Discordant pair analysis for sample efficient model evaluation.

We present a new technique for assessing the effectiveness of a classification algorithm using discordant pair analysis. This method utilizes a known performance baseline algorithm and a large unlabeled dataset with an assumed class distribution to obtain overall performance estimates by only assessing the subset of examples that the algorithms classify discordantly. Our approach offers an efficient way to evaluate the performance of an algorithm that minimizes the human adjudications needed while also maintaining precision in the evaluation and in some cases improving the evaluation quality by reducing human adjudication errors. This approach is a computationally efficient alternative to the traditional exhaustive method of performance evaluation and has the potential to improve the accuracy of performance estimates. Simulation studies show that the discordant pair method reduces the number of adjudications by over 90%, while maintaining the same level of sensitivity and specificity.

In vitro modeling accurately predicts cardiac lead fracture at 10 years.

BACKGROUND: Development of a cardiac lead fracture model has the potential to differentiate well-performing lead designs from poor performing ones and could aid in future lead development. OBJECTIVE: The purpose of this study was to demonstrate a predictive model for lead fracture and validate the results generated by the model by comparing them to observed 10-year implantable cardioverter-defibrillator lead fracture-free survival. METHODS: The model presented here uses a combination of in vivo patient data, in vitro conductor fatigue test data, and statistical simulation to predict the fracture-free survival of cardiac leads. The model was validated by comparing the results to human clinical performance data from the Medtronic Sprint Fidelis (Minneapolis, MN) models 6931 (single coil, active fixation) and 6949 (dual coil, active fixation), as well as the Quattro model 6947 (dual coil, active fixation). RESULTS: Median patient age in the single coil Fidelis 6931 population (64 years) was less than in the dual coil Fidelis 6949 and Quattro populations (68 years). Modeled and observed fracture-free survival for Quattro (>97%) was superior to that for Fidelis (<94%). The modeled survival agreed with the observed fracture-free survival data. The average model error was 0.3% (SD 1.2%). CONCLUSION: This model for cardiac lead fracture-free survival using in vivo lead bending measurements and in vitro bench testing can be used to predict lead performance as observed by alignment with field survival data.

In silico assessment of biomedical products: The conundrum of rare but not so rare events in two case studies.

In silico clinical trials, defined as "The use of individualized computer simulation in the development or regulatory evaluation of a medicinal product, medical device, or medical intervention," have been proposed as a possible strategy to reduce the regulatory costs of innovation and the time to market for biomedical products. We review some of the the literature on this topic, focusing in particular on those applications where the current practice is recognized as inadequate, as for example, the detection of unexpected severe adverse events too rare to be detected in a clinical trial, but still likely enough to be of concern. We then describe with more details two case studies, two successful applications of in silico clinical trial approaches, one relative to the University of Virginia/Padova simulator that the Food and Drug Administration has accepted as possible replacement for animal testing in the preclinical assessment of artificial pancreas technologies, and the second, an investigation of the probability of cardiac lead fracture, where a Bayesian network was used to combine in vivo and in silico observations, suggesting a whole new strategy of in silico-augmented clinical trials, to be used to increase the numerosity where recruitment is impossible, or to explore patients' phenotypes that are unlikely to appear in the trial cohort, but are still frequent enough to be of concern.

Tuberculous Empyema Post Bilateral Lung Transplant.

Tuberculous empyema in lung transplantation recipients is a rare entity, with only a handful of cases reported in the English-language literature. We are reporting a case of tuberculous empyema 3 months after uncomplicated bilateral lung transplantation. The recipient underwent video-assisted thoracic surgery for diagnosis and decortication. Both the recipient and donor lacked a history of tuberculosis or tuberculosis exposure.

Incorporation of stochastic engineering models as prior information in Bayesian medical device trials.

Evaluation of medical devices via clinical trial is often a necessary step in the process of bringing a new product to market. In recent years, device manufacturers are increasingly using stochastic engineering models during the product development process. These models have the capability to simulate virtual patient outcomes. This article presents a novel method based on the power prior for augmenting a clinical trial using virtual patient data. To properly inform clinical evaluation, the virtual patient model must simulate the clinical outcome of interest, incorporating patient variability, as well as the uncertainty in the engineering model and in its input parameters. The number of virtual patients is controlled by a discount function which uses the similarity between modeled and observed data. This method is illustrated by a case study of cardiac lead fracture. Different discount functions are used to cover a wide range of scenarios in which the type I error rates and power vary for the same number of enrolled patients. Incorporation of engineering models as prior knowledge in a Bayesian clinical trial design can provide benefits of decreased sample size and trial length while still controlling type I error rate and power.

