost is large, due not only to the initial hospitalization event, but also to the high rate of hospital readmission (5%-14%), over half of which occurs within 90 days.

Objectives To gain a better understanding of the characteristics of patients with a hospital encounter for major depressive disorder (MDD) and evaluate associated hospital resource utilization, hospital charges and costs, and hospital re-encounters. Methods Adult patients with a hospital encounter (i.e. emergency department [ED] visit only or inpatient admission) with MDD as the primary discharge diagnosis (index event) during July 2018‒March 2019 were selected from the Premier Healthcare Database. Patient characteristics, hospital resource utilization, and hospital charges and costs were evaluated during index events. During a 12-month follow-up, hospital re-encounters (MDD-related and all-cause ED visit only or inpatient readmissions) were examined. Results The study population included 77,178 patients with an index hospital encounter (ED visit only: 49.9%; inpatient admission: 50.1%) for MDD. The most common secondary mental health-related diagnosis was suicidal ideation/behavior, which was recorded in 51.8% of patients. The mean age was 38.2 years, 53.0% were female, and 72.1% were Caucasian. Among patients with an ED visit only, the mean index hospital charges and costs were $3,608 and $639, respectively. Among those with inpatient admissions, the mean length of stay was 4.9 days, and the mean index hospital charges and costs were $17,107 and $6,095, respectively. During the 12-month follow-up, 13.3% of patients in the overall study population had an MDD-related hospital re-encounter (primary or secondary discharge diagnosis code indicating MDD); nearly one-third (31.3%) occurred within 30 days post-discharge. During the follow-up, 28.1% had an all-cause hospital re-encounter with 29.7% having occurred within 30 days post-discharge. Limitations Due to constraints of the Premier Healthcare Database, healthcare resource utilization and costs outside of the hospital could not be evaluated. Conclusions Patients with a hospital encounter for MDD are relatively young, commonly have suicidal ideation/behavior, utilize substantial hospital resources, and have a high risk for a hospital re-encounter in the 30 days post-discharge.

Introduction: About 15-40% patients with inflammatory bowel diseases (IBD) are obese. Obesity has been variably associated with aggressive phenotype, complicated disease course and inferior response to biologic therapy in patients with immune-mediated inflammatory diseases, including IBD. We conducted a nationally representative cohort study to estimate and compare burden, costs and causes for hospitalization in obese vs. non-obese patients with IBD. Methods: We constructed a retrospective cohort study using the Nationwide Readmissions Database 2013, to identify patients who had been hospitalized at least once between January-June 2013, with a primary or secondary discharge diagnosis of IBD and had a follow-up ≥6 months (or until death). Patients were classified as obese based on ICD-9 codes (which have specificity and modest sensitivity); all other patients were classified as non-obese. We estimated burden of hospitalization as total days spent in hospital per month and costs per month (median with interquartile range, IQR). Causes of hospitalization were classified based on primary discharge diagnoses. Results: We identified 42,285 patients with IBD admitted during the exposure period, of which 12.4% were obese. Obese patients spent more days in hospital annually (vs. non-obese, median: 8 vs. 5 days) and per month (0.82 [IQR 0.38-1.71] vs. 0.55 [IQR 0.29-1.10], p < 0.01), without significant differences in in-hospital mortality (3% vs. 2%). Obese patients with IBD had higher hospitalization-related costs annually (vs. young, median: $17232 vs. $10818, p < 0.01) and per month ($1827 vs. $1142), though rates of IBD-related endoscopic or surgical procedures were only marginally higher (47% vs. 43%). There was considerable variability in burden of hospitalization, and obese patients with IBD in the highest quintile spent 3.3 days/month in hospital (IQR, 2.5-4.6) (vs. non-obese, 2.2 days/month), with costs of $5747/month. Obese patients were more likely to be hospitalized due to serious infections (vs. younger: 12% vs. 10%) and cardiopulmonary complications (15% vs. 10%) causes, rather than IBD complications (18% vs. 26%). Older patients were significantly more likely to have preventable admissions (28% vs. 16%). Conclusion: Obesity is associated with higher burden and costs of hospitalization in patients with IBD, and may be considered as potential modifier in decreasing burden of hospitalization.

