Model Validation Group Research Articles

Development of a nomogram to predict recurrence scores obtained using Oncotype DX in Japanese patients with breast cancer

BackgroundChemotherapy is crucial for hormone receptor-positive, human epidermal growth factor receptor 2 (HER2)-negative breast cancer, and its survival benefits may outweigh adverse events. Oncotype DX (ODX) assesses this balance; however, it is expensive. Using nomograms to identify cases requiring ODX may be economically beneficial. We aimed to identify clinicopathological variables that correlated with the recurrence score (RS) and develop a nomogram that predicted the RS.MethodsWe included 457 patients with estrogen receptor-positive, HER2-negative breast cancer with metastases in fewer than four axillary lymph nodes who underwent surgery and ODX at our hospital between 2007 and 2023. We developed nomograms and internally validated them in 310 patients who underwent surgery between 2007 and 2021 and validated the model’s performance in 147 patients who underwent surgery between 2022 and 2023.ResultsLogistic regression analysis revealed that progesterone receptor (PgR) level, histological grade (HG), and Ki67 index independently predicted the RS. A nomogram was developed using these variables to predict the RS (area under the curve [AUC], 0.870; 95% confidence interval [CI], 0.82–0.92). The nomogram was applied to the model validation group (AUC, 0.877; 95% CI, 0.80–0.95). When the sensitivity of the nomogram was 90%, the model was able to identify 52.3% low-RS and 41.2% high-RS cases not requiring ODX.ConclusionsThis was the first nomogram model developed based on data from a cohort of Japanese women. It may help determine the indications for ODX and the use of nomogram to identify cases requiring ODX may be economically beneficial.

A novel predictive model for noninvasively diagnosing bladder outlet obstruction in female patients based on clinical features and uroflowmetry parameters.

To develop and validate a simple prediction model to diagnose female bladder outlet obstruction (fBOO) because of the invasive nature of standard urodynamic studies (UDS) for diagnosing fBOO. We retrospectively analyzed the data of 728 women who underwent UDS at Tongji Hospital between 2011 and 2021. The definition of fBOO was Pdet.Qmax - 2.2 × Qmax > 5 (BOOIf > 5). Independent predictive factors of fBOO were determined by multivariable logistic regression analysis. These predictive factors were incorporated into a predictive model to assess the risk of fBOO. Of the 728 patients, 249 (34.2%) were identified as having fBOO and these women were randomly assigned to two groups, a model development group and a model validation group. Multivariate logistic regression demonstrated that age, Qmax, flow time, and voiding efficiency were independent risk factors for fBOO. The predictive model of fBOO showed a satisfactory performance, with area under the curve being 0.811 (95% confidence interval [CI] 0.771-0.850, P < 0.001), which was confirmed to be 0.820 (95% CI 0.759-0.882, P < 0.001) with external validation. The calibration curve indicated that the predicted probability had an excellent correspondence to observed frequency. Decision curve analysis demonstrated a greater clinical net benefit compared with the strategies of treat all or treat none when the predicted risk was in a range of 3% and 75%. A novel predictive model of fBOO was developed and validated based on clinical features and noninvasive test parameters in female patients with lower urinary tract symptoms. The model is a quick and easy-to-use tool to assess the risk of fBOO for urologists in their routine practice without an invasive UDS.

StepTest4all: Improving the Prediction of Cardiovascular Capacity Assessment in Young Adults.

Cardiovascular capacity, expressed as maximal oxygen uptake (VO2max), is a strong predictor of health and fitness and is considered a key measure of physiological function in the healthy adult population. The aim of this study was to investigate the influence of the physical activity levels (PAlevel) of participants in the StepTest4all (validated protocol for the estimation of VO2max in adults). The sample consisted of 69 participants, including 27 women (age 21.7 ± 3.6 years; body mass = 63.5 ± 14.8 kg; height = 1.64 ± 0.06 m; body mass index = 23.7 ± 5.3 kg/m2) and 42 men (aged 21.7 ± 3.4 years; body mass = 72.0 ± 7.3 kg; height = 1.77 ± 0.07 m; body mass index = 23.1 ± 2.1 kg/m2). The participants were assigned to one of the two groups: (i) the VO2max prediction group and (ii) the prediction model validation group. In the multiple linear regression, the following predictors of VO2max remained significant: sex (p < 0.001), physical activity level (p = 0.014), and HRR60 (p = 0.020). The prediction equation (R2 = 74.0%, SEE = 4.78) showed a close and strong relationship between the measurements and can be expressed as follows: VO2max = 17.105 + 0.260·(HRR60) + 8.563·(sex) + 4.097·(PAlevel), in which HRR60 is the magnitude of the HR decrease (bpm) in one minute immediately after stopping the step, and sex: men = 1, women = 0, and PAlevel is level 1 (low), level 2 (moderate), and level 3 (high). The StepTest4all was shown to be a suitable method for estimating cardiovascular capacity, expressed as VO2max, in young adults. Retaining PAlevel as a significant predictor allows us to better individualize the participants' VO2max.