Body surface mapping using an ECG belt to characterize electrical heterogeneity for different left ventricular pacing sites during cardiac resynchronization: Relationship with acute hemodynamic improvement.

BACKGROUND: Electrical heterogeneity (EH) during cardiac resynchronization therapy may vary with different left ventricular (LV) pacing sites. OBJECTIVE: The purpose of this study was to evaluate the relationship between such changes and acute hemodynamic response (AHR). METHODS: Two EH metrics-standard deviation of activation times and mean left thorax activation times-were computed from isochronal maps based on 53-electrode body surface mapping during baseline AAI pacing and biventricular (BiV) pacing from different pacing sites in coronary veins in 40 cardiac resynchronization therapy-indicated patients. AHR at different sites was evaluated by invasive measurement of LV-dp/dtmax at baseline and BiV pacing, along with right ventricular (RV)-LV sensing delays and QRS duration. RESULTS: The site with the greatest combined reduction in standard deviation of activation times and left thorax activation times from baseline to BiV pacing was hemodynamically optimal (defined by AHR equal to, or within 5% of, the largest dp/dt response) in 35 of 40 patients (88%). Sites with the longest RV-LV and narrowest paced QRS were hemodynamically optimal in 26 of 40 patients (65%) and 28 of 40 patients (70%), respectively. EH metrics from isochronal maps had much better accuracy (sensitivity 90%, specificity 80%) for identifying hemodynamically responsive sites (∆LV dp/dtmax ≥10%) compared with RV-LV delay (69%, 85%) or paced QRS reduction (52%, 76%). Multivariate prediction model based on EH metrics showed significant correlation (R2 = 0.53, P <.001) between predicted and measured AHR. CONCLUSION: Changes in EH from baseline to BiV pacing more accurately identified hemodynamically optimal sites than RV-LV delays or paced QRS shortening. Optimization of LV lead location by minimizing EH during BiV pacing, based on body surface mapping, may improve CRT response.

Initial hazard screening for genotoxicity of photo-transformation products of ciprofloxacin by applying a combination of experimental and in-silico testing.

Ciprofloxacin (CIP) is a broad-spectrum antibiotic found within µg/L concentration range in the aquatic environment. It is a known contributor of umuC induction in hospital wastewater samples. CIP can undergo photolysis to result in many transformation products (TPs) of mostly unknown toxicity. The aims of this study were to determine the genotoxicity of the UV mixtures and to understand the possible genotoxic role of the stable TPs. As such, CIP and its UV-irradiated mixtures were investigated in a battery of genotoxicity and cytotoxicity in vitro assays. The combination index (CI) analysis of residual CIP in the irradiated mixtures was performed for the umu assay. Further, Quantitative Structure-Activity Relationships (QSARs) predicted selected genotoxicity endpoints of the identified TPs. CIP achieved primary elimination after 128 min of irradiation but was not completely mineralized. Nine photo-TPs were identified. The irradiated mixtures were neither mutagenic in the Ames test nor genotoxic in the in vitro micronucleus (MN) test. Like CIP, the irradiated mixtures were umuC inducing. The CI analysis revealed that the irradiated mixtures and the corresponding CIP concentration in the mixtures shared similar umuC potentials. QSAR predictions suggested that the TPs may be capable of inducing chromosome aberration, MN in vivo, bacterial mutation and mammalian mutation. However, the experimental testing for a few genotoxic endpoints did not show significant genotoxic activity for the TPs present as a component of the whole mixture analysis and therefore, further genotoxic endpoints may need to be investigated to fully confirm this.

Early Performance and Safety of the Micra Transcatheter Pacemaker in Pigs.

