Prescribed Drug Use and Aneurysmal Subarachnoid Hemorrhage Incidence: A Drug-Wide Association Study.

BACKGROUND AND OBJECTIVES: Current benefits of invasive intracranial aneurysm treatment to prevent aneurysmal subarachnoid hemorrhage (aSAH) rarely outweigh treatment risks. Most intracranial aneurysms thus remain untreated. Commonly prescribed drugs reducing aSAH incidence may provide leads for drug repurposing. We performed a drug-wide association study (DWAS) to systematically investigate the association between commonly prescribed drugs and aSAH incidence. METHODS: We defined all aSAH cases between 2000 and 2020 using International Classification of Diseases codes from the Secure Anonymised Information Linkage databank. Each case was matched with 9 controls based on age, sex, and year of database entry. We investigated commonly prescribed drugs (>2% in study population) and defined 3 exposure windows relative to the most recent prescription before index date (i.e., occurrence of aSAH): current (within 3 months), recent (3-12 months), and past (>12 months). A logistic regression model was fitted to compare drug use across these exposure windows vs never use, controlling for age, sex, known aSAH risk factors, and health care utilization. The family-wise error rate was kept at p < 0.05 through Bonferroni correction. RESULTS: We investigated exposure to 205 commonly prescribed drugs between 4,879 aSAH cases (mean age 61.4, 61.2% women) and 43,911 matched controls. We found similar trends for lisinopril and amlodipine, with a decreased aSAH risk for current use (lisinopril odds ratio [OR] 0.63, 95% CI 0.44-0.90, amlodipine OR 0.82, 95% CI 0.65-1.04) and an increased aSAH risk for recent use (lisinopril OR 1.30, 95% CI 0.61-2.78, amlodipine OR 1.61, 95% CI 1.04-2.48). A decreased aSAH risk in current use was also found for simvastatin (OR 0.78, 95% CI 0.64-0.96), metformin (OR 0.58, 95% CI 0.43-0.78), and tamsulosin (OR 0.55, 95% CI 0.32-0.93). By contrast, an increased aSAH risk was found for current use of warfarin (OR 1.35, 95% CI 1.02-1.79), venlafaxine (OR 1.67, 95% CI 1.01-2.75), prochlorperazine (OR 2.15, 95% CI 1.45-3.18), and co-codamol (OR 1.31, 95% CI 1.10-1.56). DISCUSSION: We identified several drugs associated with aSAH, of which 5 drugs (lisinopril and possibly amlodipine, simvastatin, metformin, and tamsulosin) showed a decreased aSAH risk. Future research should build on these signals to further assess the effectiveness of these drugs in reducing aSAH incidence. CLASSIFICATION OF EVIDENCE: This study provides Class III evidence that some commonly prescribed drugs are associated with subsequent development of aSAH.

Prediction of aneurysmal subarachnoid hemorrhage in comparison with other stroke types using routine care data.

Aneurysmal subarachnoid hemorrhage (aSAH) can be prevented by early detection and treatment of intracranial aneurysms in high-risk individuals. We investigated whether individuals at high risk of aSAH in the general population can be identified by developing an aSAH prediction model with electronic health records (EHR) data. To assess the aSAH model's relative performance, we additionally developed prediction models for acute ischemic stroke (AIS) and intracerebral hemorrhage (ICH) and compared the discriminative performance of the models. We included individuals aged ≥35 years without history of stroke from a Dutch routine care database (years 2007-2020) and defined outcomes aSAH, AIS and ICH using International Classification of Diseases (ICD) codes. Potential predictors included sociodemographic data, diagnoses, medications, and blood measurements. We cross-validated a Cox proportional hazards model with an elastic net penalty on derivation cohorts and reported the c-statistic and 10-year calibration on validation cohorts. We examined 1,040,855 individuals (mean age 54.6 years, 50.9% women) for a total of 10,173,170 person-years (median 11 years). 17,465 stroke events occurred during follow-up: 723 aSAH, 14,659 AIS, and 2,083 ICH. The aSAH model's c-statistic was 0.61 (95%CI 0.57-0.65), which was lower than the c-statistic of the AIS (0.77, 95%CI 0.77-0.78) and ICH models (0.77, 95%CI 0.75-0.78). All models were well-calibrated. The aSAH model identified 19 predictors, of which the 10 strongest included age, female sex, population density, socioeconomic status, oral contraceptive use, gastroenterological complaints, obstructive airway medication, epilepsy, childbirth complications, and smoking. Discriminative performance of the aSAH prediction model was moderate, while it was good for the AIS and ICH models. We conclude that it is currently not feasible to accurately identify individuals at increased risk for aSAH using EHR data.

