Vision-language foundation model for echocardiogram interpretation.

The development of robust artificial intelligence models for echocardiography has been limited by the availability of annotated clinical data. Here, to address this challenge and improve the performance of cardiac imaging models, we developed EchoCLIP, a vision-language foundation model for echocardiography, that learns the relationship between cardiac ultrasound images and the interpretations of expert cardiologists across a wide range of patients and indications for imaging. After training on 1,032,975 cardiac ultrasound videos and corresponding expert text, EchoCLIP performs well on a diverse range of benchmarks for cardiac image interpretation, despite not having been explicitly trained for individual interpretation tasks. EchoCLIP can assess cardiac function (mean absolute error of 7.1% when predicting left ventricular ejection fraction in an external validation dataset) and identify implanted intracardiac devices (area under the curve (AUC) of 0.84, 0.92 and 0.97 for pacemakers, percutaneous mitral valve repair and artificial aortic valves, respectively). We also developed a long-context variant (EchoCLIP-R) using a custom tokenizer based on common echocardiography concepts. EchoCLIP-R accurately identified unique patients across multiple videos (AUC of 0.86), identified clinical transitions such as heart transplants (AUC of 0.79) and cardiac surgery (AUC 0.77) and enabled robust image-to-text search (mean cross-modal retrieval rank in the top 1% of candidate text reports). These capabilities represent a substantial step toward understanding and applying foundation models in cardiovascular imaging for preliminary interpretation of echocardiographic findings.

Deep Learning-Derived Myocardial Strain.

BACKGROUND: Echocardiographic strain measurements require extensive operator experience and have significant intervendor variability. Creating an automated, open-source, vendor-agnostic method to retrospectively measure global longitudinal strain (GLS) from standard echocardiography B-mode images would greatly improve post hoc research applications and may streamline patient analyses. OBJECTIVES: This study was seeking to develop an automated deep learning strain (DLS) analysis pipeline and validate its performance across multiple applications and populations. METHODS: Interobserver/-vendor variation of traditional GLS, and simulated effects of variation in contour on speckle-tracking measurements were assessed. The DLS pipeline was designed to take semantic segmentation results from EchoNet-Dynamic and derive longitudinal strain by calculating change in the length of the left ventricular endocardial contour. DLS was evaluated for agreement with GLS on a large external dataset and applied across a range of conditions that result in cardiac hypertrophy. RESULTS: In patients scanned by 2 sonographers using 2 vendors, GLS had an intraclass correlation of 0.29 (95% CI: -0.01 to 0.53, P = 0.03) between vendor measurements and 0.63 (95% CI: 0.48-0.74, P < 0.001) between sonographers. With minor changes in initial input contour, step-wise pixel shifts resulted in a mean absolute error of 3.48% and proportional strain difference of 13.52% by a 6-pixel shift. In external validation, DLS maintained moderate agreement with 2-dimensional GLS (intraclass correlation coefficient [ICC]: 0.56, P = 0.002) with a bias of -3.31% (limits of agreement: -11.65% to 5.02%). The DLS method showed differences (P < 0.0001) between populations with cardiac hypertrophy and had moderate agreement in a patient population of advanced cardiac amyloidosis: ICC was 0.64 (95% CI: 0.53-0.72), P < 0.001, with a bias of 0.57%, limits of agreement of -4.87% to 6.01% vs 2-dimensional GLS. CONCLUSIONS: The open-source DLS provides lower variation than human measurements and similar quantitative results. The method is rapid, consistent, vendor-agnostic, publicly released, and applicable across a wide range of imaging qualities.

Deep learning for transesophageal echocardiography view classification.

