Clinical Data Warehouse Research Articles

Prediction of postoperative lung function in lung cancer patients using machine learning models.

: Surgical resection is the standard treatment for early-stage lung cancer. Since postoperative lung function is related to mortality, predicted postoperative lung function is used to determine the treatment modality. The aim of this study was to evaluate the predictive performance of linear regression and machine learning models. We extracted data from the Clinical Data Warehouse (CDW) and developed 3 sets: set Ⅰ, the linear regression model; set Ⅱ, machine learning models omitting the missing data: and set Ⅲ, machine learning models imputing the missing data. Six machine learning models, the least absolute shrinkage and selection operator (LASSO), ridge regression, ElasticNet, Random Forest, eXtreme gradient boosting (XGBoost), and the light gradient boosting machine (LightGBM) were implemented. The forced expiratory volume in one second (FEV1) measured 6 months after surgery was defined as the outcome. Five-fold cross-validation was performed for hyperparameter tuning of the machine learning models. The dataset was split into training and test datasets at a 70:30 ratio. Implementation was done after dataset splitting in set Ⅲ. Predictive performance was evaluated by R2 and mean squared error (MSE) in the 3 sets. A total of 1,487 patients were included in sets Ⅰ and Ⅲ and 896 patients were included in set Ⅱ. In set Ⅰ, the R2 value was 0.27 and in set Ⅱ, LightGBM was the best model with the highest R2 value of 0.5 and the lowest MSE of 154.95. In set Ⅲ, LightGBM was the best model with the highest R2 value of 0.56 and the lowest MSE of 174.07. The LightGBM model showed the best performance in predicting postoperative lung function.

Accuracy of identifying pediatric acute bacterial sinusitis diagnoses in outpatient claims data.

Acute bacterial sinusitis is among the most frequent outpatient infections in children and adolescents and is well suited to study in large healthcare utilization databases, but the validity of International Classification of Diseases, 10th Revision (ICD-10) codes together with antibiotic prescriptions to identify cases of acute bacterial sinusitis has not been established. We aimed to evaluate the validity of ICD-10 codes combined with antibiotic prescriptions to identify new diagnoses of acute bacterial sinusitis among pediatric patients evaluated in the outpatient setting. Children and adolescents aged 17 years and younger with an outpatient diagnosis of acute sinusitis along with an antibiotic prescription from an ambulatory facility affiliated with the Mass General Brigham health system were identified via a clinical data warehouse. Patients were stratified by age (0-5 years, 6-11 years, and 12-17 years), and 50 cases per age group were randomly sampled. Medical records were independently reviewed by two pediatric infectious diseases physicians to assess for the documentation of a clinician-defined diagnosis of acute bacterial sinusitis. Positive predictive values (PPVs) and 95% confidence intervals (CIs) were calculated. A total of 150 patients were included in the final cohort. Frontal, maxillary, and "unspecified" sinuses accounted for 88% of the diagnoses. The positive predictive value of the algorithm to identify clinician-defined diagnoses of acute bacterial sinusitis was 92% (95% CI 87%, 95%). The PPVs were consistent across age strata. ICD-10 codes for acute sinusitis, when paired with a same-day antibiotic prescription, have a high positive predictive value among a cohort of pediatric patients, suggesting that they can be used to study new acute bacterial sinusitis diagnoses with claims.

331 Reusing EHR Phenotyping Algorithms in Practice: Developing the Colorado Diabetes EHR Research Repository (CODER)

OBJECTIVES/GOALS: The rates of computational phenotyping algorithm reuse across health systems are low, leading to a proliferation of algorithms for the same trait. We propose a framework for reusing computational phenotyping algorithms and describe the real-world deployment of this framework for the development of the Colorado Diabetes EHR Research Repository. METHODS/STUDY POPULATION: The novel phenotype reuse framework consists of 4 steps: select algorithms that are appropriate for reuse by assessing whether they are fit for purpose; extend the algorithm to account for changes in data and care practice standards; localize the algorithm to use local database standards and terminologies; optimize the algorithm by applying a data driven approach to achieve the desired local performance. To identify individuals with type 1 diabetes (T1D) or type 2 diabetes (T2D), we selected and implemented T2D algorithms in a cohort of adults with any diabetes or pre-diabetes related diagnosis code, medication, or abnormal glucose-related laboratory test in the clinical data warehouse for UCHealth and the University of Colorado. RESULTS/ANTICIPATED RESULTS: We included a total of 926,290 patients who were identified by initial filters. Patients were more likely to be female (53%), identify as non-Hispanic white (69%) and had a median age of 58 years (IQR: 41, 70). Implementation, extension, localization, & optimization through iterative chart review prioritized high sensitivity for all-cause diabetes and high specificity for T1D and T2D. Of the original cohort, 252,946 (27%) were identified by the all-cause diabetes algorithm. Of these 11,688 were identified as T1D and 135,588 as T2D. After optimization the all-cause diabetes algorithm had 88% sensitivity, 90% specificity, 74% positive predictive value (PPV), and 96% negative predictive value (NPV). Our algorithms for T1D and T2D had high specificity (100% and 99%, respectively) and PPV (100 and 96% respectively). DISCUSSION/SIGNIFICANCE: Developing computational phenotyping algorithms is expensive and time consuming, yet algorithm reuse is low due to a lack of practical approaches for reusing algorithms. We demonstrate application of a novel framework for algorithm reuse, yielding good alignment of algorithm performance with study goals for identifying individuals with diabetes.