To determine whether hospital discharge diagnoses can be used for intensive care unit (ICU)-related activities. Comparison between the diagnoses coded by physicians at the time of ICU admission and those diagnoses coded by medical records personnel. University hospital adult surgical ICU. Consecutive ICU admissions (n = 622). None. The ICU admission and hospital discharge codes were compared in two ways. Initially, each discharge code was subtracted from the corresponding ICU admission code. There was no difference in 150 (24%) cases. In 216 (35%) patients, the codes differed by +/-10. In 221 (36%) instances, the codes differed by >200. The secondary discharge diagnoses were also compared with the ICU admission diagnoses. In 56 patients, the ICU admission diagnosis was one of the secondary diagnoses. The second comparison involved having two physicians not associated with the study examine each pair of codes to determine if the two diagnoses were medically different. Review of the codes by physicians not involved in the study found that in 318 (48%) patients, the two diagnoses were not different, i.e., the codes were either the same or the codes were so similar as not to functionally change the actual diagnosis. The primary discharge diagnosis often failed to reflect the reason for ICU admission, making it impossible to consistently establish the reason for ICU admission from the discharge data. The reason for ICU admission was also frequently not included among the secondary discharge diagnoses. Administrative data are therefore not useful for ICU quality management and other functions. Intensivists need to establish their own databases.

Discharge diagnostic data from hospital administrative databases are often used to inform decisions relating to a variety of vital applications. These may include the allocation of resources, quality-of-care assessments, clinical research and formulation of healthcare policy. Accurately coded and reliably captured patient discharge data are of paramount importance for any hospital and health system to function efficiently. To retrospectively examine the reliability of the International Classification of Diseases version 10 (ICD-10) discharge coding in Red Cross War Memorial Children's Hospital (RCWMCH)'s administrative database for primary and secondary discharge diagnoses, and to formulate recommendations for improvement to the current system. This study was a retrospective folder review of 450 patient admissions to the short-stay and general paediatric wards at RCWMCH between 1 August 2013 and 1 September 2014. The principal investigator (PI) completed ICD-10 discharge coding for each admission and compared it with the corresponding admission data captured for each patient in the Clinicom (Siemens Medical Solutions, Germany) health information system. Agreement comparison was done to 4- and 3-character ICD-10 code specificity. Of the initial 450 randomly selected folders, 396 (88%) were analysed during the folder review process. The median number of total diagnoses (primary diagnosis plus secondary diagnoses) coded by the PI folder review was 3, with a distribution of 1 - 10 (interquartile range (IQR) 2 - 4). The median number of total diagnoses coded in Clinicom was 1, with a distribution of 1 - 3 (IQR 1 - 1). Agreement of primary diagnosis coding to 4 characters was 26.3%, with slight improvement to 34.3% when assessed to 3 characters. Agreement of secondary diagnoses to 4 characters was 14.9%, and 27.7% when assessed to 3 characters. Reliability of administrative ICD-10 discharge data from RCWMCH is poor. Inadequacies regarding the employment of dedicated and/or adequately trained coding personnel may significantly contribute to the problem and should be addressed.

The diagnosis and treatment of catatonia via telemedicine: A case report and proposed diagnostic criteria

