Figure - available via license: Creative Commons Attribution-NonCommercial 3.0 Unported
Content may be subject to copyright.
Validity of Various Case Definitions for Identifying Cases of Acute Hemorrhagic Stroke, SAH, and ICH
Source publication
Purpose
The performance of the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes for identifying acute hemorrhagic stroke in Taiwan’s National Health Insurance claims database has not been assessed. This study aimed to construct and validate the case definitions for acute hemorrhagic stroke based on I...
Contexts in source publication
Context 1
... 98.6% (95% CI, 97.6-99.3%). Table 1 shows the validity of various case definitions to identify cases of acute hemorrhagic stroke as a whole or cases of SAH or ICH separately. The case definition of using the ICD-10-CM code of I60 or I61 as the primary diagnosis to identify acute hemorrhagic stroke yielded a PPV and sensitivity of 98.2% and 93.1%, respectively. ...Context 2
... The case definitions tested in this study showed different tradeoffs between the PPV and sensitivity, as shown in a study that investigated various case definitions for stroke in Korean NHI claims data. 13 In this regard, the study results presented in Table 1 may help researchers choose an ICD-10-CM coding definition that suits their research purposes to identify an acute hemorrhagic stroke. For studies where a validated method for identifying all patients with acute hemorrhagic stroke, SAH, or ICH is crucial, the case definitions that use ICD-10-CM codes in the first three diagnosis fields to identify the corresponding condition are highly recommended. ...Similar publications
Background: Heart failure (HF) is a complex syndrome associated with significant morbidity and healthcare costs. Electronic health records (EHRs) are widely used to identify patients with HF and other phenotypes. Despite widespread use of EHRs for phenotype algorithm development, it is unclear if the characteristics of identified populations mirror...
The research aim is to assess the two-week program containing complex dolphin assisted therapy activities intended for psychosocial and neurosensory functions of children with psychiatric and behavioral disorders. The study involved 140 school-aged (7-17 years, SD=3.36) subjects: 26 girls and 114 boys. The subjects were diagnosed with mental, behav...
Background: Sustainable development goal 3 (SDG 3) includes an ambitious target of reducing the global maternal mortality rate (MMR) to less than 70 per 100 000 births by 2030. To reach this target, countries need an accurate picture of the levels and causes of maternal deaths. A standardization of the cause of death attribution will improve interp...
Objectives
To study trends of utilization, outcomes, and cost of care in patients undergoing undergoing transcatheter mitral valve repair (TMVr) with end‐stage renal disease (ESRD).
Background
Renal disease has been known to be a predictor of poor outcome in patients with mitral valve disease. Outcome data for patients with ESRD undergoing TMVr re...
Background-The accuracy of identifying sarcoidosis cases in electronic medical records (EMR) using diagnostic codes is unknown.
Methods-To estimate the statistical performance of using diagnostic codes, ICD-9 and ICD-10, to identify sarcoidosis cases in EMR, we searched the San Francisco and Palo Alto Veterans Affairs (VA) medical centers EMR and r...
Citations
... We extracted the diagnosis codes, based on the International Classification of Diseases (ICD), from the hospitalization records in the study database for all patients. Stroke was defined as ICD-9 codes 430, 431, 433, and 434 [24,25] or ICD-10 codes I60, I61, and I63 [26,27]. The secondary outcome measures were hospital admission for stroke within 30, 90, and 365 days, stroke severity, and stroke location. ...
... Incident mortality and CV events were assessed during the follow-up period between 1 day after the index event until 1 year of follow-up. The primary outcome was defined using ensuing ICD-10 codes to create a 3-point major adverse CV events (3P-MACE) composite of death, acute myocardial infarction (I21, I46, I24.9) (12), and acute CV events (G45-46, I61-63) (13,14). Secondary outcomes included the individual components of the 3P-MACE as well as acute heart failure (I50.21, ...