Transesophageal echocardiography (TEE) imaging is a vital tool used in the evaluation of complex cardiac pathology and the management of cardiac surgery patients. A key limitation to the application of deep learning strategies to intraoperative and intraprocedural TEE data is the complexity and unstructured nature of these images. In the present study, we developed a deep learning-based, multi-category TEE view classification model that can be used to add structure to intraoperative and intraprocedural TEE imaging data. More specifically, we trained a convolutional neural network (CNN) to predict standardized TEE views using labeled intraoperative and intraprocedural TEE videos from Cedars-Sinai Medical Center (CSMC). We externally validated our model on intraoperative TEE videos from Stanford University Medical Center (SUMC). Accuracy of our model was high across all labeled views. The highest performance was achieved for the Trans-Gastric Left Ventricular Short Axis View (area under the receiver operating curve [AUC] = 0.971 at CSMC, 0.957 at SUMC), the Mid-Esophageal Long Axis View (AUC = 0.954 at CSMC, 0.905 at SUMC), the Mid-Esophageal Aortic Valve Short Axis View (AUC = 0.946 at CSMC, 0.898 at SUMC), and the Mid-Esophageal 4-Chamber View (AUC = 0.939 at CSMC, 0.902 at SUMC). Ultimately, we demonstrate that our deep learning model can accurately classify standardized TEE views, which will facilitate further downstream deep learning analyses for intraoperative and intraprocedural TEE imaging.

Electrocardiographic deep learning for predicting post-procedural mortality: a model development and validation study.

BACKGROUND: Preoperative risk assessments used in clinical practice are insufficient in their ability to identify risk for postoperative mortality. Deep-learning analysis of electrocardiography can identify hidden risk markers that can help to prognosticate postoperative mortality. We aimed to develop a prognostic model that accurately predicts postoperative mortality in patients undergoing medical procedures and who had received preoperative electrocardiographic diagnostic testing. METHODS: In a derivation cohort of preoperative patients with available electrocardiograms (ECGs) from Cedars-Sinai Medical Center (Los Angeles, CA, USA) between Jan 1, 2015 and Dec 31, 2019, a deep-learning algorithm was developed to leverage waveform signals to discriminate postoperative mortality. We randomly split patients (8:1:1) into subsets for training, internal validation, and final algorithm test analyses. Model performance was assessed using area under the receiver operating characteristic curve (AUC) values in the hold-out test dataset and in two external hospital cohorts and compared with the established Revised Cardiac Risk Index (RCRI) score. The primary outcome was post-procedural mortality across three health-care systems. FINDINGS: 45 969 patients had a complete ECG waveform image available for at least one 12-lead ECG performed within the 30 days before the procedure date (59 975 inpatient procedures and 112 794 ECGs): 36 839 patients in the training dataset, 4549 in the internal validation dataset, and 4581 in the internal test dataset. In the held-out internal test cohort, the algorithm discriminates mortality with an AUC value of 0·83 (95% CI 0·79-0·87), surpassing the discrimination of the RCRI score with an AUC of 0·67 (0·61-0·72). The algorithm similarly discriminated risk for mortality in two independent US health-care systems, with AUCs of 0·79 (0·75-0·83) and 0·75 (0·74-0·76), respectively. Patients determined to be high risk by the deep-learning model had an unadjusted odds ratio (OR) of 8·83 (5·57-13·20) for postoperative mortality compared with an unadjusted OR of 2·08 (0·77-3·50) for postoperative mortality for RCRI scores of more than 2. The deep-learning algorithm performed similarly for patients undergoing cardiac surgery (AUC 0·85 [0·77-0·92]), non-cardiac surgery (AUC 0·83 [0·79-0·88]), and catheterisation or endoscopy suite procedures (AUC 0·76 [0·72-0·81]). INTERPRETATION: A deep-learning algorithm interpreting preoperative ECGs can improve discrimination of postoperative mortality. The deep-learning algorithm worked equally well for risk stratification of cardiac surgeries, non-cardiac surgeries, and catheterisation laboratory procedures, and was validated in three independent health-care systems. This algorithm can provide additional information to clinicians making the decision to perform medical procedures and stratify the risk of future complications. FUNDING: National Heart, Lung, and Blood Institute.

Artificial Intelligence in Ventricular Arrhythmias and Sudden Death.

Sudden cardiac arrest due to lethal ventricular arrhythmias is a major cause of mortality worldwide and results in more years of potential life lost than any individual cancer. Most of these sudden cardiac arrest events occur unexpectedly in individuals who have not been identified as high-risk due to the inadequacy of current risk stratification tools. Artificial intelligence tools are increasingly being used to solve complex problems and are poised to help with this major unmet need in the field of clinical electrophysiology. By leveraging large and detailed datasets, artificial intelligence-based prediction models have the potential to enhance the risk stratification of lethal ventricular arrhythmias. This review presents a synthesis of the published literature and a discussion of future directions in this field.