The International Coalition of Medicines Regulatory Authorities recently announced a collaboration to address the expanded use of Real-World Data (RWD) to support regulatory decision-making.1 Drug safety and comparative effectiveness studies using electronic health records and medical claims databases are being used increasingly because compared with randomized clinical trials (RCTs) they can offer larger populations and be completed more quickly at a fraction of the cost. While database studies deliver results with impressive efficiency, this efficiency can come at a scientific cost. Instead of designing data collection to meet the scientific needs of a particular study, database studies analyze (secondary) data not designed to support scientific research. Consequently, database studies are more vulnerable than RCTs to biases arising from inaccurate data. A challenge in using database studies to guide health policy decisions, therefore, is ensuring that database studies are not only efficient, but also accurate. One potential source of error in database studies is outcome misclassification. Diagnoses recorded in health encounter databases are influenced by factors such as variation in patient care-seeking behavior, access to care, quality of care, clinical challenges of differential diagnosis, and diagnoses that can indicate suspected or rule-out conditions. To address uncertainties in identifying health outcomes, database researchers often identify outcomes using combinations of diagnoses, procedures, treatments, medical specialties, place of service, and other factors in the form of an algorithm (also called a computable phenotype or operational definition).2, 3 Owing to limitations in source data, however, even the best algorithms are unlikely to identify every case of interest (false negative errors) and can identify as cases individuals who do not have the outcome (false positive errors). Because errors in identifying outcomes can distort study results, it has been proposed that "validation of algorithms is essential to avoid misclassification bias."4 A recent FDA Guidance for RWD advises that "FDA expects validation of the outcome variable to minimize outcome misclassification."5 Regulators impose high scientific standards and even though validation is not widely practiced in epidemiology, validation is routinely requested for studies submitted to regulators. Many public health researchers receive little formal training in validation, however, and validation studies are not covered thoroughly in texts.6 Consequently, many researchers are guided by convention and validation tends to be conducted perfunctorily. As regulators contemplate wider use of database studies and despite routine use of validation, the impact of outcome misclassification bias in these studies is rarely assessed and remains largely unknown. Analytic methods to adjust results for misclassification have been described but have not been widely adopted.7-17 In this Commentary, we highlight shortcomings in popular validation methods and strategies for reducing misclassification bias. Basic concepts of validation have been described recently for epidemiology4, 6 and specifically as practiced in pharmacoepidemiology.3 In brief, outcome validation entails comparing outcome classification in the database with outcome classification in a gold (or reference) standard (e.g., medical records or disease registry data). Because the gold standard is typically more expensive and time- consuming to obtain than the information in the database used for the main study, the gold standard is usually used for a sample of the study population to assess the performance of the case identifying algorithm.2, 3, 16 Algorithm performance is assessed by parameters such as positive predictive value (PPV), negative predictive value (NPV), sensitivity, and specificity. The conventional strategy is to report the proportion of potential cases that was confirmed by the gold standard, or PPV (Table 1).2-4, 16-18 A high PPV means the outcome of interest is confirmed for a large proportion of potential cases and is taken to suggest the algorithm is fit for purpose, with a PPV of 70%–80% or greater considered to indicate a high-performing algorithm.3, 16, 18 For a couple of reasons, the conventional practice of reporting an effect estimate and a PPV does not avoid or minimize bias from outcome misclassification. First, a PPV indicates the proportion of people identified as cases who are noncases (false positive errors), but does not address cases the algorithm failed to identify (false negative errors). While the importance of each kind error varies by setting, each type of error can bias study results.7-17 Second, to understand the impact of misclassification on study results requires parameters that estimate false positive (specificity or PPV) and false negative (sensitivity or NPV) errors in each of the comparison groups.7-17 Differences between comparison groups is readily seen as important in the context of confounding bias, but less easily recognized for misclassification bias. Even in the presence of a high PPV, small differences between comparison groups in misclassification errors can introduce important bias and lead to incorrect conclusions.9, 16 A simulation study demonstrated that a result showing a negative association (RR = 0.87 (95% CI 0.76, 0.99)) with an outcome PPV of 93% could be the result of outcome misclassification masking an even stronger association (RR = 0.70 (95% CI 0.60, 0.81)).16 Alternatively, the negative result could be entirely spurious due to misclassification bias creating an association when in fact there was none (RR = 1.0 (95% CI 0.88, 1.13)).16 Simply put, reporting an effect estimate accompanied by a high PPV does