Development and external validation of MRI-based RAS mutation status prediction model for liver metastases of colorectal cancer.

The mutation status of rat sarcoma viral oncogene homolog (RAS) has prognostic significance and serves as a key predictive biomarker for the effectiveness of antiepidermal growth factor receptor (EGFR) therapy. However, there remains a lack of effective models for predicting RAS mutation status in colorectal liver metastases (CRLMs). This study aimed to construct and validate a diagnostic model for predicting RAS mutation status among patients undergoing hepatic resection for CRLMs. A diagnostic multivariate prediction model was developed and validated in patients with CRLMs who had undergone hepatectomy between 2014 and 2020. Patients from Institution A were assigned to the model development group (i.e., Development Cohort), while patients from Institutions B and C were assigned to the external validation groups (i.e., Validation Cohort_1 and Validation Cohort_2). The presence of CRLMs was determined by examination of surgical specimens. RAS mutation status was determined by genetic testing. The final predictors, identified by a group of oncologists and radiologists, included several key clinical, demographic, and radiographic characteristics derived from magnetic resonance images. Multiple imputation was performed to estimate the values of missing non-outcome data. A penalized logistic regression model using the adaptive least absolute shrinkage and selection operatorpenalty was implemented to select appropriate variables for the development of the model. A single nomogram was constructed from the model. The performance of the prediction model, discrimination,and calibration were estimated and reported by the area under the receiver operating characteristic curve (AUC) and calibration plots. Internal validation with a bootstrapping procedure and external validation of the nomogram were assessed. Finally, decision curve analyses were used to characterize the clinical outcomes of the Development and Validation Cohorts. A total of 173 patients were enrolled in this study between January 2014 and May 2020. Of the 173 patients, 117 patients from Institution A were assigned to the Model Development group, while 56 patients (33 from Institution B and 23 from Institution C) were assigned to the Model Validation groups. Forty-six (39.3%) patients harbored RAS mutations in the Development Cohort compared to 14 (42.4%) in Validation Cohort_1 and 8 (34.8%) in Validation Cohort_2. The final model contained the following predictor variables: time of occurrence of CRLMs, location of primary lesion, type of intratumoral necrosis, and early enhancement of liver parenchyma. The diagnostic model based on clinical and MRI data demonstrated satisfactory predictive performance in distinguishing between mutated and wild-type RAS, with AUCs of 0.742 (95% confidence interval [CI]: 0.651─0.834), 0.741 (95% CI: 0.649─0.836), 0.703 (95% CI: 0.514─0.892), and 0.708 (95% CI: 0.452─0.964) in the Development Cohort, bootstrapping internal validation, external Validation Cohort_1 and Validation Cohort_2, respectively. The Hosmer-Lemeshow goodness-of-fit values for the Development Cohort, Validation Cohort_1 and Validation Cohort_2 were 2.868 (p = 0.942), 4.616 (p = 0.465),and 6.297 (p = 0.391), respectively. Integrating clinical, demographic, and radiographic modalities with a magnetic resonance imaging-based approach may accurately predict the RAS mutation status of CRLMs, thereby aiding in triage and possibly reducing the time taken to perform diagnostic and life-saving procedures. Our diagnostic multivariate prediction model may serve as a foundation for prognostic stratification and therapeutic decision-making.

Development and validation of a nomogram to predict cardiac death after radiotherapy for esophageal cancer.