Artificial intelligence applications in cardio-oncology: Leveraging high dimensional cardiovascular data.

Growing evidence suggests a wide spectrum of potential cardiovascular complications following cancer therapies, leading to an urgent need for better risk-stratifying and disease screening in patients undergoing oncological treatment. As many cancer patients undergo frequent surveillance through imaging as well as other diagnostic testing, there is a wealth of information that can be utilized to assess one's risk for cardiovascular complications of cancer therapies. Over the past decade, there have been remarkable advances in applying artificial intelligence (AI) to analyze cardiovascular data obtained from electrocardiograms, echocardiograms, computed tomography, and cardiac magnetic resonance imaging to detect early signs or future risk of cardiovascular diseases. Studies have shown AI-guided cardiovascular image analysis can accurately, reliably and inexpensively identify and quantify cardiovascular risk, leading to better detection of at-risk or disease features, which may open preventive and therapeutic opportunities in cardio-oncology. In this perspective, we discuss the potential for the use of AI in analyzing cardiovascular data to identify cancer patients at risk for cardiovascular complications early in treatment which would allow for rapid intervention to prevent adverse cardiovascular outcomes.

Cardiac microstructural alterations measured by echocardiography identify sex-specific risk for heart failure.

OBJECTIVE: Established preclinical imaging assessments of heart failure (HF) risk are based on macrostructural cardiac remodelling. Given that microstructural alterations may also influence HF risk, particularly in women, we examined associations between microstructural alterations and incident HF. METHODS: We studied N=2511 adult participants (mean age 65.7±8.8 years, 56% women) of the Framingham Offspring Study who were free of cardiovascular disease at baseline. We employed texture analysis of echocardiography to quantify microstructural alteration, based on the high spectrum signal intensity coefficient (HS-SIC). We examined its relations to incident HF in sex-pooled and sex-specific Cox models accounting for traditional HF risk factors and macrostructural alterations. RESULTS: We observed 94 new HF events over 7.4±1.7 years. Individuals with higher HS-SIC had increased risk for incident HF (HR 1.67 per 1-SD in HS-SIC, 95% CI 1.31 to 2.13; p<0.0001). Adjusting for age and antihypertensive medication use, this association was significant in women (p=0.02) but not men (p=0.78). Adjusting for traditional risk factors (including body mass index, total/high-density lipoprotein cholesterol, blood pressure traits, diabetes and smoking) attenuated the association in women (HR 1.30, p=0.07), with mediation of HF risk by the HS-SIC seen for a majority of these risk factors. However, the HS-SIC association with HF in women remained significant after adjusting for relative wall thickness (representing macrostructure alteration) in addition to these risk factors (HR 1.47, p=0.02). CONCLUSIONS: Cardiac microstructural alterations are associated with elevated risk for HF, particularly in women. Microstructural alteration may identify sex-specific pathways by which individuals progress from risk factors to clinical HF.

Cardiac microstructural alterations in immune-inflammatory myocardial disease: a retrospective case-control study.

BACKGROUND: Immune-inflammatory myocardial disease contributes to multiple chronic cardiac processes, but access to non-invasive screening is limited. We have previously developed a method of echocardiographic texture analysis, called the high-spectrum signal intensity coefficient (HS-SIC) which assesses myocardial microstructure and previously associated with myocardial fibrosis. We aimed to determine whether this echocardiographic texture analysis of cardiac microstructure can identify inflammatory cardiac disease in the clinical setting. METHODS: We conducted a retrospective case-control study of 318 patients with distinct clinical myocardial pathologies and 20 healthy controls. Populations included myocarditis, atypical chest pain/palpitations, STEMI, severe aortic stenosis, acute COVID infection, amyloidosis, and cardiac transplantation with acute rejection, without current rejection but with prior rejection, and with no history of rejection. We assessed the HS-SIC's ability to differentiate between a broader diversity of clinical groups and healthy controls. We used Kruskal-Wallis tests to compare HS-SIC values measured in each of the clinical populations with those in the healthy control group and compared HS-SIC values between the subgroups of cardiac transplantation rejection status. RESULTS: For the total sample of N = 338, the mean age was 49.6 ± 20.9 years and 50% were women. The mean ± standard error of the mean of HS-SIC were: 0.668 ± 0.074 for controls, 0.552 ± 0.049 for atypical chest pain/palpitations, 0.425 ± 0.058 for myocarditis, 0.881 ± 0.129 for STEMI, 1.116 ± 0.196 for severe aortic stenosis, 0.904 ± 0.116 for acute COVID, and 0.698 ± 0.103 for amyloidosis. Among cardiac transplant recipients, HS-SIC values were 0.478 ± 0.999 for active rejection, 0.594 ± 0.091 for prior rejection, and 1.191 ± 0.442 for never rejection. We observed significant differences in HS-SIC between controls and myocarditis (P = 0.0014), active rejection (P = 0.0076), and atypical chest pain or palpitations (P = 0.0014); as well as between transplant patients with active rejection and those without current or prior rejection (P = 0.031). CONCLUSIONS: An echocardiographic method can be used to characterize tissue signatures of microstructural changes across a spectrum of cardiac disease including immune-inflammatory conditions.