CovAID: Identification of factors associated with severe COVID-19 in patients with inflammatory rheumatism or autoimmune diseases.

Autoimmune/inflammatory rheumatic diseases (AIRDs) patients might be at-risk of severe COVID-19. However, whether this is linked to the disease or to its treatment is difficult to determine. This study aimed to identify factors associated with occurrence of severe COVID-19 in AIRD patients and to evaluate whether having an AIRD was associated with increased risk of severe COVID-19 or death. Two databases were analyzed: the EDS (Entrepôt des Données de Santé, Clinical Data Warehouse), including all patients followed in Paris university hospitals and the French multi-center COVID-19 cohort [French rheumatic and musculoskeletal diseases (RMD)]. First, in a combined analysis we compared patients with severe and non-severe COVID-19 to identify factors associated with severity. Then, we performed a propensity matched score case-control study within the EDS database to compare AIRD cases and non-AIRD controls. Among 1,213 patients, 195 (16.1%) experienced severe COVID-19. In multivariate analysis, older age, interstitial lung disease (ILD), arterial hypertension, obesity, sarcoidosis, vasculitis, auto-inflammatory diseases, and treatment with corticosteroids or rituximab were associated with increased risk of severe COVID-19. Among 35,741 COVID-19 patients in EDS, 316 having AIRDs were compared to 1,264 Propensity score-matched controls. AIRD patients had a higher risk of severe COVID-19 [aOR = 1.43 (1.08-1.87), p = 0.01] but analysis restricted to rheumatoid arthritis and spondyloarthritis found no increased risk of severe COVID-19 [aOR = 1.11 (0.68-1.81)]. In this multicenter study, we confirmed that AIRD patients treated with rituximab or corticosteroids and/or having vasculitis, auto-inflammatory disease, and sarcoidosis had increased risk of severe COVID-19. Also, AIRD patients had, overall, an increased risk of severe COVID-19 compares general population.

The effect of telemental versus in-person mental health consults in the emergency department on 30-day utilization and processes of care.

We sought to characterize how telemental health (TMH) versus in-person mental health consults affected 30-day postevaluation utilization outcomes and processes of care in Veterans presenting to the emergency department (ED) and urgent care clinic (UCC) with acute psychiatric complaints. This exploratory retrospective cohort study was conducted in an ED and UCC located in a single Veterans Affairs system. A mental health provider administered TMH via iPad. The primary outcome was a composite of return ED/UCC visits, rehospitalizations, or death within 30 days. The following processes of care were collected during the index visit: changes to home psychiatric medications, admission, involuntary psychiatric hold placement, parenteral benzodiazepine or antipsychotic medication use, and physical restraints or seclusion. Data were abstracted from the Veterans Affairs electronic health record and the Clinical Data Warehouse. Multivariable logistic regression was performed. Adjusted odds ratios (aORs) with their 95% confidence intervals (95% CIs) were reported. Of the 496 Veterans in this analysis, 346 (69.8%) received TMH, and 150 (30.2%) received an in-person mental health evaluation. There was no significant difference in the primary outcome of 30-day return ED/UCC, rehospitalization, or death (aOR 1.47, 95% CI 0.87-2.49) between the TMH and in-person groups. TMH was significantly associated with increased ED/UCC length of stay (aOR 1.46, 95% CI 1.03-2.06) and decreased use of involuntary psychiatric holds (aOR 0.42, 95% CI 0.23-0.75). There were no associations between TMH and the other processes-of-care outcomes. TMH was not significantly associated with the 30-day composite outcome of return ED/UCC visits, rehospitalizations, and death compared with traditional in-person mental health evaluations. TMH was significantly associated with increased ED/UCC length of stay and decreased odds of placing an involuntary psychiatric hold. Future studies are required to confirm these findings and, if confirmed, explore the potential mechanisms for these associations.