Patients frequently die from cardiac causes after radiotherapy for esophageal cancer. Early detection of cardiac death risk in these patients is crucial to improve clinical decision-making and prognosis. Thus, we modeled the risk of cardiac death after irradiation for esophageal cancer. A retrospective analysis of 37,599 esophageal cancer cases treated with radiotherapy in the SEER database between 2000 and 2018 was performed. The selected cases were randomly assigned to the model development group (n = 26,320) and model validation group (n = 11,279) at a ratio of 7:3. We identified the risk factors most commonly associated with cardiac death by least absolute shrinkage and selection operator regression analysis (LASSO). The endpoints for model development and validation were 5- and 10-year survival rates. The net clinical benefit of the models was evaluated by decision curve analysis (DCA) and concordance index (C-index). The performance of the models was further assessed by creating a receiver operating characteristic curve (ROC) and calculating the area under the curve (AUC). Kaplan-Meier (K-M) survival analysis was performed on the probability of death. Patients were classified according to death probability thresholds. Five- and ten-year survival rates for the two groups were shown using K-M curves. The major risk factors for cardiac death were age, surgery, year of diagnosis, sequence of surgery and radiotherapy, chemotherapy and a number of tumors, which were used to create the nomogram. The C-indexes of the nomograms were 0.708 and 0.679 for the development and validation groups, respectively. DCA showed the good net clinical benefit of nomograms in predicting 5- and 10-year risk of cardiac death. The model exhibited moderate predictive power for 5- and 10-year cardiac mortality (AUC: 0.833 and 0.854, respectively), and for the development and validation cohorts (AUC: 0.76 and 0.813, respectively). Our nomogram may assist clinicians in making clinical decisions about patients undergoing radiotherapy for esophageal cancer based on early detection of cardiac death risk.

Network pharmacology-based study on the mechanism of ShenKang injection in diabetic kidney disease through Keap1/Nrf2/Ho-1 signaling pathway

ObjectiveTo study the effect of ShenKang Injection (SKI) on the kidneys of DKD rats and its effect on oxidative stress mediated by the Keap1/Nrf2/Ho-1 signaling pathway through network pharmacology and in vivo and in vitro experiments. MethodsSKI drug targets were screened by TCMSP, DKD targets were screened by GenGards, OMIM, Drugbank, TTD, and Disgenet databases, and the two intersected for PPI network analysis and target prediction was performed by GO and KEGG. A total of 40 SD rats were randomly divided into 10 in the control group and 30 in the model group. After the model group was fed 8 W with high-sugar and high-fat diets, a DKD model was constructed by one-time intraperitoneal injection of streptozotocin (35 mg/kg). According to the weight, the model animals were randomly divided into three groups: 8 for model validation group, 8 for Irbesartan (25 mg/kg daily) group, and 8 for SKI group (5 ml/kg). Gavaged deionized water was given to the control group and the model validation group equally. The general conditions of the rats were observed, their body weights measured and their urine volumes recorded for 24 h. After the intervention of 16 W, serum was collected to detect Urea, Scr, blood lipids, and oxidative stress and lipid peroxidation indicators; Transmission electron microscopy, HE and Mallory staining were used to observe the pathological morphology of renal tissue. Immunohistochemistry and RT-PCR were used to detect the expression of Keap1, Nrf2, Ho-1, Gpx4 proteins and mRNA in rat kidney tissues. HK-2 cells were cultured in vitro and divided into: the control group, AGEs (200 μg/ml) group and AGEs + SKI group. The cell activity of the groups was detected using CCK-8 after 48 h of cell culture, and ROS were detected using fluorescent probes. Gpx4 expression was detected by immunofluorescence, while Keap1, Nrf2, Ho-1, and Gpx4 were detected by Western Blot. ResultsNetwork pharmacological analysis predicted that SKI may delay DKD kidney injury by affecting redox-related signaling pathways and mitigating AGEs-induced oxidative stress. In the animal experiment, compared with the model validation group, the general state of rats in the SKI group was improved, and 24-hour urine protein levels were significantly reduced, and the Scr in the serum was reduced. A decreasing trend was seen in Urea, and TC, TG, and LDL levels significantly decreased and the levels of ROS, LPO and MDA were significantly lowered. Pathological staining showed that renal interstitial fibrosis was significantly improved, and electron microscopy showed that foot process effacement was alleviated. Immunohistochemistry and RT-PCR showed decreased expression of Keap1 protein and mRNA in kidney tissues of the SKI group. Additionally, Nrf2, Ho-1, and Gpx4 proteins and mRNA were expressed significantly. In the cell experiment, after 48 h treatment with AGEs, ROS in HK-2 cells increased significantly and cell activity decreased significantly, while cell activity in AGEs + SKI group increased significantly and ROS decreased. The expression of Keap1 protein in HK-2 cells in the AGEs + SKI group decreased, while the expression of Nrf2, Ho-1 and Gpx4 proteins increased significantly. ConclusionSKI can protect kidney function in DKD rats, delay DKD progression, inhibit AGEs-induced oxidative stress damage in HK-2 cells, and the mechanism of SKI to improve DKD may be achieved by activating the Keap1/Nrf2/Ho-1 signal transduction pathway.