High-Throughput Precision Phenotyping of Left Ventricular Hypertrophy With Cardiovascular Deep Learning.

IMPORTANCE: Early detection and characterization of increased left ventricular (LV) wall thickness can markedly impact patient care but is limited by under-recognition of hypertrophy, measurement error and variability, and difficulty differentiating causes of increased wall thickness, such as hypertrophy, cardiomyopathy, and cardiac amyloidosis. OBJECTIVE: To assess the accuracy of a deep learning workflow in quantifying ventricular hypertrophy and predicting the cause of increased LV wall thickness. DESIGN, SETTINGS, AND PARTICIPANTS: This cohort study included physician-curated cohorts from the Stanford Amyloid Center and Cedars-Sinai Medical Center (CSMC) Advanced Heart Disease Clinic for cardiac amyloidosis and the Stanford Center for Inherited Cardiovascular Disease and the CSMC Hypertrophic Cardiomyopathy Clinic for hypertrophic cardiomyopathy from January 1, 2008, to December 31, 2020. The deep learning algorithm was trained and tested on retrospectively obtained independent echocardiogram videos from Stanford Healthcare, CSMC, and the Unity Imaging Collaborative. MAIN OUTCOMES AND MEASURES: The main outcome was the accuracy of the deep learning algorithm in measuring left ventricular dimensions and identifying patients with increased LV wall thickness diagnosed with hypertrophic cardiomyopathy and cardiac amyloidosis. RESULTS: The study included 23 745 patients: 12 001 from Stanford Health Care (6509 [54.2%] female; mean [SD] age, 61.6 [17.4] years) and 1309 from CSMC (808 [61.7%] female; mean [SD] age, 62.8 [17.2] years) with parasternal long-axis videos and 8084 from Stanford Health Care (4201 [54.0%] female; mean [SD] age, 69.1 [16.8] years) and 2351 from CSMS (6509 [54.2%] female; mean [SD] age, 69.6 [14.7] years) with apical 4-chamber videos. The deep learning algorithm accurately measured intraventricular wall thickness (mean absolute error [MAE], 1.2 mm; 95% CI, 1.1-1.3 mm), LV diameter (MAE, 2.4 mm; 95% CI, 2.2-2.6 mm), and posterior wall thickness (MAE, 1.4 mm; 95% CI, 1.2-1.5 mm) and classified cardiac amyloidosis (area under the curve [AUC], 0.83) and hypertrophic cardiomyopathy (AUC, 0.98) separately from other causes of LV hypertrophy. In external data sets from independent domestic and international health care systems, the deep learning algorithm accurately quantified ventricular parameters (domestic: R2, 0.96; international: R2, 0.90). For the domestic data set, the MAE was 1.7 mm (95% CI, 1.6-1.8 mm) for intraventricular septum thickness, 3.8 mm (95% CI, 3.5-4.0 mm) for LV internal dimension, and 1.8 mm (95% CI, 1.7-2.0 mm) for LV posterior wall thickness. For the international data set, the MAE was 1.7 mm (95% CI, 1.5-2.0 mm) for intraventricular septum thickness, 2.9 mm (95% CI, 2.4-3.3 mm) for LV internal dimension, and 2.3 mm (95% CI, 1.9-2.7 mm) for LV posterior wall thickness. The deep learning algorithm accurately detected cardiac amyloidosis (AUC, 0.79) and hypertrophic cardiomyopathy (AUC, 0.89) in the domestic external validation site. CONCLUSIONS AND RELEVANCE: In this cohort study, the deep learning model accurately identified subtle changes in LV wall geometric measurements and the causes of hypertrophy. Unlike with human experts, the deep learning workflow is fully automated, allowing for reproducible, precise measurements, and may provide a foundation for precision diagnosis of cardiac hypertrophy.