BACKGROUND: The Micra® Transcatheter Pacing System (TPS; Medtronic Inc., Minneapolis, MN, USA) is a miniaturized single-chamber pacemaker system that is delivered via catheter through the femoral vein. In this study, the electrical performance was compared between the TPS and a traditional leaded pacemaker. In addition, the safety profile of the two systems was compared by thorough monitoring for a number of adverse events. METHODS: The TPS was implanted in the right ventricular apex of 10 Yucatan mini pigs and a Medtronic single-lead pacemaker (SLP) was implanted in the right ventricular apex of another 10 pigs and connected to a traditional pacemaker. The electrical performance of all devices was monitored for 12 weeks. The safety profile of each system was characterized using x-ray, computed tomography, ultrasound, blood work, and necropsy to monitor for a variety of adverse events. RESULTS: At implant the mean pacing thresholds were 0.58 ± 0.17 V @0.24 ms and 0.75 ± 0.29 V @0.21 ms for the TPS and the SLP respectively. After 12 weeks, mean thresholds were 0.94 ± 0.46 V and 1.85 ± 0.75 V (P < 0.0001). There were two pulmonary emboli that were small and past the tertiary branch, and one occurred in each arm. There were also two infections with one in each arm. There were no dislodgements (macro or micro), tissue injury, tamponade, or valve injury. CONCLUSIONS: Overall, despite the 10-fold size reduction of the Micra TPS, it appears to perform similarly and have a similar safety profile to a traditional pacemaker system.

Transformation products of antibiotic and cytostatic drugs in the aquatic cycle that result from effluent treatment and abiotic/biotic reactions in the environment: an increasing challenge calling for higher emphasis on measures at the beginning of the pipe.

Pharmaceuticals may undergo transformation into new products during almost all possible processes along their life-cycle. This could either take place in the natural water environment and/or during water treatment processes. Numerous studies that address the issue of such transformation products (TPs) have been published, describing selected aspects of TPs in the environment and their formation within effluent and water treatment processes. In order to exemplify the number and quality of information published on TPs, we selected 21 active pharmaceutical ingredients from the groups of antibiotics and antineoplastics, and assessed the knowledge about their TPs that has been published until the end of May 2012. The goal of this work was to demonstrate, that the quality of data on pharmaceutical TPs greatly differs in terms of the availability of chemical structures for each TP, rather than to provide an exhaustive database of available TPs. The aim was to point out the challenge going along with so many TPs formed under different treatment and environmental conditions. An extensive review in the form of a table showing the existing data on 158 TPs for 15 compounds, out of 21 investigated, was presented. Numerous TPs are the result of different treatments and environmental processes. However, also numerous different TPs may be formed within only one type of treatment, applied under sometimes even very similar treatment conditions and treatments times. In general, the growing number of elucidated TPs is rationalized by ineffective removal treatments. Our results demonstrate a severe risk of drowning in much unrelated and non-assessable data, both from a scientific and from a technical treatment-related point of view. Therefore, limiting the input of pharmaceuticals into effluents as well as improving their (bio) degradability and elimination behavior, instead of only relying on advanced effluent treatments, is urgently needed. Solutions that focus on this "beginning of the pipe" approach should minimize the adverse effects of parent compounds by reducing and formation of TPs and their entrance into the natural environment.

Radiographic predictors of lead conductor fracture.

BACKGROUND: Lead fracture is a limiting factor in high voltage lead durability. Fractures noted with the Medtronic Fidelis leads provide an opportunity to examine factors captured on implant chest x-ray that correlate with risk for lead conductor fracture. We evaluated contributory factors in a large population of fractures. METHODS AND RESULTS: We conducted a retrospective case-control study at 8 Canadian centers that routinely capture anterior posterior and lateral chest x-rays within 2 weeks of implant. Cases were patients that experienced confirmed Medtronic Fidelis 6949 lead fracture based on standard definitions, matched one-to-one to controls for date of implant, sex, and age with normally functioning Fidelis leads from the same center. Select chart data and x-rays were collected for all patients. Radiographic measurements by ≥2 individuals per case/control were blinded to patient status. The data were analyzed using a time to failure multivariable Cox proportional hazards model with stratification for each matched pair. X-ray pairs from 111 fracture patients were compared with 111 controls (age 61.5±12.8 years, 75% male, 221 model 6949 leads). Six parameters included in the statistical analysis were significantly associated with risk of fracture, including slack/tortuosity measures, pulse generator and superior vena cava coil location, and angle of lead exit from the pocket. CONCLUSIONS: Pocket, intravascular and intracardiac lead characteristics on x-ray correlate with risk of lead conductor fracture. These observations may be useful to direct implant technique to optimize lead durability. Validation in larger populations and other lead models may inform the application of these results.