Expression of mir-25-3p, CARD9 and Surviving in Acute Pancreatitis and Predictive Value for Patient Outcome.

The clinical data of 246 patients with acute pancreatitis who met the inclusion and exclusion criteria in our hospital from May 2018 to May 2020 were collected as the modeling group, and 96 patients were used as the model validation group. To observe the expression of mir-25-3p, CARD9 and Survivin in patients with acute pancreatitis. To analyze the prognostic factors of acute pancreatitis by univariate and multivariate analysis, and to establish and validate the prognostic model of acute pancreatitis. Results: There was no significant difference in general data between the two groups (P>0.05). Of 246 AP patients, 217 survived and 29 died. The APACHEI score, BISAP score, CRP, lipase, lactate, mir-25-3p, CARD9 and Survivin in the survival group were lower than those in the death group, and the differences were statistically significant (P<0.05). There was no significant difference in other indexes (P>0.05). APACHEI score, BISAP score, CRP, lipase, lactate, mir-25-3p, CARD9 and Survivin were included in multivariate logistic regression analysis. Survival =1 and death =0 were the dependent variables. BISAP score, CRP, lipase, lactate, mir-25-3p, CARD9 and Survivin were all protective factors for the survival of AP patients. Log(P)=-1.648×BISAP score -0.045×CRP-0.013× lipase-0.205 × lactate-1.339 × Mir-25-3P-2.701 ×CARD9-1.663×Survivin+43.925. The survival protective factors of AP patients were incorporated into R software to establish the nomogram prediction model.

Spatial modelling of gully headcuts using UAV data and four best-first decision classifier ensembles (BFTree, Bag-BFTree, RS-BFTree, and RF-BFTree)

Despite the importance of delineating spatial modelling of gully headcuts (GHs) in erosion-prone environments, assessments of the factors that control the occurrence of headcuts is lacking. To fill this gap in the research, we identified 129 GHS field surveys. These 129 cases were randomly divided into two groups: 90 GHs (70%) for model training and 39 GHs (30%) for model validation. Subsequently, new unmanned aerial vehicle (UAV) imagery is used to develop spatial modelling to predict the location of GHs at sites prone to soil erosion in Golestan Province, Iran. Mapping GHs enables evaluation of 4 machine-learning techniques (or ensembles) – best-first decision tree (BFTree), bagging best-first decision tree (Bag-BFTree), random-subspace best-first decision tree (RS-BFTree), and rotation-forest best-first decision tree (RF-BFTree) – for modelling GHs. We use the information-gain ratio method to analyze the relationships between GHS and 22 GH conditioning factors. The 4 ensemble outputs are validated using a receiver operating characteristic (ROC) curve. The areas under the curves (AUCs) for prediction rates of the ensemble methods applied to the training group are BFTree – 88.3%, Bag-BFTree – 92.7%, RS-BFTree – 95.7%, and RF-BFTree – 93.2%. The AUCs for the model-validation group cases, however, are 84.9%, 94.1%, 97.4%, and 9.18%, respectively. Therefore, RS-BFTree is, statistically, the most effective ensemble method for accurate modelling of GHs. Variable-importance analyses using information-gain ratio indicate that out of 22 GH-influential factors, land use, slope degree, and slope-length are of more importance in developing of GH occurrence. Finally, to address the need for detailed observations and highly accurate erosion data in the field, UAV image-acquisition technologies are demonstrated.

Risk score predictive of mortality in status epilepticus according to a national database.