How to evaluate deep learning for cancer diagnostics - factors and recommendations.

The large volume of data used in cancer diagnosis presents a unique opportunity for deep learning algorithms, which improve in predictive performance with increasing data. When applying deep learning to cancer diagnosis, the goal is often to learn how to classify an input sample (such as images or biomarkers) into predefined categories (such as benign or cancerous). In this article, we examine examples of how deep learning algorithms have been implemented to make predictions related to cancer diagnosis using clinical, radiological, and pathological image data. We present a systematic approach for evaluating the development and application of clinical deep learning algorithms. Based on these examples and the current state of deep learning in medicine, we discuss the future possibilities in this space and outline a roadmap for implementations of deep learning in cancer diagnosis.

Gender Trends in Authorship of Foot and Ankle Academic Literature Over 24 Years.

Underrepresentation of females in surgery is reflected in research productivity across academic medicine, with male faculty being more likely to publish research than their female counterparts. In this study, we aimed to describe the representation and longevity of female investigators among the authors of articles in 3 foot and ankle research journals from 1993 to 2017. In this retrospective bibliometric analysis, authors from 3 prominent foot and ankle research journals (Foot and Ankle International, The Journal of Foot and Ankle Surgery, and Foot and Ankle Clinics) were identified. The proportion of female authors who were first, middle, and senior authors and the total publication count per author were determined. From 1993 to 2017, 8132 original articles were published and a total of 6597 (81.1%) had an accessible author list. This allowed us to identify 25,329 total authors, of whom 22,961 (90.7%) were successfully matched to a gender. A total of 9273 unique authors were identified (females, 19.2%). Female representation increased for first and senior authors over the years from 6.5% and 5.9% (1993 to 1997) to 16.9% and 13.1% (2013 to 2017, p < .001), respectively. However, compared with male authors, female authors published fewer articles (mean: 1.7 versus 2.4, p < .001). Of the 2691 authors who first published during 2006 to 2011, 369 authors (13%), consisting of 8.1% females and 15% males (p < .001), continued to publish 5 years after their initial publication. Female representation in academic foot and ankle research has increased >2-fold over the past 2 decades. But despite these advances, compared to male authors, female authors are less likely to continue publishing 5 years after initial publication, and on average publish fewer articles.

National trends and complication rates for invasive extraoperative electrocorticography in the USA.

Invasive electrocorticography (ECoG) is used in patients when it is difficult to localize epileptogenic foci for potential surgical resection. As MR neuroimaging has improved over the past decade, we hypothesized the utilization of ECoG diminishing over time. Using the USA Nationwide Inpatient Sample, we collected demographic and complication data on patients receiving ECoG over the years 1988-2008 and compared this to patients with medically refractory epilepsy during the same time period. A total of 695 cases using extraoperative ECoG were identified, corresponding to 3528 cases nationwide and accounting for 1.1% of patients with refractory epilepsy from 1988-2008. African Americans were less likely to receive ECoG than whites, as were patients with government insurance in comparison to those with private insurance. Large, urban, and academic hospitals were significantly more likely to perform ECoG than smaller, rural, and private practice institutions. The most frequent complication was cerebrospinal fluid leak (11.7%) and only one death was reported from the entire cohort, corresponding to an estimated six patients nationally. Invasive ECoG is a relatively safe procedure offered to a growing number of patients with refractory epilepsy each year. However, these data suggest the presence of demographic disparities in those patients receiving ECoG, possibly reflecting barriers due to race and socioeconomic status. Among patients with nonlocalized seizures, ECoG often represents their only hope for surgical treatment. We therefore must further examine the indications and efficacy of ECoG, and more work must be done to understand if and why ECoG is preferentially performed in select socioeconomic groups.