Abstract
Context
Romosozumab, a monoclonal sclerostin antibody, is a recently approved highly potent anti-osteoporotic agent with osteoanabolic properties. Clinical use of Romosozumab is hindered by the fear of adverse cardiovascular (CV) events raised following the pivotal ARCH-trial.
Objective
To assess real-world CV safety of romosozumab vs. alternative osteoanabolic therapies used for treatment of severe osteoporosis.
Design
Data was obtained from TriNetX, a global federated health research network including real-time electronic medical records from 113 healthcare organizations with a total of 136,460,930 patients across 16 countries at time of analysis. Inclusion criteria were age ≥ 40 years, a diagnosis of osteoporosis and prescription of romosozumab or a PTH analog (teriparatide/abaloparatide) during 8.2019-8.2022. 1:1 propensity score matched cohorts were created using demographic variables, comorbidities, and medications. Kaplan–Meier analysis was used to estimate the probability of the outcomes.
Outcomes
Incident 3-point major adverse CV event or death (3P-MACE) during 1-year of follow-up after the initial prescription.
Results
5,626 and 15,986 patients met the criteria for romosozumab and PTH analog cohorts, respectively, with 5,610 patients per group following propensity score matching. 3P-MACE was significantly less frequent in the romosozumab vs. PTH analog cohort (158 vs 211 patients with an outcome, p=0.003) with reductions in the individual components of the composite outcome: myocardial ischemic events (31 vs 58, p=0.003); cerebrovascular events 56 vs 79, p=0.037; deaths (83 vs 104, p=0.099).
Conclusions
In a diverse real-world setting, prescription of romosozumab for osteoporosis is associated with less adverse CV events when compared to PTH analog therapy.
... We estimated the incidence of stroke in HIV-infected patients who had their first emergency visit or hospitalization for stroke events in the period from the date of the initial diagnosis (i.e., index date) to the end of the follow-up period, date of death, or December 31, 2017. By modifying the coding algorithm proposed by Hsieh et al., [31] the incidence of major stroke events is defined as the first occurrence of strokes in this study, including ischemic stroke (ICD-9-CM: 433.xx, 434.xx, and ICD-10-CM: I63), hemorrhagic stroke (ICD-9-CM: 430, 431, and ICD-10-CM: I60, I61) and transient ischemic attack (TIA) (ICD-9-CM: 435.xx, and ICD-10-CM: I60-64, G45) in emergency or inpatient medical records from the first clinical visit for HIV to the end of the observation period [32,33]. A total of 509 patients aged ≥ 18 years and with an incident stroke were identified. ...
Background
The incidence of stroke is increasing among younger people with human immunodeficiency virus (HIV). The burden of stroke has shifted toward the young people living with HIV, particularly in low- and middle-income countries. People infected with herpes zoster (HZ) were more likely to suffer stroke than the general population. However, the association of HZ infection with the incidence of stroke among patients with HIV remains unclear.
Methods
A nested case–control study was conducted with patients with HIV registered in the Taiwan National Health Insurance Research Database in 2000–2017. A total of 509 stroke cases were 1:10 matched to 5090 non-stroke controls on age, sex, and date of first stroke diagnosis. Logistic regression models were used to estimate the odds ratio and 95% confidence intervals (CI) of stroke incidence.
Results
The odds ratio of stroke was significantly higher in the HIV-infected population with HZ (adjusted odds ratio [AOR]: 1.85, 95% CI: 1.42–2.41). A significantly increased AOR of stroke was associated with hypertension (AOR: 3.53, 95% CI: 2.86–4.34), heart disease (AOR: 2.32, 95% CI: 1.54–3.48), chronic kidney disease (AOR: 1.82, 95% CI: 1.16–2.85), hepatitis C virus infection (AOR: 1.49, 95% CI: 1.22–1.83), hyperlipidemia (OR: 1.41, 95% CI: 1.12–1.78), and treatment with protease inhibitors (AOR: 1.33, 95% CI: 1.05–1.69).