Status epilepticus (SE) is a serious neurologic condition with high morbidity and mortality rates. This study aimed to develop and validate a risk score that is predictive of mortality in patients with SE using clinical factors without electrocardiography. The inclusion criteria of this study were all patients diagnosed with SE and treated between 2005 and 2015. We retrospectively searched for eligible patients using the International Classification of Diseases, Tenth Revision (ICD-10) code for SE (G41) in the national Universal Health Coverage database. The outcome was death at discharge or within 30days after discharge. Factors-associated death was analyzed using stepwise logistic regression analysis. Risk scores were developed based on the final logistic regression model. The final model was also validated. There were 10924 patients used for model development and 10808 used for model validation. The formula to determine the risk score for SE mortality was 5×shock+4×age over 60 years old+3.5×heart diseases+3×acute renal failure+3×septicemia+2.5×central nervous system infection+2.5×age 41-60 years old+2×cancer+2×chronic renal failure+1.5×age 21-40 years old+1×pneumonia+1×respiratory failure+1×anemia. The risk scores of greater than 4 indicated risk for mortality with a sensitivity of 78.20% and specificity of 75.38%. The area under the receiver-operating characteristic (ROC) curve for death in the final model was 83.59%. The area under the ROC curve for the model validation group was 83.52%. SE patients who had a risk score of 4 or more were at high risk for death. Physicians should be aware of the high mortality rate in these particular patients.

Fingerprint verification prediction model in hand dermatitis.

Hand dermatitis associated fingerprint changes is a significant problem and affects fingerprint verification processes. This study was done to develop a clinically useful prediction model for fingerprint verification in patients with hand dermatitis. A case-control study involving 100 patients with hand dermatitis. All patients verified their thumbprints against their identity card. Registered fingerprints were randomized into a model derivation and model validation group. Predictive model was derived using multiple logistic regression. Validation was done using the goodness-of-fit test. The fingerprint verification prediction model consists of a major criterion (fingerprint dystrophy area of ≥25%) and two minor criteria (long horizontal lines and long vertical lines). The presence of the major criterion predicts it will almost always fail verification, while presence of both minor criteria and presence of one minor criterion predict high and low risk of fingerprint verification failure, respectively. When none of the criteria are met, the fingerprint almost always passes the verification. The area under the receiver operating characteristic curve was 0.937, and the goodness-of-fit test showed agreement between the observed and expected number (P=0.26). The derived fingerprint verification failure prediction model is validated and highly discriminatory in predicting risk of fingerprint verification in patients with hand dermatitis.

Risk models to improve safety of dispensing high-alert medications in community pharmacies

To determine whether sociotechnical probabilistic risk assessment can create accurate approximations of detailed risk models that describe error pathways, estimate the incidence of preventable adverse drug events (PADEs) with high-alert medications, rank the effectiveness of interventions, and provide a more informative picture of risk in the community pharmacy setting than is available currently. Developmental study. 22 community pharmacies representing three U.S. regions. Model-building group: six pharmacists and three technicians. Model validation group: 11 pharmacists; staff at two pharmacies observed. A model-building team built 10 event trees that estimated the incidence of PADEs for four high-alert medications: warfarin, fentanyl transdermal systems, oral methotrexate, and insulin analogs. Validation of event tree structure and incidence of defined PADEs with targeted medications. PADEs with the highest incidence included dispensing the wrong dose/strength of warfarin as a result of data entry error (1.83/1,000 prescriptions), dispensing warfarin to the wrong patient (1.22/1,000 prescriptions), and dispensing an inappropriate fentanyl system dose due to a prescribing error (7.30/10,000 prescriptions). PADEs with the lowest incidence included dispensing the wrong drug when filling a warfarin prescription (9.43/1 billion prescriptions). The largest quantifiable reductions in risk were provided by increasing patient counseling (27-68% reduction), conducting a second data entry verification process during product verification (50-87% reduction), computer alerts that can't be bypassed easily (up to 100% reduction), opening the bag at the point of sale (56% reduction), and use of barcoding technology (almost a 100,000% increase in risk if technology not used). Combining two or more interventions resulted in further overall reduction in risk. The risk models define thousands of ways process failures and behavioral elements combine to lead to PADEs. This level of detail is unavailable from any other source.