National trends in surgery for sinonasal malignancy and the effect of hospital volume on short-term outcomes.

OBJECTIVES/HYPOTHESIS: To characterize trends in the management of sinonasal malignancy with a focus on the impact of hospital volume on surgical outcomes. STUDY DESIGN: Retrospective cohort study. METHODS: Time trends were analyzed among patients admitted for surgical resection of sinonasal malignancy in the Nationwide Inpatient Sample (NIS) between 1988 and 2009. Subset analysis was performed on cohorts with skull base or orbital involvement or on cohorts who underwent neck dissection. Patient characteristics and hospital attributes were correlated with morbidity and mortality. RESULTS: Over 22 years, we identified 3,850 cases from 879 hospitals. A total of 24.3% of patients had complications and 0.8% of hospitalizations resulted in mortality. Cases with skull base or orbital involvement, or cases including neck dissection, had more complications and a longer length of stay. Prevalence of neck dissection increased over time. Thirty-two hospitals averaged more than five cases per year, accounting for 28% of all surgeries. These centers were large (73.3%), urban (96.7%), teaching (90%) institutions and performed more high-risk cases: 32.4% of neck dissections, 44.6% of orbital cases, and 43.1% of skull base cases. Compared to lower-volume centers, these centers had more cardiopulmonary and electrolyte complications, but no difference was observed in the lengths of stay. A greater proportion of cases were recently performed at high-volume centers. CONCLUSION: Over time, complicated surgeries were more likely to occur at higher-volume hospitals without significant changes in surgical complication rates. High-volume centers had increased rates of cardiopulmonary and electrolyte complications, likely representing complex postoperative management, but these were not associated with higher mortality. LEVEL OF EVIDENCE: 2c.

Trends in surgical treatment for trigeminal neuralgia in the United States of America from 1988 to 2008.

Current surgical treatments for refractory trigeminal neuralgia (TN) include microvascular decompression (MVD), percutaneous rhizotomy, and stereotactic radiosurgery (SRS). We aimed to map the trends of utilization of these procedures in the USA and examine factors associated with morbidities and discharge outcome. We performed a retrospective cohort study with time trends of patients admitted to US hospitals for TN between 1988 and 2008 who received MVD, percutaneous rhizotomy, or SRS as reported in the Nationwide Inpatient Sample. Univariate and multivariate analyses were conducted to examine patient demographics, hospital characteristics, and other hospitalization factors affecting complications and discharges. The use of MVD increased significantly by 194% from 1988 to 2008 while rhizotomy decreased by 92%. The use of radiosurgery, introduced in the early 1990s, peaked in 2004 and has declined since. Univariate analysis revealed patient age, length of hospitalization, hospital teaching status, and hospital patient volume to be associated with discharge and complications. Multivariate analysis showed that for MVD, younger age and high hospital volume were predictive of a good discharge outcome. For rhizotomy, age, median income, urban location, and hospital volumes were associated with discharge outcome, but only teaching status, urban location, and hospital volume were associated with complications. For SRS, patient age and length of stay were found to be important by multivariate analysis on discharge. Mortality rates for MVD (0.22%), rhizotomy (0.42%), and SRS (0.12%) were low. The clinical practices for surgical treatment of TN have evolved over time with the rise of MVD and dwindling of rhizotomy procedures.

Relationship between hospital surgical volume, lobectomy rates, and adverse perioperative events at US epilepsy centers.