Conclusions
Our findings suggest that HZ concurrent with HIV may increase the risk of stroke. The incidence rates of stroke were independent of common risk factors, suggesting strategies for early prevention of HZ infection among people living with HIV.
... The growth of natural language processing (NLP) techniques is fundamental for automatically transforming clinical notes into structured data that can be directly processed using machine learning (ML) algorithms. The utilization of NLP in the clinical domain is increasing with various applications including the identification of biomedical hypotheses from radiology reports [9], nursing documentation [10], and discharge summaries [11]. However, frameworks built on NLP applied to clinical narratives have not been widely used in clinical settings to aid decision support systems or work processes. ...
Acute coronary syndrome (ACS) is a leading cause of mortality and morbidity. Predicting the associated risks of patients with chest pain using electronic health record data can help identify those needing more tailored care. This study proposes the development of a reliable prediction framework to serve as a diagnostic support tool for preventing misdiagnoses among patients with clinical concerns for ACS. Data were collected from an urban, demographically diverse hospital in Detroit, Michigan, for patients presenting to the emergency department (ED) with primary chief complaints of chest pain from January 2017 to August 2020. This study incorporated term frequency-inverse document frequency features from free-text summaries, which contain anecdotal symptom descriptions and are among the first data points provided upon entering the ED. The analysis included 16,096 patients with clinical concerns for ACS and trained three machine learning models, logistic regression, AdaBoost, and linear discriminant analysis, across different data processing stages to predict patients with ACS from non-ACS etiology. The AdaBoost model outperformed the other two models with an accuracy of 94% and an F1-score of 0.943 in predicting ACS on the testing data. This study identified key independent factors from patient demographics, comorbidities, and clinical narrative data that predicted ACS in patients. The prediction framework can serve as a decision-support tool to classify ACS and inform physicians about better ACS risk factors.
... The ICD-9-CM codes for ICH have a sensitivity of 82% and a specificity of 93% as compared with medical record review, 21,22 while the ICD-10-CM code has been validated previously with a positive predictive value of 88.6% and sensitivity of 98.6%. 23 We excluded patients with any diagnosis codes for trauma (ICD-9-CM codes 800-806, 839.1, 839.2, 847.0, 850-854, 900.x, 920.x, 952.2, and 959.0, and ICD-10-CM codes S00-S1). The ICD-9-CM codes were applicable till October 1, 2015, after which the ICD-10-CM diagnosis codes were introduced. ...
... The diagnosis codes for ICH and ischemic stroke, however, have been previously validated to have high specificity and positive predictive value. [21][22][23]28,29 Moreover, ICD-9-CM and ICD-10-CM codes for major diagnoses in these types of administrative datasets have generally been shown to be reliable. 47 Third, we had follow-up data only on patients who were followed up in NY or FL for subsequent care after ICH, and our study hence did not capture ischemic stroke events in patients who received care in other states. ...
Background:
Intracerebral hemorrhage (ICH) is associated with an increased risk of ischemic stroke. Whether there are racial and ethnic disparities in the risk of ischemic stroke after ICH is poorly understood. We therefore aimed to test the hypothesis that non-Hispanic Black and Hispanic ICH patients have a higher risk of ischemic stroke compared with non-Hispanic White ICH patients.
Methods:
We performed a retrospective cohort study using the Healthcare Cost and Utilization Project (HCUP) on all hospitalizations at all nonfederal hospitals in Florida from 2005 to 2018 and New York from 2006 to 2016. Race and ethnicity were coded as a single variable in HCUP. We included patients with an ICH, and without a prior or concomitant diagnosis of ischemic stroke, ascertained using validated International Classification of Diseases-Clinical Modification-9 and 10 diagnosis codes. Using Cox proportional hazard models, we studied the relationship between race and risk of ischemic stroke starting from the time of discharge from ICH hospitalization, after adjustment of demographics and vascular comorbidities.