Towards an individualized protocol for workload increments in cardiopulmonary exercise testing in children and adolescents with cystic fibrosis

BackgroundThere is no single optimal exercise testing protocol for children and adolescents with cystic fibrosis (CF) that differs widely in age and disease status. The aim of this study was to develop a CF-specific, individualized approach to determine workload increments for a cycle ergometry testing protocol. MethodsA total of 409 assessments consisting of maximal exercise data, anthropometric parameters, and lung function measures from 160 children and adolescents with CF were examined. 90% of the database was analyzed with backward linear regression with peak workload (Wpeak) as the dependent variable. Afterwards, we [1] used the remaining 10% of the database (model validation group) to validate the model's capacity to predict Wpeak and [2] validated the protocol's ability to provide a maximal effort within a 10±2minute time frame in 14 adolescents with CF who were tested using this new protocol (protocol validation group). ResultsNo significant differences were seen in Wpeak and predicted Wpeak in the model validation group or in the protocol validation group. Eight of 14 adolescents with CF in the protocol validation group performed a maximal effort, and seven of them terminated the test within the 10±2minute time frame. Backward linear regression analysis resulted in the following equation: Wpeak (W)=−142.865+2.998×Age (years)−19.206×Sex (0=male; 1=female)+1.328×Height (cm)+23.362×FEV1 (L) (R=.89; R2=.79; SEE=21). Bland–Altman analysis showed no systematic bias between the actual and predicted Wpeak. ConclusionWe developed a CF-specific linear regression model to predict peak workload based on standard measures of anthropometry and FEV1, which could be used to calculate individualized workload increments for a cycle ergometry testing protocol.

Limited Sampling Models and Bayesian Estimation for Mycophenolic Acid Area under the Curve Prediction in Stable Renal Transplant Patients Co-Medicated with Ciclosporin or Sirolimus

Mycophenolate mofetil is a prodrug of mycophenolic acid (MPA), an immunosuppressive agent used in combination with corticosteroids and calcineurin inhibitors or sirolimus for the prevention of acute rejection after solid organ transplantation. Although MPA has a rather narrow therapeutic window and its pharmacokinetics show considerable intra- and interindividual variability, dosing guidelines recommend a standard dosage regimen of 0.5-1.0 g twice daily in adult renal, liver and cardiac transplant recipients. The main objective of the present study was to develop a method to predict the MPA area under the plasma concentration-time curve during one 12-hour dosing interval (AUC(12)) by using multiple linear regression models and maximum a posteriori (MAP) Bayesian estimation methods in patients co-medicated with ciclosporin or sirolimus, aiming to individualize the dosage regimen of mycophenolate mofetil. Pharmacokinetic profiles of MPA and mycophenolic acid glucuronide (MPAG), the main metabolite of MPA, were obtained from 40 stable adult renal allograft recipients on three different occasions: the day before switching from ciclosporin to sirolimus co-medication (+/-7.4 months post-transplantation; period I) and at 60 days and 270 days after the switch (periods II and III). Blood samples for determination of MPA and MPAG concentrations in plasma were taken at 0 hours (pre-dose) and at 0.33, 0.66, 1.25, 2, 4, 6, 8 and 12 hours after oral intake of mycophenolate mofetil. The MPA AUC(12) was calculated by the trapezoidal method (the observed AUC(12)). Patients were randomly divided into (i) a model-building test group (n = 27); and (ii) a model-validation group (n = 13). Multiple linear regression models were developed, based on three sampling times after drug administration. Subsequently, a population pharmacokinetic model describing MPA and MPAG plasma concentrations was developed using nonlinear mixed-effects modelling and a Bayesian estimator based on the population pharmacokinetic model was used to predict the MPA AUC(12) based on three sampling times taken within 2 hours following dosing. Fifty-two percent of the observed AUC(12) values (three periods) in the 40 patients receiving a fixed dose of mycophenolate mofetil 750 mg twice daily were outside the recommended therapeutic range (30-60 microg x h/mL). The failure of the standard dose to yield an AUC(12) value within the therapeutic range was especially pronounced during the first study period. Of the multiple linear regression models that were tested, the equation based on the 0-hour (pre-dose), 0.66- and 2-hour sampling times showed the best predictive performance in the validation group: r2 = 0.79, relative root mean square error (rRMSE) = 14% and mean relative prediction error (MRPE) = 0.9%. The pharmacokinetics of MPA and MPAG were best described by a two-compartment model with first-order absorption and elimination for MPA, plus a compartment for MPAG, also including a gastrointestinal compartment and enterohepatic cycling in the case of sirolimus co-medication. The ratio of aminotransferase liver enzymes (AST and ALT) and the glomerular filtration rate significantly influenced MPA glucuronidation and MPAG renal excretion, respectively. Bayesian estimation of the MPA AUC(12) based on 0-, 1.25- and 2-hour sampling times predicted the observed AUC(12) values of the patients in the validation group, with the following predictive performance characteristics: r2 = 0.93, rRMSE = 12.4% and MRPE = -0.4%. Use of the developed multiple linear regression equation and Bayesian estimator, both based on only three blood sampling times within 2 hours following a dose of mycophenolate mofetil, allowed an accurate prediction of a patient's MPA AUC(12) for therapeutic drug monitoring and dose individualization. These findings should be validated in a randomized prospective trial.