Characterization of photo-transformation products of the antibiotic drug Ciprofloxacin with liquid chromatography-tandem mass spectrometry in combination with accurate mass determination using an LTQ-Orbitrap.

The presence of pharmaceuticals, especially antibiotics, in the aquatic environment is of growing concern. Several studies have been carried out on the occurrence and environmental risk of these compounds. Ciprofloxacin (CIP), a broad-spectrum anti-microbial second-generation fluoroquinolone, is widely used in human and veterinary medicine. In this work, photo-degradation of CIP in aqueous solution using UV and xenon lamps was studied. The transformation products (TPs), created from CIP, were initially analyzed by an ion trap in the MS, MS/MS and MS(3) modes. These data were used to clarify the structures of the degradation products. Furthermore, the proposed products were confirmed by accurate mass measurement and empirical formula calculation for the molecular ions of TPs using LTQ-Orbitrap XL mass spectrometer. The degree of mineralization, the abundance of detected TPs and degradation pathways were determined. Eleven TPs were detected in the present study. TP1, which was never detected before, was structurally characterized in this work. All TPs still retained the core quinolone structure, which is responsible for the biological activity. As mineralization of CIP and its transformation products did not happen, the formation of stable TPs can be expected in waste water treatment and in surface water with further follow-up problems.

Abrasion and fatigue resistance of PDMS containing multiblock polyurethanes after accelerated water exposure at elevated temperature.

Segmented polyurethane multiblock polymers containing polydimethylsiloxane and polyether soft segments form tough and easily processed thermoplastic elastomers (PDMS-urethanes). Two commercially available examples, PurSil 35 (denoted as P35) and Elast-Eon E2A (denoted as E2A), were evaluated for abrasion and fatigue resistance after immersion in 85 °C buffered water for up to 80 weeks. We previously reported that water exposure in these experiments resulted in a molar mass reduction, where the kinetics of the hydrolysis reaction is supported by a straight forward Arrhenius analysis over a range of accelerated temperatures (37-85 °C). We also showed that the ultimate tensile properties of P35 and E2A were significantly compromised when the molar mass was reduced. Here, we show that the reduction in molar mass also correlated with a reduction in both the abrasion and fatigue resistance. The instantaneous wear rate of both P35 and E2A, when exposed to the reciprocating motion of an ethylene tetrafluoroethylene (ETFE) jacketed cable, increased with the inverse of the number averaged molar mass (1/Mn). Both materials showed a change in the wear surface when the number-averaged molar mass was reduced to ≈ 16 kg/mole, where a smooth wear surface transitioned to a 'spalling-like' pattern, leaving the wear surface with ≈ 0.3 mm cracks that propagated beyond the contact surface. The fatigue crack growth rate for P35 and E2A also increased in proportion to 1/Mn, after the molar mass was reduced below a critical value of ≈30 kg/mole. Interestingly, this critical molar mass coincided with that at which the single cycle stress-strain response changed from strain hardening to strain softening. The changes in both abrasion and fatigue resistance, key predictors for long term reliability of cardiac leads, after exposure of this class of PDMS-urethanes to water suggests that these materials are susceptible to mechanical compromise in vivo.

Left ventricular midwall fibrosis as a predictor of mortality and morbidity after cardiac resynchronization therapy in patients with nonischemic cardiomyopathy.