not indicate the presence, direction or magnitude of outcome misclassification bias. FDA recommends including a quantitative bias analysis to demonstrate whether and how outcome misclassification might affect study results. Using validation to inform bias analyses has implications for the design of validation studies.6 The conventional approach to validation in database studies is to calculate one PPV, whereas to support bias analysis validation studies would need to stratify by exposure status in calculating a PPV for each comparison group.6-17 Although conceptually straightforward, stratifying validation samples by exposure status would necessitate larger validation samples to maintain precision. Second, unless there are sound reasons to disregard false negative errors, estimates of sensitivity or NPV are also needed.7-17 Quantitative bias analyses are usually presented using estimates of sensitivity and specificity which can be challenging to estimate in database studies. Bias analysis methods have also been developed for predictive values,6-12, 16, 17 but while estimates of PPV are commonplace, estimates of NPV in studies of rare outcomes can require large sample sizes and are reported rarely. It is sometimes thought that PPV is the only bias parameter that can be estimated in database studies.16 Brenner and Gefeller8 developed a simple bias analysis that uses PPVs in each comparison group so is well-suited to correct for outcome misclassification bias in pharmacoepidemiology.8 Using only PPVs requires that sensitivity be nondifferential, however, which is a tenuous assumption rarely supported by data.19-23 They also present an equation for a simple bias analysis that replaces the nondifferentiality assumption with estimates of PPV and sensitivity in each comparison group.8, 17 Estimating sensitivity in a database study is not as straightforward as estimating PPV. To estimate sensitivity, we need as a denominator a sample of true cases. When studying rare outcomes, a random sample of the study population would need to be quite large to identify a useful sample of true cases. By excluding noncases using conditional sampling, however, it may be possible to target a sample that is enriched in people more likely to be cases.24 For instance, one could design a "screening" algorithm aimed at capturing all (or nearly all) of the cases so that it has near-perfect sensitivity, while still being specific enough to exclude many noncases.25 Validating a random sample of people identified by such as screening algorithm could generate sufficient cases to be used to estimate sensitivity of algorithms designed to have high PPVs (Table 1). Suppose one were interested in occurrence of serious acute nonfatal myocardial infarction (AMI). A primary algorithm might require a primary hospital discharge diagnosis of AMI.26, 27 To assess sensitivity of this algorithm in the study population, a more sensitive screening algorithm could be developed by relaxing the requirement that the AMI diagnosis be in the primary position thus including a discharge diagnosis of AMI in any position.28 One might also broaden the list of diagnoses to include diagnoses associated with AMI, such as unstable angina. Such a screening algorithm might identify all AMI cases of interest and include a larger proportion of AMI cases than a simple random sample of the population. One could then obtain gold standard information on AMI diagnosis for a random sample of people identified by the screening algorithm. Because the population identified by the screening algorithm is enriched in AMI cases, a much smaller sample of people who meet the screening algorithm is needed to obtain a desired number of AMI cases than a sample of the entire study population. After outcome classification using the gold standard, the confirmed AMI cases could then serve as the denominator for estimating the sensitivity of the primary algorithm (e.g., a primary discharge diagnosis of AMI). By designing the validation sample to include people from each comparison group, one could estimate values of PPV and sensitivity of the primary algorithm in each comparison group. These bias parameter estimates could then be used in a bias analysis to correct the RR estimate for outcome misclassification bias to the accuracy of the gold standard.8, 17 Conventional practice in safety and effectiveness studies using validation is to report an effect estimate along with a PPV to assess algorithm performance. A problem with this approach is that we cannot infer from a PPV the impact of outcome misclassification errors on study results. A high PPV does not mean outcome misclassification bias is negligible, and a low PPV does not mean that misclassification bias is important. A preferable approach is to use validation in RWD to inform quantitative bias analysis that corrects effect estimates for misclassification bias to the accuracy of the gold standard. Such a strategy requires larger validation samples and additional resources to complete database studies but would not substantially affect the efficiency advantage of database studies compared with RCTs. Validation for RWD studies is common only for studies submitted to regulators mainly because regulators request it. The proposed approach of using validation to inform bias analysis also is likely to be adopted only if regulators request it. We recommend that regulators strongly consider requesting validation studies to inform quantitative bias analyses. Studies that use quantitative bias analysis to correct results for misclassification bias would strengthen the validity of RWD studies so that they can be used more confidently to support societal and personal health policy decisions. The authors declare no conflict of interest.