OBJECT: Epilepsy surgery remains significantly underutilized. The authors recently reported that the number of lobectomies for localized intractable epilepsy in the US has not changed despite the implementation of clear evidence-based guidelines 10 years ago supporting early referral for surgery. To better understand why epilepsy surgery continues to be underused, the authors' objective was to carefully examine hospital-related factors related to the following: 1) where patients are being admitted for the evaluation of epilepsy, 2) rates of utilization for surgery across hospitals, and 3) perioperative morbidity between hospitals with low versus high volumes of epilepsy surgery. METHODS: The authors performed a population-based cohort study of US hospitals between 1990 and 2008 using the Nationwide Inpatient Sample (NIS), stratifying epilepsy surgery rates and trends as well as perioperative morbidity rates by hospital surgical volume. RESULTS: The number of lobectomies for epilepsy performed at high-volume centers (> 15 lobectomies/year) significantly decreased between 1990 and 2008 (F = 20.4, p < 0.001), while significantly more procedures were performed at middle-volume hospitals (5-15 lobectomies/year) over time (F = 16.1, p < 0.001). No time trend was observed for hospitals performing fewer than 5 procedures per year. However, patients admitted to high-volume centers were significantly more likely to receive lobectomy than those at low-volume hospitals (relative risk 1.05, 95% CI 1.03-1.08, p < 0.001). Also, the incidence of perioperative adverse events was significantly higher at low-volume hospitals (12.9%) than at high-volume centers (6.1%) (relative risk 1.08, 95% CI 1.03-1.07, p < 0.001). CONCLUSIONS: Hospital volume is an important predictor of epilepsy surgery utilization and perioperative morbidity. Patients with medically refractory epilepsy should be referred to a comprehensive epilepsy treatment center for surgical evaluation by an experienced clinical team.

Obstetric complications of fibroids.

The effect of uterine fibroids on fecundity and pregnancy outcome is difficult to determine with any degree of accuracy; this is due, in large part, to the lack of adequate large clinical trials. In general, the literature tends to underestimate the prevalence of fibroids in pregnancy and overestimate the complications that are attributed to them. In contrast to popular opinion, most fibroids do not exhibit a significant change in volume during pregnancy, although those that do increase in size tend to do so primarily in the first trimester. Although most pregnancies are unaffected by the presence of uterine fibroids, large submucosal and retro-placental fibroids seem to impart a greater risk for complications, including pain (degeneration), vaginal bleeding, placental abruption, IUGR, and preterm labor and birth. Preconception myomectomy to improve reproductive outcome can be considered on an individual basis, but likely has a place only in women who have recurrent pregnancy loss, large submucosal fibroids, and no other identifiable cause for recurrent miscarriage. Antepartum myomectomy should be reserved for women who have subserosal or pedunculated fibroids and intractable fibroid pain that are unresponsive to medical therapy and who are in the first or second trimester of pregnancy. Myomectomy at the time of cesarean delivery is associated with significant morbidity (hemorrhage) and should be pursued with caution and only in select patients.

Pregnancy outcomes in women with congenital heart disease.

BACKGROUND: Pregnant women with congenital heart disease are at increased risk for cardiac and neonatal complications, yet risk factors for adverse outcomes are not fully defined. METHODS AND RESULTS: Between January 1998 and September 2004, 90 pregnancies at age 27.7+/-6.1 years were followed in 53 women with congenital heart disease. Spontaneous abortions occurred in 11 pregnancies at 10.8+/-3.7 weeks, and 7 underwent elective pregnancy termination. There were no maternal deaths. Primary maternal cardiac events complicated 19.4% of ongoing pregnancies, with pulmonary edema in 16.7% and sustained arrhythmias in 2.8%. Univariate risk factors included prior history of heart failure (odds ratio [OR], 15.5), NYHA functional class > or =2 (OR, 5.4), and decreased subpulmonary ventricular ejection fraction (OR, 7.7). Independent predictors were decreased subpulmonary ventricular ejection fraction and/or severe pulmonary regurgitation (OR, 9.0) and smoking history (OR, 27.2). Adverse neonatal outcomes occurred in 27.8% of ongoing pregnancies and included preterm delivery (20.8%), small for gestational age (8.3%), respiratory distress syndrome (8.3%), intraventricular hemorrhage (1.4%), intrauterine fetal demise (2.8%), and neonatal death (1.4%). A subaortic ventricular outflow tract gradient >30 mm Hg independently predicted an adverse neonatal outcome (OR, 7.5). Cardiac risk assessment was improved by including decreased subpulmonary ventricular systolic function and/or severe pulmonary regurgitation (OR, 10.3) in a previously proposed risk index developed in pregnant women with acquired and congenital heart disease. CONCLUSIONS: Maternal cardiac and neonatal complication rates are considerable in pregnant women with congenital heart disease. Patients with impaired subpulmonary ventricular systolic function and/or severe pulmonary regurgitation are at increased risk for adverse cardiac outcomes.