Body Segment Parameter Estimation of the Human Lower Leg Using an Elliptical Model with Validation from DEXA

Accurate estimates of human body segment parameters (BSPs) are required for kinetic analyses of motion. The purpose of this study was to develop a geometric model of the human lower leg based on the mass distribution properties of the segment. Forty subjects were recruited from 4 human populations. Each population was randomly divided equally into model development (MD) and model validation (MV) groups. Participants underwent frontal and sagittal plane dual energy X-ray absorptiometry (DEXA) scans and anthropometric measurements. Leg BSPs were calculated from the scan information and mass distribution properties in the two planes were determined. Further, a geometric model was developed based on the ensemble averages of the mass distribution information from the MD groups. The model was applied to the MV groups and mean absolute errors were calculated for each BSP and each population. Finally, BSP estimates from literature sources were also determined and compared against DEXA. The model developed produced the lowest errors overall. Additionally, the results showed that the model developed estimated BSPs for all four populations with consistent accuracy whereas the other 4 models tested provided different levels of accuracy depending on the age and gender categories of the group tested. The results of this study present a model that accurately estimates BSPs of the lower leg for individuals varying in age, gender, race, and morphology. This study also presents a modelling technique that may successfully provide similar results for other body segments.

Predictors of Long-term Mortality in Patients With Cirrhosis of the Liver Admitted to a Medical ICU

The long-term survival of patients with cirrhosis of the liver admitted to the ICU has not been described.The main objectives were to determine the 1-year and 5-year mortality rates of a cohort of patients admitted to a medical ICU (MICU), and to identify the risk factors that may predict long-term outcomes.This is a cohort study. We used a model-building (MB) and model validation (MV) procedure that has previously been described to determine the risk factors for overall mortality.An MICU in a major referral medical center.Four hundred twenty consecutive patients admitted to the ICU from January 1, 1993, through October 31, 1998, met the inclusion criteria of diagnosis of liver failure, cirrhosis, chronic liver disease, variceal bleeding, hepatic encephalopathy, or hepatorenal syndrome. Patients with acute liver failure who had undergone liver transplantation, or candidates for orthotopic liver transplantation were excluded.None.The 1-year mortality rate was 69%, and the 5-year mortality rate was 77%. The median survival time was 1 month. The independent predictors of mortality in patients in the MB group who retained their significance in the MV group were as follows: an acute physiology, age, and chronic health evaluation (APACHE) III score of >/= 90 (hazard ratio [HR], 2.2; 95% confidence interval [CI], 1.6 to 2.8; p < 0.0001), the use of pressors (HR, 2.5; 95% CI, 1.9 to 3.2; p < 0.0001), and jaundice (HR, 1.7; 95% CI, 1.4 to 2.2; p < 0.0001). Patients with all three risk factors (ie, APACHE III score >/= 90, use of pressors, and jaundice) had a 92% 1-month mortality rate compared to 11.2% for patients with no risk factors.Patients admitted to an MICU with underlying cirrhosis who are not eligible for liver transplantation have a poor long-term prognosis, even if they survive the ICU admission, particularly as the number of risk factors increases.