OBJECTIVES: The aim of this study was to determine whether left ventricular (LV) midwall fibrosis, detected by midwall hyperenhancement (MWHE) on late gadolinium enhancement cardiovascular magnetic resonance (CMR) imaging, predicts mortality and morbidity in patients with dilated cardiomyopathy (DCM) undergoing cardiac resynchronization therapy (CRT). BACKGROUND: Midwall fibrosis predicts mortality and morbidity in patients with DCM. METHODS: Patients with DCM with (+) or without (-) MWHE (n = 20 and n = 77, respectively) as well as 161 patients with ischemic cardiomyopathy (ICM) undergoing CRT (n = 258) were followed up for a maximum of 8.7 years. RESULTS: Among patients with DCM, +MWHE predicted cardiovascular mortality (hazard ratio [HR]: 18.6; 95% confidence intervals [CI]: 3.51 to 98.5; p = 0.0008), total mortality or hospitalization for major adverse cardiovascular events (HR: 7.57; 95% CI: 2.71 to 21.2; p < 0.0001), and cardiovascular mortality or heart failure hospitalizations (HR: 9.56; 95% CI: 2.72 to 33.6; p = 0.0004), independent of New York Heart Association class, QRS duration, atrial fibrillation, LV volumes, LV ejection fraction, and a CMR-derived measure of dyssynchrony. Among patients with DCM and ICM, the risk of cardiovascular mortality for DCM +MWHE (adjusted HR: 18.5; 95% CI: 3.93 to 87.3; p = 0.0002) was similar to that for ICM (adjusted HR: 21.0; 95% CI: 5.06 to 87.2; p < 0.0001). Both DCM +MWHE and ICM were predictors of pump failure death as well as sudden cardiac death. LV reverse remodeling was observed in DCM -MWHE and in ICM but not in DCM +MWHE. CONCLUSIONS: Midwall fibrosis is an independent predictor of mortality and morbidity in patients with DCM undergoing CRT. The outcome of DCM with midwall fibrosis is similar to that of ICM. This relationship is mediated by both pump failure and sudden cardiac death.

Effect of bipolar electrode spacing on phrenic nerve stimulation and left ventricular pacing thresholds: an acute canine study.

BACKGROUND: Phrenic nerve stimulation (PNS) is a common complication of cardiac resynchronization therapy when left ventricular (LV) pacing occurs via a coronary vein. The purpose of this study was to evaluate the effects of bipolar electrode spacing on PNS and LV pacing thresholds. METHODS AND RESULTS: Electrophysiology catheters with standard (2 mm-5 mm-2 mm) or modified (1 mm-5 mm-1 mm) interelectrode spacing was, respectively, inserted in a posterior/lateral cardiac vein in a randomized order in 6 anesthetized dogs via jugular access. The phrenic nerve was dissected via a left minithoracotomy and repositioned over the vein as close as possible to one of the electrodes. The presence of PNS was verified (ie, PNS threshold <2 V at 0.5 ms in unipolar configuration). Bipolar pacing was delivered using the electrode closest to the phrenic nerve as the cathode, and multiple bipolar electrode spacing configurations were tested. During bipolar pacing, PNS threshold increased as bipolar electrode spacing was reduced (P<0.05), whereas LV pacing thresholds did not change significantly (P>0.05). Compared with a standard bipolar electrode spacing of 20 mm for LV leads, 1 and 2 mm bipolar electrode spacing resulted in a PNS threshold increase of 5.5±2.2 V (P=0.003) and 2.8±1.7 V (P<0.001), respectively. Similarly, PNS threshold increased by 6.5±3.7 V with 1 mm and by 3.8±1.9 V with 2 mm bipolar pacing (both P<0.001), compared with unipolar pacing. CONCLUSIONS: This study suggests that reducing LV bipolar electrode spacing from the standard 20 mm to 1 or 2 mm may significantly increase the PNS threshold without compromising LV pacing thresholds.

Effect of insulation material in aging pacing leads: comparison of impedance and other electricals: time-dependent pacemaker insulation changes.