Integrating a new knowledge organisation system for monoclonal antibodies for therapeutic use authorised in Europe into HeTOP terminology-ontology server

Monoclonal antibodies (MAs) are increasingly used in the therapeutic arsenal. Clinical Data Warehouses (CDWs) offer unprecedented opportunities for research on real-word data. The objective of this work is to develop a knowledge organization system on MAs for therapeutic use (MATUs) applicable in Europe to query CDWs from a multi-terminology server (HeTOP). After expert consensus, three main health thesauri were selected: the MeSH thesaurus, the National Cancer Institute thesaurus (NCIt) and the SNOMED CT. These thesauri contain 1,723 MAs concepts, but only 99 (5.7 %) are identified as MATUs.The knowledge organisation system proposed in this article is a six-level hierarchical system according to their main therapeutic target. It includes 193 different concepts organised in a cross lingual terminology server, which will allow the inclusion of semantic extensions. Ninety nine (51.3 %) MATUs concepts and 94 (48.7 %) hierarchical concepts composed the knowledge organisation system. Two separates groups (an expert group and a validation group) carried out the selection, creation and validation processes.Queries identify, for unstructured data, 83 out of 99 (83.8 %) MATUs corresponding to 45,262 patients, 347,035 hospital stays and 427,544 health documents, and for structured data, 61 out of 99 (61.6 %) MATUs corresponding to 9,218 patients, 59,643 hospital stays and 104,737 hospital prescriptions. The volume of data in the CDW demonstrated the potential for using these data in clinical research, although not all MATUs are present in the CDW (16 missing for unstructured data and 38 for structured data).The knowledge organisation system proposed here improves the understanding of MATUs, the quality of queries and helps clinical researchers retrieve relevant medical information. The use of this model in CDW allows for the rapid identification of a large number of patients and health documents, either directly by a MATU of interest (e.g. Rituximab) but also by searching for parent concepts (e.g. Anti-CD20 Monoclonal Antibody).

Abstract P4-02-27: Preliminary Attainability Assessment of Real-World Data for Answering Clinical Questions about Breast Cancer Brain Metastases

Abstract Background Brain metastases is one of breast cancer’s leading causes of death, and the incidence is increasing with advances in diagnosis and treatment techniques. Nevertheless, these patients have fewer opportunities to participate in prospective randomized clinical trials because of the design challenges of increasing target population heterogeneity or differentiating the definition of endpoints and how the endpoints are evaluated. Real-world data has the potential that reaches the information at a blinded area of RCT approaches. In recent years, observational studies using real-world data have gotten attention for their capability to enhance an understanding of hard-to-reach RCT approaches. However, the yet incomplete gold standard of the study protocol and the unpredictable ’hidden labor’ of secondary use of clinical data become barriers to clinical researchers. In the study, we demonstrate data attainability assessment focused on clinicians’ clinical unmet needs in breast cancer brain metastases (BCBM) as a preliminary step of the observational study for screening study feasibility on a certain data source. Methods A breast cancer registry based on the clinical data warehouse (CDW) of Samsung Medical Center has been used as the data source (N=45,219, up to 31 Dec 2021). A total of 5 clinical questions (CQ) were presented as the result of the interview of two breast cancer experts about BCBM for the study (Table 1). Additionally, we conducted an attainability assessment for the population size and core variables constrained by a clinical question and a survey of 7 breast cancer clinicians for evaluation of the recognized clinical significance and research methods’ suitability for each CQ. A working group was formed for data attainability tests withal a person with experience in clinical research for at least several years across interdisciplinary areas, including clinical expertise, medical informatics, and epidemiology. For the first step, we declare the operational definition for the study population and core variables in which the research question is inherent conceptually. And appropriate data fields and value sets are figured out from the data set. Data attainability is examined via RWD extraction on the population size and availability of core variables. Result The collected five CQs were mostly about the relationship between brain metastases, systemic metastases, and systemic treatments. Other CQs were related to the relationship between brain metastases incidence and the type of neoadjuvant treatment or patients’ described symptoms. Assessment of the importance of clinical questions aligned well, especially for the questions with higher scores. (r=- · 98) For CQ E, filtered out not to fit the SMC BCR dataset at the early stage of the data attainability screening process due to lack of explicit data field or semantically related value sets for one of the core variables, ‘neurological symptom.’ For CQ A, B, and D, we could confirm enough population size and dataset for conducting the study, but additional logical reconstruction of data elements had needed. For example, we used the ‘clinical subtype’ variable, which is provided from SMC BCR, as pre-processed feature variable using an expert-knowledge-based algorithm. For CQ C, we were able to obtain pertinent data for the study only with direct data field match and explicit code set definition. Conclusion To effectively use CDW, the variables have to be interpreted and defined in a clinically meaningful way with the cooperation of clinicians and data science experts. For observational studies based on RWD, understanding the data source contents and clarifying the research question enough to translate to the data level is essential. The step-by-step protocol presented in the study could be applied to the other clinical researcher using RWD at preliminary feasibility screening. Table 1. List of clinical questions from expert interviews Citation Format: Min-Jeong Kim, Hyo Jung Kim, Danbee Kang, Hee Kyung Ahn, Soo-Yong Shin, Seri Park, Juhee Cho, Yeon Hee Park. Preliminary Attainability Assessment of Real-World Data for Answering Clinical Questions about Breast Cancer Brain Metastases [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P4-02-27.