Knowledge related to acute heart failure (HF) and its management in "real-world" clinical practice is limited. The authors delineated characteristics and drug therapy for hospitalized HF patients and their association with clinical and economic outcomes. The NDCHealth (National Data Corporation, Atlanta, GA) database, containing billing data from a geographically diverse sample of approximately 300 hospitals, was screened for admissions in 2003 of either a primary or secondary discharge diagnosis of HF. Of 2.5 million admissions screened, 496,534 patients (19.7%) had a primary (131,057) or secondary (365,477) discharge diagnosis of HF. Mean age was 73.1+/-13.9 years, and 55.4% were women. Mean length of hospital stay was 8.7+/-28.6 days, and in-hospital mortality was 7.1%. Mean hospital cost per admission was dollars 18,667. Admissions with HF as a secondary diagnosis were associated with a worse prognosis. HF commonly exists in hospitalized patients and is associated with significant morbidity, mortality, and economic burden, irrespective of whether it is the primary diagnosis or a secondary comorbidity.

Objectives: This study evaluated the frequency of hospital readmissions for venous thromboembolism (VTE) and the associated costs and length of stay (LOS) among acute medically ill patients in the US using a real-world claims database analysis.Methods: Patients (≥40 years of age) at risk of VTE due to hospitalization for acute medical illnesses, based on primary hospital discharge diagnosis codes, were identified from the MarketScan databases between July 1, 2011 and March 31, 2015. Patients were required to have continuous insurance enrollment in the 6 months prior to initial (index) hospitalizations (baseline period) and in the 6 months after hospital discharge (follow-up period). The proportions of patients with VTE-related (diagnosis at any position) and VTE as primary diagnosis hospital readmissions during the follow-up period were evaluated. The associated costs and LOS for such readmissions were also determined, as well as time to VTE-related readmissions.Results: Of the study population (n = 12,785; mean age = 68.3 years), most were hospitalized primarily for infectious diseases (35.2%), followed by respiratory diseases (27.9%), cancer (15.7%), heart failure (11.8%), ischemic stroke (8.1%), and rheumatic diseases (1.4%). Of the overall study population, 2.1% (n = 268) had a VTE-related hospital readmission in the 6 months following discharge of their index hospitalization, of which 36.6% (n = 98) were for a primary diagnosis of VTE. Approximately 25.4% of the VTE-related hospital readmissions occurred within the first 30 days of discharge and 58.2% within 90 days. The mean cost for a hospital readmission with a primary diagnosis of VTE was $18,681 (mean LOS = 5.0 days); for readmissions with a primary diagnosis of DVT and PE, mean costs were $14,719 and $23,305, respectively.Conclusions: Among this study population of patients hospitalized for acute medical illnesses, some experienced a VTE event requiring re-hospitalization, with 25% occurring within the first 30 days after hospital discharge.

National Burden of Emergency Department Care for Sickle Disease: Impact of Age, Insurance Status, Income, Hospital Type and Location On Subsequent Hospital Admission.

A troponin study on patients with ischemic stroke, intracerebral hemorrhage and subarachnoid hemorrhage: Type II myocardial infarction is significantly associated with stroke severity, discharge disposition and mortality

Purpose: IBD is a chronic inflammatory disease characterized by periods of relapses and remissions. Hospitalizations for exacerbations of IBD are common. Readmission has been deemed a marker of poor quality of care. The aim of this study was to better understand and characterize the rates of unplanned readmissions in IBD patients at 30 days, 90 days, 180 days and 1 yr and to identify the reasons for the unplanned readmissions. Methods: We retrospectively reviewed the electronic database of our institution (3 large urban academic hospitals) to identify consecutive patients admitted between Jan 2009 to Sept 2010 with a primary discharge diagnosis of either UC or CD. Index admission was defined as first unplanned admission with primary discharge diagnosis of either UC or CD during this period. Readmission was defined as unplanned admission (due to any cause) occurring within 30 days, 90 days, 180 days or 1 year from the index admission. Data collected included demographics, type of IBD, length of stay, and primary discharge diagnosis for both index and readmission. Results: A total of 558 pts with IBD accounted for 1,056 admissions to the health system within this time period. Of these a total of 209 pts had an index admission with primary discharge diagnosis of either UC or CD. The demographics of the cohort were-mean age 37.5 yrs, 115 females, 149 whites, 150 had CD and 59 had UC. There were a total of 160 unplanned readmissions during the study period. The unplanned readmission rates were - 8.6% (18/209) at 30 days, 18% (37/209) at 90 days, 30.6% (64/209) at 180 days and 36.4% at 1 year. The mean length of stay (LOS) of all index admissions was 7.5 days and mean LOS of all readmissions was 7 days. Among the 76 pts readmitted at least once at 1 year, 38 were females, 47 were whites and 56 pts had CD. At 1 year, 33 pts were readmitted at least twice and 17 were readmitted more than 2 times. Eighty-five of the 160 readmissions had a primary discharge diagnosis of IBD (either CD or UC), while the rest of the 75 had other diagnoses at discharge. Infection was the most common other discharge diagnosis of readmissions (28/160), followed by postoperative complications (10/160), abdominal pain (5/160), bowel obstruction (4/160), pancreatitis (4/160), cholecystitis (4/160). Conclusion: The unplanned readmission rate in IBD pts is high, with more than one third of pts being readmitted at least once within 1 yr of discharge. The most common reasons for readmissions were IBD exacerbations and infections. Future large studies are needed to define national rates of readmission and risk factors for readmissions in IBD pts. This has the potential to better understand and possibly alter the standard of care. Disclosure: Dr. Gary R. Lichtenstein - Consultant: Abbott Corporation, Consultant, Research: Alaven, Research: Bristol-Myers Squibb, Consultant, Research: Centocor Orthobiotech, Consultant: Elan, Consultant, Research: Ferring, Consultant: Hospira, Consultant: Meda Pharmaceuticals, Consultant: Millenium Pharmaceuticals, Consultant: Pfizer Pharmaceuticals, Consultant, Research: Proctor and Gamble, Consultant, Research: Prometheus Laboratories, Inc., Consultant, Research: Salix Pharmaceuticals, Consultant: Santarus, Consultant: Schering-Plough Corporation, Consultant, Research: Shire Pharmaceuticals, Consultant: Takeda, Consultant, Research: UCB, Consultant, Research: Warner Chilcotte, Consultant: Wyeth.