Results:
We included 91 342 patients with ICH-62% non-Hispanic White, 18% non-Hispanic Black, and 12% Hispanic patients. Non-Hispanic Black and Hispanic patients were younger and had a higher prevalence of cardiovascular comorbidities; however, atrial fibrillation was more prevalent among non-Hispanic White patients. During a median follow-up period of 4.4 years (interquartile range, 1.5-8.1), an incident ischemic stroke occurred in 3377 (6%) non-Hispanic White, 1323 (8%) non-Hispanic Black, and 844 (8%) Hispanic patients. In adjusted Cox models, the risk of an ischemic stroke was significantly higher among non-Hispanic Black patients (hazard ratio, 1.6 [95% CI, 1.5-1.8]) and Hispanic patients (hazard ratio, 1.4 [95% CI, 1.3-1.5]), compared with non-Hispanic White patients. Similar results were obtained in sensitivity analyses when using death as a competing risk and after excluding patients with atrial fibrillation and valvular heart disease.
Conclusions:
In a large heterogeneous cohort of patients with ICH, we found that non-Hispanic Black and Hispanic patients had a significantly higher risk of ischemic stroke compared with non-Hispanic White patients.
... This variation in the validity of primary and secondary diagnoses is in line with previous validation studies of bleeding and other diagnosis codes. 1,3,27,28 The lower accuracy of secondary diagnoses might arise through the definition of primary and secondary diagnoses, as the primary diagnosis has to describe the primary event for each patient contact, while secondary diagnoses are optional. 5 As such, another level of attention might be given to primary diagnoses, which might generate the variation in PPVs for primary and secondary diagnoses. ...
Purpose
The majority of bleeding diagnoses in the Danish National Patient Registry have not been validated despite extensive use in epidemiological research. Therefore, we examined the positive predictive value (PPV) of non-traumatic bleeding diagnoses in the Danish National Patient Registry.
Study Design
Population-based validation study.
Patients and Methods
Based on a manual review of electronic medical records, we estimated the PPV of diagnostic coding (International Classification of Diseases, Tenth Revision (ICD‐10)) for non-traumatic bleeding for all patients ≥65 years of age with any hospital contact in the North Denmark Region during March–December 2019 as registered in the Danish National Patient Registry. We calculated PPVs and associated 95% confidence intervals (CI) for non-traumatic bleeding diagnoses overall and stratified according to primary or secondary diagnosis, and according to major anatomical sites.
Results
A total of 907 electronic medical records were available for review. The population mean age was 79.33 years (standard deviation (SD)=7.73) and 57.6% were males. Primary bleeding diagnoses accounted for 766 of the records and 141 were secondary bleeding diagnoses. The overall PPV for bleeding diagnoses was 94.0% (95% CI: 92.3–95.4). The PPV was 98.7% (95% CI: 97.6–99.3) for the primary diagnoses and 68.8% (95% CI: 60.7–75.9) for the secondary diagnoses. When stratified according to subgroups of major anatomical sites, the PPVs ranged between 94.1% and 100% for the primary diagnoses, and between 53.8% and 100% for secondary diagnoses.
Conclusion
The overall validity of non-traumatic bleeding diagnoses in the Danish National Patient Registry is high and considered acceptable for epidemiological research. However, PPVs were substantially higher for primary than for secondary diagnosis.
... The CA_SID is an administrative database with diagnoses based upon ICD-10 codes. While coding errors may exist, the sensitivity and positive predictive value of utilizing ICD-10 codes for identifying acute hemorrhagic strokes is 99% and 89%, respectively [18]. Patients in this sample were selected based upon admission date, so individual admitted in December 2020 but discharged after December 2020 were not included, leading to a loss of some patients. ...