Weight-Based Bisphosphonate Administration for Multiple Myeloma Patients and the Risks of Skeletal Complications

High-dose bisphosphonate for multiple myeloma patients might elevate risks of skeletal complications earlier than general expectations. This study aims to find incidences of atypical femoral fracture (AFF) and medication-related osteonecrosis of the jaw (MRONJ), elucidate their risk factors, and suggest cut-off values for the safer dosing of high-dose bisphosphonate treatment. By using the clinical data warehouse of a single institute, retrospective cohort data of multiple myeloma-diagnosed patients with high-dose bisphosphonate (pamidronate or zoledronate) treatment from 2009 to 2019 was extracted. Among 644 patients, the incidence of prominent AFF requiring surgery was 0.93% (6/644) and MRONJ was diagnosed in 11.8% (76/644) of the study population. For both AFF and MRONJ, the total potency-weighted sum of total dose per body weight (OR = 1.010, p = 0.005) were significant on logistic regression. Cutoffs of the potency-weighted total dose (mg) per body weight (kg) for AFF and MRONJ were 77.00 and 57.70 mg/kg, respectively. After roughly one year of treatment with high-dose zoledronate (about four years for pamidronate), an earlier thorough re-evaluation of skeletal complications should be taken. Body weight adjustments for accumulative dose calculation in terms of permissible dosing should be taken into consideration.

Deep learning approach to detection of colonoscopic information from unstructured reports

BackgroundColorectal cancer is a leading cause of cancer deaths. Several screening tests, such as colonoscopy, can be used to find polyps or colorectal cancer. Colonoscopy reports are often written in unstructured narrative text. The information embedded in the reports can be used for various purposes, including colorectal cancer risk prediction, follow-up recommendation, and quality measurement. However, the availability and accessibility of unstructured text data are still insufficient despite the large amounts of accumulated data. We aimed to develop and apply deep learning-based natural language processing (NLP) methods to detect colonoscopic information.MethodsThis study applied several deep learning-based NLP models to colonoscopy reports. Approximately 280,668 colonoscopy reports were extracted from the clinical data warehouse of Samsung Medical Center. For 5,000 reports, procedural information and colonoscopic findings were manually annotated with 17 labels. We compared the long short-term memory (LSTM) and BioBERT model to select the one with the best performance for colonoscopy reports, which was the bidirectional LSTM with conditional random fields. Then, we applied pre-trained word embedding using large unlabeled data (280,668 reports) to the selected model.ResultsThe NLP model with pre-trained word embedding performed better for most labels than the model with one-hot encoding. The F1 scores for colonoscopic findings were: 0.9564 for lesions, 0.9722 for locations, 0.9809 for shapes, 0.9720 for colors, 0.9862 for sizes, and 0.9717 for numbers.ConclusionsThis study applied deep learning-based clinical NLP models to extract meaningful information from colonoscopy reports. The method in this study achieved promising results that demonstrate it can be applied to various practical purposes.

Do Individualized Patient-Specific Situations Predict the Progression Rate and Fate of Knee Osteoarthritis? Prediction of Knee Osteoarthritis.

Factors affecting the progression rate and fate of osteoarthritis need to be analyzed when considering patient-specific situation. This study aimed to identify the rate of remarkable progression and fate of primary knee osteoarthritis based on patient-specific situations. Between May 2003 and May 2019, 83,280 patients with knee pain were recruited for this study from the clinical data warehouse. Finally, 2492 knees with pain that were followed up for more than one year were analyzed. For analyzing affecting factors, patient-specific information was categorized and classified as demographic, radiologic, social, comorbidity disorders, and surgical intervention data. The degree of contribution of factors to the progression rate and the fate of osteoarthritis was analyzed. Bone mineral density (BMD), Kellgren-Lawrence (K-L) grade, and physical occupational demands were major contributors to the progression rate of osteoarthritis. Hypertension, initial K-L grade, and physical occupational demands were major contributors to the outcome of osteoarthritis. The progression rate and fate of osteoarthritis were mostly affected by the initial K-L grade and physical occupational demands. Patients who underwent surgical intervention for less than five years had the highest proportion of initial K-L grade 2 (49.0%) and occupations with high physical demand (41.3%). In identifying several contributing factors, the initial K-L grade and physical occupational demands were the most important factors. BMD and hypertension were also major contributors to the progression and fate of osteoarthritis, and the degree of contribution was lower compared to the two major factors.