Background and PurposeThe claims data of the Korean National Health Insurance (NHI) system can be useful in stroke research. The aim of this study was to validate the accuracy of hospital discharge data used for NHI claims in identifying acute stroke and use of thrombolytic therapy.MethodsThe hospital discharge data of 1,811 patients with stroke-related diagnosis codes were obtained from Jeju National University Hospital (JNUH) and Seoul Medical Center (SMC). Three algorithms were tested to identify discharges with acute stroke [ischemic stroke (IS), intracranial hemorrhage (ICH), or subarachnoid hemorrhage (SAH)]: 1) all diagnosis codes up to nine positions, 2) one primary diagnosis and one secondary diagnosis, and 3) only one primary diagnosis code. Reviews of medical records were considered the gold standards.ResultsOverall, the degree of agreement (κ) was higher for algorithms 1 and 2 than for algorithm 3, and the sensitivity and specificity of the first two algorithms for IS and SAH were both >90%, with almost perfect agreement (κ=0.83-0.84) in the JNUH data set. Regarding ICH, only algorithm 1 yielded an almost perfect agreement (κ=0.82). In the SMC data set, almost perfect agreement was found for both ICH and SAH in all three algorithms. In contrast, the three algorithms yielded a range of agreement levels, though all substantial, for IS. Almost perfect agreement was obtained for use of thrombolytic therapy in both data sets (κ=0.91-0.99).ConclusionsDischarge with hemorrhagic stroke and use of thrombolytic therapy were identified with high reliability in administrative discharge data. A substantial level of agreement was also obtained for IS, despite variation between the algorithms and data sets.

Predictive value of stroke and transient ischemic attack discharge diagnoses in The Danish National Registry of Patients

Discharge ICD-9-CM (International Classification of Diseases, 9th Revision, Clinical Modification) codes have been used to identify patients with acute stroke for epidemiological, quality of care, and cost studies. The aim of this study was to determine if the accuracy of the primary ICD-9-CM codes for ischemic stroke is improved by modifier codes and how specific codes reflect stroke subtype diagnoses. Available hospital charts for all patients discharged from a single hospital between May 1995 and June 1997 with ICD-9-CM codes 433 (occlusion and stenosis of precerebral arteries), 434 (occlusion of cerebral arteries), or 436 (acute but ill-defined cerebrovascular disease) listed in the first position were reviewed. The primary discharge diagnosis was verified, and a presumed stroke subtype was assigned on the basis of information provided in the medical record. Charts were available for 175 of the 198 identified patients (88%). Of these, 61% had an acute ischemic stroke (code 433, 4%; 434, 82%; 436, 79%) with the remaining patients having other conditions. Of the 130 patients with a modifier code indicating cerebral infarction, 79% had an acute stroke; of the 45 patients with a modifier code indicating an absence of cerebral infarction, 7% had acute stroke (sensitivity, 0.97; specificity, 0.60). The codes with the highest proportions of ischemic stroke cases were 434.11 (embolic occlusion of cerebral arteries with infarction, 85%), 434.91 (unspecified occlusion of precerebral arteries with infarction, 82%), and 436 (79%), with a combined sensitivity of 0.81 and specificity of 0.90. On review, 73% of patients with code 434.11 had embolic strokes, and 47% of those with code 436 had an identified stroke cause. Of patients with code 434.91, 39% had stroke of uncertain cause, 25% "lacunar," 17% atherothrombosis, and 15% embolism. Despite the use of modifier codes, 15% to 20% of patients with the indicated primary ICD-9-CM codes have conditions other than acute ischemic stroke. Although the proportion of patients with acute stroke increased from 61% to 79% with the use of modifier codes, the inclusion of modifier codes did not have an appreciable effect on the accuracy of the coding if patients with code 433 are excluded. Assignment of presumed ischemic stroke subtype is particularly inaccurate.