Objectives:
Patients with severe intracerebral hemorrhage (ICH) often suffer from impaired capacity and rely on surrogates for decision-making. Restrictions on visitors within healthcare facilities during the pandemic may have impacted care and disposition for patient with ICH. We investigated outcomes of ICH patients during the COVID-19 pandemic compared to a pre-pandemic period.
Materials and methods:
We conducted a retrospective review of ICH patients from two sources: (1) University of Rochester Get With the Guidelines database and (2) the California State Inpatient Database (SID). Patients were divided into 2019-2020 pre-pandemic and 2020 pandemic groups. We compared mortality, discharge, and comfort care/hospice. Using single-center data, we compared 30-day readmissions and follow-up functional status.
Results:
The single-center cohort included 230 patients (n = 122 pre-pandemic, n = 108 pandemic group), and the California SID included 17,534 patients (n = 10,537 pre-pandemic, n = 6,997 pandemic group). Inpatient mortality was no different before or during the pandemic in either cohort. Length of stay was unchanged. During the pandemic, more patients were discharged to hospice in the California SID (8.4% vs. 5.9%, p<0.001). Use of comfort care was similar before and during the pandemic in the single center data. Survivors in both datasets were more likely to be discharged home vs. facility during the pandemic. Thirty-day readmissions and follow-up functional status in the single-center cohort were similar between groups.
Conclusions:
Using a large database, we identified more ICH patients discharged to hospice during the COVID-19 pandemic and, among survivors, more patients were discharged home rather than healthcare facility discharge during the pandemic.
... Hemorrhagic stroke included SAH and ICH. ICD-10 codes for SAH (I60), ICH (I61), ischemic hemorrhage (I63), and unspecified strokes (I64) have been validated previously [32][33][34][35][36]. Follow-up was conducted from 1992-1999, to 31 December 2019. ...
(1) Background: We investigated whether weight changes affect the association between smoking cessation and stroke risk; (2) Methods: Overall, 719,040 males were categorized into eight groups according to smoking status (sustained smokers, non-smokers, long-term quitters (quit > 4 years), and recent quitters (quit < 4 years)) and post-cessation weight change (−5 kg, −5.0 to 0.1 kg, maintainers, 0.1–5.0 kg, and >5.0 kg). The hazard ratios (HR) and 95% confidence intervals (CI) for incident total, ischemic, and hemorrhagic strokes, including subarachnoid and intracerebral hemorrhage, were calculated using Cox proportional hazard models; (3) Results: We detected 38,730 strokes (median follow-up, 25.7 years), including 30,609 ischemic and 9055 hemorrhagic strokes. For recent quitters with a >5.0 kg or 0.1–5.0 kg weight increase, maintainers, or those who lost 0.1–5 kg, the multivariable HR for total stroke was 0.73 (95% CI, 0.67–0.79), 0.78 (95% CI, 0.74–0.82), 0.77 (95% CI, 0.69–0.85), 0.84 (95% CI, 0.77–0.90), and 1.06 (95% CI, 0.92–1.23), respectively, compared with that of sustained smokers; (4) Conclusions: Comparable patterns were obtained for stroke subtypes. Thus, we strongly recommend quitting smoking, as weight gain after quitting smoking does not alter the stroke-related benefits.
... Patients with stroke (ischemic stroke, nontraumatic intracerebral hemorrhage, and nontraumatic subarachnoid hemorrhage) were the positive control and were identified using standard, validated ICD-10-CM codes. [18][19][20] Measurements Any inpatient or emergency department encounter for seizure, defined as having the discharge diagnosis ICD-10-CM code G40.x in any diagnostic position, was the primary outcome. However, we excluded diagnosis codes for idiopathic epilepsies (G40.0x and G40.3x), and other conditions that are unlikely a result of PRES (G40.Ax, G40.Bx, G40.42, G40.81x, G40.83x, G40.82x, G40.89x). ...
Objective:
Patients with posterior reversible encephalopathy syndrome (PRES) can develop seizures during the acute phase. We sought to determine the long-term risk of seizure after PRES.