Long-term stroke and major bleeding risk in patients with non-valvular atrial fibrillation: A comparative analysis between non-vitamin K antagonist oral anticoagulants and warfarin using a clinical data warehouse.

Non-vitamin K antagonist oral anticoagulants (NOACs) has been the drug of choice for preventing ischemic stroke in patients with atrial fibrillation (AF) since 2014. Many studies based on claim data revealed that NOACs had comparable effect to warfarin in preventing ischemic stroke with fewer hemorrhagic side effects. We analyzed the difference in clinical outcomes according to the drugs in patients with AF based on the clinical data warehouse (CDW). We extracted data of patients with AF from our hospital's CDW and obtained clinical information including test results. All claim data of the patients were extracted from National Health Insurance Service, and dataset was constructed by combining it with CDW data. Separately, another dataset was constructed with patients who could obtain sufficient clinical information from the CDW. The patients were divided NOAC and warfarin groups. The occurrence of ischemic stroke, intracranial hemorrhage, gastrointestinal bleeding, and death were confirmed as clinical outcome. The factors influencing the risk of clinical outcomes were analyzed. The patients who were diagnosed AF between 2009 and 2020 were included in the dataset construction. In the combined dataset, 858 patients were treated with warfarin, 2,343 patients were treated with NOACs. After the diagnosis of AF, the incidence of ischemic stroke during follow-up was 199 (23.2%) in the warfarin group, 209 (8.9%) in the NOAC group. Intracranial hemorrhage occurred in 70 patients (8.2%) among the warfarin group, 61 (2.6%) of the NOAC group. Gastrointestinal bleeding occurred in 69 patients (8.0%) in the warfarin group, 78 patients (3.3%) in the NOAC group. NOAC's hazard ratio (HR) of ischemic stroke was 0.479 (95% CI 0.39-0.589, p < 0.0001), HR of intracranial hemorrhage was 0.453 (95% CI 0.31-0.664, p < 0.0001), and HR of gastrointestinal bleeding was 0.579 (95% CI 0.406-0.824, p = 0.0024). In the dataset constructed using only CDW, the NOAC group also had a lower risk of ischemic stroke and intracranial hemorrhage than warfarin group. In this CDW based study, NOACs are more effective and safer than warfarin in patients with AF even with long-term follow-up. NOACs should be used to prevent ischemic stroke in patients with AF.

Association of the perfusion index with postoperative acute kidney injury: a retrospective study.

Many studies have examined the risk factors for postoperative acute kidney injury (AKI), but few have focused on intraoperative peripheral perfusion index (PPI) that has recently been shown to be associated with postoperative morbidity and mortality. Therefore, this study aimed to evaluate the relationship between intraoperative PPI and postoperative AKI under the hypothesis that lower intraoperative PPI is associated with AKI occurrence. We retrospectively searched electronic medical records to identify patients who underwent surgery at the general surgery department from May 2021 to November 2021. Patient baseline characteristics, pre- and post-operative laboratory test results, comorbidities, intraoperative vital signs, and discharge profiles were obtained from the Institutional Clinical Data Warehouse and VitalDB. Intraoperative PPI was the primary exposure variable, and the primary outcome was postoperative AKI. Overall, 2,554 patients were identified and 1,586 patients were included in our analysis. According to Kidney Disease Improving Global Outcomes (KDIGO) criteria, postoperative AKI occurred in 123 (7.8%) patients. We found that risks of postoperative AKI increased (odds ratio: 2.00, 95% CI [1.16, 3.44], P = 0.012) when PPI was less than 0.5 for more than 10% of surgery time. Other risk factors for AKI occurrence were male sex, older age, higher American Society of Anesthesiologists physical status, obesity, underlying renal disease, prolonged operation time, transfusion, and emergent operation. Low intraoperative PPI was independently associated with postoperative AKI.