BACKGROUND: There has been concern over declining bipolar (BP) impedance (Z) in aging polyurethane (PU) cardiac pacing leads. Subsequently, a prospective study was conducted comparing BP Z, threshold (Th), and R-wave sensing amplitude of 55D PU-insulated (Model 4024, Medtronic, Inc., Minneapolis, MN, USA) and silicone-insulated (Model 5024) leads. METHODS: This study was initiated by The Iowa Heart Center. Patients with Model 4024 (N = 162) or 5024 (N = 120) pacing leads with at least 6 years implant time were enrolled and followed for an additional 5 years. RESULTS: There was a significant drop in the mean BP Z for the Model 4024 population, between enrollment (6 years) and the final endpoint (11 years), which was in contrast to the Model 5024 which did not see a significant drop in its mean BP Z for this same period. The trend difference seen in the means between the two models was statistically significant (P < 0.0001). In addition, a statistically significant relationship was found between dropping BP Z and rising Th (P < 0.0001). The analysis showed that if BP Z dropped below 200 ohms, the probability of having a >3X increase over baseline, in Th at 2.5 V, increases from approximately 3-7% to as high as 30%. CONCLUSIONS: A significant drop in BP Z observed in the PU-insulated Model 4024 lead was not present in the silicone-insulated Model 5024 lead. The statistically significant relationship between dropping BP Z and rising Th helps to understand how to better manage patients with aging leads.

Genotoxic effect of ciprofloxacin during photolytic decomposition monitored by the in vitro micronucleus test (MNvit) in HepG2 cells.

PURPOSE: Ciprofloxacin (CIP), a broad-spectrum, second-generation fluoroquinolone, has frequently been found in hospital wastewaters and effluents of sewage treatment plants. CIP is scarcely biodegradable, has toxic effects on microorganisms and is photosensitive. The aim of this study was to assess the genotoxic potential of CIP in human HepG2 liver cells during photolysis. METHODS: Photolysis of CIP was performed in aqueous solution by irradiation with an Hg lamp, and transformation products were monitored by HPLC-MS/MS and by the determination of dissolved organic carbon (DOC). The cytotoxicity and genotoxicity of CIP and of the irradiated samples were determined after 24 h of exposure using the WST-1 assay and the in vitro micronucleus (MN) test in HepG2 cells. RESULTS: The concentration of CIP decreased during photolysis, whereas the content of DOC remained unchanged. CIP and its transformation products were not cytotoxic towards HepG2 cells. A concentration-dependent increase of MN frequencies was observed for the parent compound CIP (lowest observed effect level, 1.2 µmol L(-1)). Furthermore, CIP and the irradiated samples were found to be genotoxic with a significant increase relative to the parent compound after 32 min (P < 0.05). A significant reduction of genotoxicity was found after 2 h of irradiation (P < 0.05). CONCLUSIONS: Photolytic decomposition of aqueous CIP leads to genotoxic transformation products. This proves that irradiated samples of CIP are able to exert heritable genotoxic effects on human liver cells in vitro. Therefore, photolysis as a technique for wastewater treatment needs to be evaluated in detail in further studies, not only for CIP but in general.

State school nutrition and physical activity policy environments and youth obesity.

BACKGROUND: With the epidemic of childhood obesity, there is national interest in state-level school policies related to nutrition and physical activity, policies adopted by states, and relationships to youth obesity. PURPOSE: This study develops a comprehensive state-level approach to characterize the overall obesity prevention policy environment for schools and links the policy environments to youth obesity for each state. METHODS: Using 2006 School Health Policies and Programs Study (SHPPS) state data, qualitative and quantitative methods were used (2008-2009) to construct domains of state-level school obesity prevention policies and practices, establish the validity and reliability of the domain scales, and examine their associations with state-level obesity prevalence among youth aged 10-17 years from the 2003 National Survey of Children's Health. RESULTS: Nearly 250 state-level obesity prevention-policy questions were identified from the SHPPS. Three broad policy topic areas containing 100 food service and nutrition (FSN) questionnaire items; 146 physical activity and education (PAE) items; and two weight assessment (WA) items were selected. Principal components analysis and content validity assessment were used to further categorize the items into six FSN, ten PAE, and one WA domain. Using a proportional scaled score to summarize the number of policies adopted by states, it was found that on average states adopted about half of the FSN (49%), 38% of the PAE, and 17% of the WA policies examined. After adjusting for state-level measures of ethnicity and income, the average proportion of FSN policies adopted by states was correlated with the prevalence of youth obesity at r =0.35 (p=0.01). However, no correlation was found between either PAE or WA policies and youth obesity (PAE policies at r =0.02 [p=0.53] and WA policies at r =0.16 [p=0.40]). CONCLUSIONS: States appear to be doing a better job adopting FSN policies than PA or WA policies, and adoption of policies is correlated with youth obesity. Continued monitoring of these policies seems to be warranted.