Methods:
We performed a retrospective cohort study using statewide all-payer claims data from 2016-2018 from nonfederal hospitals in 11 US states. Adults admitted with PRES were compared to adults admitted with stroke, an acute cerebrovascular disorder associated with long-term risk of seizure. The primary outcome was seizure diagnosed during an emergency room visit or hospital admission after the index hospitalization. The secondary outcome was status epilepticus. Diagnoses were determined using previously validated ICD-10-CM codes. Patients with seizure diagnoses before or during the index admission were excluded. We used Cox regression to evaluate the association of PRES with seizure, adjusting for demographics and potential confounders.
Results:
We identified 2095 patients hospitalized with PRES and 341,809 with stroke. Median follow-up was 0.9 years (IQR, 0.3-1.7) in the PRES group and 1.0 years (IQR, 0.4-1.8) in the stroke group. Crude seizure incidence per 100 person-years was 9.5 after PRES and 2.5 after stroke. After adjustment for demographics and comorbidities, patients with PRES had a higher risk of seizure than patients with stroke (HR, 2.9; 95% CI, 2.6-3.4). Results were unchanged in a sensitivity analysis that applied a two-week washout period to mitigate detection bias. A similar relationship was observed for the secondary outcome of status epilepticus.
Interpretation:
PRES was associated with an increased long-term risk of subsequent acute care utilization for seizure compared to stroke.
... We secondarily assessed the rates of seizure and stroke diagnoses and prescription of antiseizure medication. We determined stroke and seizure diagnoses through the presence of diagnosis codes [17][18][19][20]. We report whether a patient was prescribed an antiseizure medication using CTC codes, excluding medications most likely to be used for sedation including benzodiazepines (Additional file 1: Table S1). ...
Background
Seizures, strokes, and intracranial hemorrhage are common and feared complications in children receiving extracorporeal membrane oxygenation (ECMO) support. Researchers and clinicians have proposed and deployed methods for monitoring and detecting neurologic injury, but best practices are unknown. We sought to characterize clinicians’ approach to electroencephalography (EEG) and brain imaging modalities in children supported by ECMO.
Methods
We performed a retrospective observational cohort study among US Children’s Hospitals participating in the Pediatric Health Information System (PHIS) from 2016 to 2021. We identified hospitalizations containing ECMO support. We stratified these admissions by pediatric, neonatal, cardiac surgery, and non-cardiac surgery. We characterized the frequency of EEG, cranial ultrasound, brain computed tomography (CT), magnetic resonance imaging (MRI), and transcranial Doppler during ECMO hospitalizations. We reported key diagnoses (stroke and seizures) and the prescription of antiseizure medication. To assess hospital variation, we created multilevel logistic regression models.
Results
We identified 8746 ECMO hospitalizations. Nearly all children under 1 year of age (5389/5582) received a cranial ultrasound. Sixty-two percent of the cohort received an EEG, and use increased from 2016 to 2021 (52–72% of hospitalizations). There was marked variation between hospitals in rates of EEG use. Rates of antiseizure medication use (37% of hospitalizations) and seizure diagnoses (20% of hospitalizations) were similar across hospitals, including high and low EEG utilization hospitals. Overall, 37% of the cohort received a CT and 36% received an MRI (46% of neonatal patients). Stroke diagnoses (16% of hospitalizations) were similar between high- and low-MRI utilization hospitals (15% vs 17%, respectively). Transcranial Doppler (TCD) was performed in just 8% of hospitalizations, and 77% of the patients who received a TCD were cared for at one of five centers.
Conclusions
In this cohort of children at high risk of neurologic injury, there was significant variation in the approach to EEG and neuroimaging in children on ECMO. Despite the variation in monitoring and imaging, diagnoses of seizures and strokes were similar across hospitals. Future work needs to identify a management strategy that appropriately screens and monitors this high-risk population without overuse of resource-intensive modalities.