Patient similarity and other artificial intelligence machine learning algorithms in clinical decision aid for shared decision-making in the Prevention of Cardiovascular Toxicity (PACT): a feasibility trial design

BackgroundThe many improvements in cancer therapies have led to an increased number of survivors, which comes with a greater risk of consequent/subsequent cardiovascular disease. Identifying effective management strategies that can mitigate this risk of cardiovascular complications is vital. Therefore, developing computer-driven and personalized clinical decision aid interventions that can provide early detection of patients at risk, stratify that risk, and recommend specific cardio-oncology management guidelines and expert consensus recommendations is critically important.ObjectivesTo assess the feasibility, acceptability, and utility of the use of an artificial intelligence (AI)-powered clinical decision aid tool in shared decision making between the cancer survivor patient and the cardiologist regarding prevention of cardiovascular disease.DesignThis is a single-center, double-arm, open-label, randomized interventional feasibility study. Our cardio-oncology cohort of > 4000 individuals from our Clinical Research Data Warehouse will be queried to identify at least 200 adult cancer survivors who meet the eligibility criteria. Study participants will be randomized into either the Clinical Decision Aid Group (where patients will use the clinical decision aid in addition to current practice) or the Control Group (current practice). The primary endpoint of this study is to assess for each patient encounter whether cardiovascular medications and imaging pursued were consistent with current medical society recommendations. Additionally, the perceptions of using the clinical decision tool will be evaluated based on patient and physician feedback through surveys and focus groups.SummaryThis trial will determine whether a clinical decision aid tool improves cancer survivors’ medication use and imaging surveillance recommendations aligned with current medical guidelines.Trial registrationClinicalTrials.Gov Identifier: NCT05377320

Effects of rifampin coadministration on the pharmacokinetics of digoxin: a real-world data approach.

Digoxin, a cardiac glycoside, is commonly prescribed to treat heart failure and atrial fibrillation. Because digoxin acts as a substrate of P-glycoprotein (P-gp), its blood concentration may be reduced by P-gp inducers such as rifampin. To assess the real-world implications of this drug-drug interaction, a retrospective analysis was carried out on the Clinical Data Warehouse at Seoul National University Hospital between 2012 and 2017. Eleven patients who received both digoxin and rifampin with satisfying the inclusion/exclusion criteria were identified. The Ctrough values of digoxin monotherapy were compared to those of the combination therapy with rifampin. Results demonstrated that the systemic exposure of orally administered digoxin decreased by 40% with the concurrent use of rifampin. Clinicians should be aware of potential drug interactions between digoxin and rifampin, as adjustments to digoxin dosage might be necessary for patients receiving rifampin or other P-gp inducer drugs.

Patient Identification and Tumor Identification Management: Quality Program in a Cancer Multicentric Clinical Data Warehouse.

The Regional Basis of Solid Tumor (RBST), a clinical data warehouse, centralizes information related to cancer patient care in 5 health establishments in 2 French departments. To develop algorithms matching heterogeneous data to "real" patients and "real" tumors with respect to patient identification (PI) and tumor identification (TI). A graph database programed in java Neo4j was used to build the RBST with data from ~20 000 patients. The PI algorithm using the Levenshtein distance was based on the regulatory criteria identifying a patient. A TI algorithm was built on 6 characteristics: tumor location and laterality, date of diagnosis, histology, primary and metastatic status. Given the heterogeneous nature and semantics of the collected data, the creation of repositories (organ, synonym, and histology repositories) was required. The TI algorithm used the Dice coefficient to match tumors. Patients matched if there was complete agreement of the given name, surname, sex, and date/month/year of birth. These parameters were assigned weights of 28%, 28%, 21%, and 23% (with 18% for year, 2.5% for month, and 2.5% for day), respectively. The algorithm had a sensitivity of 99.69% (95% confidence interval [CI] [98.89%, 99.96%]) and a specificity of 100% (95% CI [99.72%, 100%]). The TI algorithm used repositories, weights were assigned to the diagnosis date and associated organ (37.5% and 37.5%, respectively), laterality (16%) histology (5%), and metastatic status (4%). This algorithm had a sensitivity of 71% (95% CI [62.68%, 78.25%]) and a specificity of 100% (95% CI [94.31%, 100%]). The RBST encompasses 2 quality controls: PI and TI. It facilitates the implementation of transversal structuring and assessments of the performance of the provided care.