Added Predictive Value of Female-Specific Factors and Psychosocial Factors for the Risk of Stroke in Women Under 50.

BACKGROUND AND OBJECTIVES: Female-specific factors and psychosocial factors may be important in the prediction of stroke but are not included in prediction models that are currently used. We investigated whether addition of these factors would improve the performance of prediction models for the risk of stroke in women younger than 50 years. METHODS: We used data from the Stichting Informatievoorziening voor Zorg en Onderzoek, population-based, primary care database of women aged 20-49 years without a history of cardiovascular disease. Analyses were stratified by 10-year age intervals at cohort entry. Cox proportional hazards models to predict stroke risk were developed, including traditional cardiovascular factors, and compared with models that additionally included female-specific and psychosocial factors. We compared the risk models using the c-statistic and slope of the calibration curve at a follow-up of 10 years. We developed an age-specific stroke risk prediction tool that may help communicating the risk of stroke in clinical practice. RESULTS: We included 409,026 women with a total of 3,990,185 person-years of follow-up. Stroke occurred in 2,751 women (incidence rate 6.9 [95% CI 6.6-7.2] per 10,000 person-years). Models with only traditional cardiovascular factors performed poorly to moderately in all age groups: 20-29 years: c-statistic: 0.617 (95% CI 0.592-0.639); 30-39 years: c-statistic: 0.615 (95% CI 0.596-0.634); and 40-49 years: c-statistic: 0.585 (95% CI 0.573-0.597). After adding the female-specific and psychosocial risk factors to the reference models, the model discrimination increased moderately, especially in the age groups 30-39 (Δc-statistic: 0.019) and 40-49 years (Δc-statistic: 0.029) compared with the reference models, respectively. DISCUSSION: The addition of female-specific factors and psychosocial risk factors improves the discriminatory performance of prediction models for stroke in women younger than 50 years.

Cardiovascular Risk Prediction in Men and Women Aged Under 50 Years Using Routine Care Data.

Background Prediction models for risk of cardiovascular events generally do not include young adults, and cardiovascular risk factors differ between women and men. Therefore, this study aimed to develop prediction models for first-ever cardiovascular event risk in men and women aged 30 to 49 years. Methods and Results We included patients aged 30 to 49 years without cardiovascular disease from a Dutch routine care database. Outcome was defined as first-ever cardiovascular event. Our reference models were sex-specific Cox proportional hazards models based on traditional cardiovascular predictors, which we compared with models using 2 predictor subsets with the 20 or 50 most important predictors based on the Cox elastic net model regularization coefficients. We assessed the C-index and calibration curve slopes at 10 years of follow-up. We stratified our analyses based on 30- to 39-year and 40- to 49-year age groups at baseline. We included 542 141 patients (mean age 39.7, 51% women). During follow-up, 10 767 cardiovascular events occurred. Discrimination of reference models including traditional cardiovascular predictors was moderate (women: C-index, 0.648 [95% CI, 0.645-0.652]; men: C-index, 0.661 [95%CI, 0.658-0.664]). In women and men, the Cox proportional hazard models including 50 most important predictors resulted in an increase in C-index (0.030 and 0.012, respectively), and a net correct reclassification of 3.7% of the events in women and 1.2% in men compared with the reference model. Conclusions Sex-specific electronic health record-derived prediction models for first-ever cardiovascular events in the general population aged <50 years have moderate discriminatory performance. Data-driven predictor selection leads to identification of nontraditional cardiovascular predictors, which modestly increase performance of models.