Modeling and integrating interactions involving the CYP450 enzyme system in a multi-terminology server: Contribution to information extraction from a clinical data warehouse

IntroductionThe cytochrome P450 (CYP450) enzyme system is involved in the metabolism of certain drugs and is responsible for most drug interactions. These interactions result in either an enzymatic inhibition or an enzymatic induction mechanism that has an impact on the therapeutic management of patients. Detecting these drug interactions will allow for better predictability in therapeutic response. Therefore, computerized solutions can represent a valuable help for clinicians in their tasks of detection. ObjectiveThe objective of this study is to provide a structured data-source of interactions involving the CYP450 enzyme system. These interactions are aimed to be integrated in the cross-lingual multi-terminology server HeTOP (Health Terminologies and Ontologies Portal), to support the query processing of the clinical data warehouse (CDW) EDSaN (Entrepôt de Données de Santé Normand). Material and methodsA selection and curation of drug components (DCs) that share a relationship with the CYP450 system was performed from several international data sources. The DCs were linked according to the type of relationship which can be substrate, inhibitor, or inducer. These relationships were then integrated into the HeTOP server. To validate the CYP450 relationships, a semantic query was performed on the CDW, whose search engine is founded on HeTOP data (concepts, terms, and relations). ResultsA total of 776 DCs are associated by a new interaction relationship, integrated in HeTOP, by 14 enzymes. These are CYP450 1A2, 2A6, 2B6, 2C8, 2C9, 2C18, 2C19, 2D6, 2E1, 3A4, 3A7, 11B1,11B2 mitochondrial and P-glycoprotein, constituting a total of 2,088 relationships. A general modelling of cytochromic interactions was performed. From this model, 233,006 queries were processed in less than two hours, demonstrating the usefulness and performance of our CDW implementation. Moreover, they showed that in our university hospital, the concurrent prescription that could cause a cytochromic interaction is Bisoprolol with Amiodarone by enzymatic inhibition for 2,493 patients. DiscussionThe queries submitted to the CDW EDSaN allowed to highlight the most prescribed molecules simultaneously and potentially responsible for cytochromic interactions. In a second step, it would be interesting to evaluate the real clinical impact by looking for possible adverse effects of these interactions in the patients' files. Other computational solutions for cytochromic interactions exist. The impact of CYP450 is particularly important for drugs with narrow therapeutic window (NTW) as they can lead to increased toxicity or therapeutic failure. It is also important to define which drug component is a pro-drug and to considerate the many genetic polymorphisms of patients. ConclusionThe HeTOP server contains a non-negligible number of relationships between drug components and CYP450 from multiple reference sources. These data allow us to query our Clinical Data Warehouse to highlight these cytochromic interactions. It would be interesting in the future to assess the actual clinical impact in hospital reports.

Dental developmental complications in pediatric hematopoietic stem cell transplantation patients: A study using CMC clinical data warehouse.

This study aimed to investigate the prevalence and extent of dental developmental complications in patients who have undergone pediatric hematopoietic stem cell transplantation (SCT) and identify the risk factors. We retrospectively investigated the clinical data warehouse of the Catholic Medical Center information system for identifying patients who: 1) visited the Department of Pediatrics between 2009 and 2019, 2) underwent SCT under the age of 10, and 3) had panoramic radiographs. Thus 153 patients were included in this study. The prevalence and extent of tooth agenesis, microdontia, and root malformation were assessed using panoramic radiographs obtained after SCT, and the risk factors were analyzed using regression analysis. All 153 patients had at least one dental anomaly. When grouped according to the age at initial chemotherapy (≤ 2.5; 2.6-5.0; 5.1-7.5; > 7.5 years), the prevalence of agenesis showed statistically significant differences among the different age groups (P < 0.001). The prevalence of agenesis was highest in the youngest age group. As the initial age at chemotherapy increased, the number of affected teeth per patient decreased for all three anomalies. The location of the affected tooth was also influenced by the age at initial chemotherapy. Regression analysis demonstrated that young age at initial chemotherapy was a risk-increasing factor for tooth agenesis and microdontia. The age at initial chemotherapy may be a critical factor in determining the type, extent, and location of dental complications after SCT. These results suggest that careful dental follow-up and timely treatment are recommended for pediatric patients undergoing SCT.

Healthcare trajectory of critically ill patients with necrotizing soft tissue infections: a multicenter retrospective cohort study using the clinical data warehouse of Greater Paris University Hospitals

BackgroundNecrotizing skin and soft tissue infections (NSTIs) are rare but serious and rapidly progressive infections characterized by necrosis of subcutaneous tissue, fascia and even muscle. The care pathway of patients with NSTIs is poorly understood. A better characterization of the care trajectory of these patients and a better identification of patients at risk of a complicated evolution, requiring prolonged hospitalization, multiple surgical re-interventions, or readmission to the intensive care unit (ICU), is an essential prerequisite to improve their care. The main objective of this study is to obtain large-scale data on the care pathway of these patients. We performed a retrospective multicenter observational cohort study in 13 Great Paris area hospitals, including patients hospitalized between January 1, 2015 and December 31, 2019 in the ICU for surgically confirmed NSTIs.Results170 patients were included. The median duration of stay in ICU and hospital was 8 (3–17) and 37 (14–71) days, respectively. The median time from admission to first surgical debridement was 1 (0–2) day but 69.9% of patients were re-operated with a median of 1 (0–3) additional debridement. Inter-hospital transfer was necessary in 52.4% of patients. 80.2% of patients developed organ failures during the course of ICU stay with 51.8% of patients requiring invasive mechanical ventilation, 77.2% needing vasopressor support and 27.7% renal replacement therapy. In-ICU and in-hospital mortality rates were 21.8% and 28.8%, respectively. There was no significant difference between patients with abdomino-perineal NSTIs (n = 33) and others (n = 137) in terms of in-hospital or ICU mortality. Yet, immunocompromised patients (n = 43) showed significantly higher ICU and in-hospital mortality rates than non-immunocompromised patients (n = 127) (37.2% vs. 16.5%, p = 0.009, and 53.5% vs. 20.5%, p < 0.001). Factors associated with a complicated course were the presence of a polymicrobial infection (adjusted odds ratio [aOR = 3.18 (1.37–7.35); p = 0.007], of a bacteremia [aOR = 3.29 (1.14–9.52); p = 0.028] and a higher SAPS II score [aOR = 1.05 (1.02–1.07); p < 0.0001]. 62.3% of patients were re-hospitalized within 6 months.ConclusionIn this retrospective multicenter study, we showed that patients with NSTI required complex management and are major consumers of care. Two-thirds of them underwent a complicated hospital course, associated with a higher SAPS II score, a polymicrobial NSTI and a bacteremia.

ELECTRONIC HEALTH RECORD SCREENING FOR PERSONS LIVING WITH DEMENTIA IN D-CARE

Abstract Pragmatically identifying persons living with dementia (PLWD) for research is a persistent challenge, reflecting limitations in data elements. As part of the D-CARE study, the recruitment coordinator and analyst collaborated to create a screening algorithm to identify potentially eligible patients. A combination SAS/R script was applied on a monthly basis (Oct 2020-Dec 2021) to a clinical data warehouse, identifying living patients 50 years or older with ICD-10 codes for dementia or related conditions within the past two years, and an encounter within the past year. This process identified 6,478 unique patients, leading to the recruitment of 837 PLWD into the study (22% Black or Hispanic). Algorithmic components that enhanced recruitment efforts included the incorporation of data elements necessary to contact and enroll patients, formatting data consistent with reporting requirements, and tracking information on primary care provider, diagnosis dates with the set of providers that agreed to recruit for the trial.

Characteristics and Evolution of Cerebral Aneurysms Among Adults Living With HIV: A Retrospective, Longitudinal Case Series

Objective To describe the characteristics and evolution of cerebral aneurysms in a large cohort of adults living with human immunodeficiency virus (ALWH). Background HIV-associated vasculopathy, which can predispose to cerebral aneurysm development, may result from an inflammatory process whereby HIV viral proteins lead to increased chemoattractant and adhesion particles in the vessel wall. Additionally, adventitial layer inflammation has been shown to be strongly associated with HIV infection and vascular changes that may lead to cerebral aneurysm formation. A recent population-based matched cohort study demonstrated a higher rate of subarachnoid hemorrhage in ALWH. It is important to expand the current understanding of risk factors for aneurysm development and the longitudinal course of patients with HIV-associated vasculopathy to help reduce this source of potential morbidity. Design/Methods The Clinical Data Warehouse was queried for all adult patients evaluated at Boston Medical Center between January 1, 2000, and October 22, 2021 with a history of HIV and at least one cerebral aneurysm. Charts were reviewed for variables including timing of HIV and aneurysm diagnoses, antiretroviral (ART) duration, additional aneurysm risk factors, development of new aneurysms/growth of existing aneurysms, interventions, and clinical outcomes. Results A total of 50 patients (52% female) were identified, including 82 cerebral aneurysms. Forty-six percent of patients with a nadir CD4 count less than 200 cells/mm3 (N = 13) developed new aneurysms or were found to have aneurysm growth over time compared with 29% of patients with a CD4 nadir above 200 cells/mm3 (N = 21). New aneurysms were found or existing aneurysms grew in 67% of those not on ART at time of aneurysm diagnosis (N = 6), 38% of those with inconsistent ART use (N = 8), and 21% of those with consistent ART (N = 19). Conclusions Among ALWH, lower CD4 nadir and inconsistent ART use may contribute to aneurysm formation or growth. Further studies are needed to more thoroughly characterize this trend.