Genetic Risk Score for Intracranial Aneurysms: Prediction of Subarachnoid Hemorrhage and Role in Clinical Heterogeneity.

BACKGROUND: Recently, common genetic risk factors for intracranial aneurysm (IA) and aneurysmal subarachnoid hemorrhage (ASAH) were found to explain a large amount of disease heritability and therefore have potential to be used for genetic risk prediction. We constructed a genetic risk score to (1) predict ASAH incidence and IA presence (combined set of unruptured IA and ASAH) and (2) assess its association with patient characteristics. METHODS: A genetic risk score incorporating genetic association data for IA and 17 traits related to IA (so-called metaGRS) was created using 1161 IA cases and 407 392 controls from the UK Biobank population study. The metaGRS was validated in combination with risk factors blood pressure, sex, and smoking in 828 IA cases and 68 568 controls from the Nordic HUNT population study. Furthermore, we assessed association between the metaGRS and patient characteristics in a cohort of 5560 IA patients. RESULTS: Per SD increase of metaGRS, the hazard ratio for ASAH incidence was 1.34 (95% CI, 1.20-1.51) and the odds ratio for IA presence 1.09 (95% CI, 1.01-1.18). Upon including the metaGRS on top of clinical risk factors, the concordance index to predict ASAH hazard increased from 0.63 (95% CI, 0.59-0.67) to 0.65 (95% CI, 0.62-0.69), while prediction of IA presence did not improve. The metaGRS was statistically significantly associated with age at ASAH (ß=-4.82×10-3 per year [95% CI, -6.49×10-3 to -3.14×10-3]; P=1.82×10-8), and location of IA at the internal carotid artery (odds ratio=0.92 [95% CI, 0.86-0.98]; P=0.0041). CONCLUSIONS: The metaGRS was predictive of ASAH incidence, although with limited added value over clinical risk factors. The metaGRS was not predictive of IA presence. Therefore, we do not recommend using this metaGRS in daily clinical care. Genetic risk does partly explain the clinical heterogeneity of IA warranting prioritization of clinical heterogeneity in future genetic prediction studies of IA and ASAH.

Developing Clinical Prediction Models Using Primary Care Electronic Health Record Data: The Impact of Data Preparation Choices on Model Performance.

Objective: To quantify prediction model performance in relation to data preparation choices when using electronic health records (EHR). Study Design and Setting: Cox proportional hazards models were developed for predicting the first-ever main adverse cardiovascular events using Dutch primary care EHR data. The reference model was based on a 1-year run-in period, cardiovascular events were defined based on both EHR diagnosis and medication codes, and missing values were multiply imputed. We compared data preparation choices based on (i) length of the run-in period (2- or 3-year run-in); (ii) outcome definition (EHR diagnosis codes or medication codes only); and (iii) methods addressing missing values (mean imputation or complete case analysis) by making variations on the derivation set and testing their impact in a validation set. Results: We included 89,491 patients in whom 6,736 first-ever main adverse cardiovascular events occurred during a median follow-up of 8 years. Outcome definition based only on diagnosis codes led to a systematic underestimation of risk (calibration curve intercept: 0.84; 95% CI: 0.83-0.84), while complete case analysis led to overestimation (calibration curve intercept: -0.52; 95% CI: -0.53 to -0.51). Differences in the length of the run-in period showed no relevant impact on calibration and discrimination. Conclusion: Data preparation choices regarding outcome definition or methods to address missing values can have a substantial impact on the calibration of predictions, hampering reliable clinical decision support. This study further illustrates the urgency of transparent reporting of modeling choices in an EHR data setting.