Total-body skeletal muscle mass: estimation by a new dual-energy X-ray absorptiometry method

Skeletal muscle (SM) is an important body-composition component that remains difficult and impractical to quantify by most investigators outside of specialized research centers. A large proportion of total-body SM is found in the extremities, and a large proportion of extremity lean soft tissue is SM. A strong link should thus exist between appendicular lean soft tissue (ALST) mass and total-body SM mass. The objective was to develop prediction models linking ALST estimated by dual-energy X-ray absorptiometry (DXA) with total-body SM quantified by multislice magnetic resonance imaging in healthy adults. ALST and total-body SM were evaluated with a cross-sectional design in adults [body mass index (in kg/m(2)) < 35] with an SM-prediction model developed and validated in model-development and model-validation groups, respectively. The model-development and model-validation groups included 321 and 93 ethnically diverse adults, respectively. ALST alone was highly correlated with total-body SM (model 1: R(2) = 0.96, SEE = 1.63 kg, P < 0.001), although multiple regression analyses showed 2 additional predictor variables: age (model 2: 2-variable combined R(2) = 0.96, SEE = 1.58 kg, P < 0.001) and sex (model 3: 3-variable combined R(2) = 0.96, SEE = 1.58 kg, P < 0.001). All 3 models performed well in the validation group. An SM-prediction model based on the SM-ALST ratio was also developed, although this model had limitations when it was applied across all subjects. Total-body SM can be accurately predicted from DXA-estimated ALST, thus affording a practical means of quantifying the large and clinically important SM compartment.

Predictors of Mortality and Resource Utilization in Cirrhotic Patients Admitted to the Medical ICU

Cirrhotic patients admitted to the medical ICU (MICU) are associated with high mortality rates and high resource utilization. This study identifies specific predictors of increased mortality and resource utilization and uses them to develop and validate prognostic models in cirrhotic patients admitted to the MICU. Cirrhotic patients admitted to the MICU were identified from the Critical Care Section database (January 1993 to October 1998). Clinical data were extracted from chart review including hospital course variables, mortality, and length of stay (LOS). Total cost per case (TCPC) was obtained from the Transition System INC: Multivariate logistic and linear regression analyses identified the independent predictors of increased mortality and resource utilization used for model building (MB) and model validation (MV). A total of 582 cases were randomized to the MB and MV groups. Each group contained 240 cases after exclusion criteria were applied. The MICU mortality rate was 36.6%, and the in-hospital mortality rate was 49.0%. Acute physiology, age, and chronic health evaluation (APACHE) III score (odds ratio [OR], 4.7; 95% confidence interval [CI], 2.70 to 8.16; p < 0.001), mechanical ventilation (OR, 4.57; 95% CI, 2.35 to 8.34); p < 0.001), and the use of pressors (OR, 7.57; 95% CI, 4.35 to 13.18; p < 0.001) were independent predictors of MICU mortality. APACHE III score (OR, 4.96; 95% CI, 2.97 to 8.29; p < 0.001), the use of pressors (OR, 6.55; 95% CI, 3.66 to 11.72; p < 0.001), and acute renal failure (ARF) (OR, 4.31; 95% CI, 2.41 to 7.71; p < 0.001) were independent predictors of in-hospital mortality. Increased LOS in the MICU was associated with mechanical ventilation, ARF, bronchoscopy, bacteremia, use of pressors, transjugular intrahepatic portosystemic shunt (TIPS), and never received cardiopulmonary resuscitation (CPR) (p < 0.005). Source of admission, platelet transfusion, bacteremia, pneumonia, and never received CPR were independently associated with increased total LOS (p < 0.001). Mechanical ventilation, platelet transfusion, bronchoscopy, TIPS, sepsis, and never received CPR were independent predictors of increased TCPC (p < 0.001). Simple prognostic models for mortality and resource utilization have been developed for cirrhotic patients admitted to the MICU.

Model reference adaptive control of heart rate during wheelchair ergometry

This study used parametric methods to identify a model of the wheelchair user's heart rate response to changes in physical workload while on a wheelchair dynamometer. A model reference adaptive control (MRAC) algorithm was developed based upon the results of the system identification process. Based upon subject's maximum speed, ten random levels from zero to maximal speed were selected to form a tracking input vector. Two autoregressive models with exogenous inputs were developed using a stepwise regression-based algorithm. Analysis of variance results imply that the model development and model validation groups were different. Linear regression was used to compare the model predicted heart rates with recorded heart rates for the validation group. The application of indirect MRAC appears to be feasible for controlling heart rate kinetics of people with paraplegia or lower limb impairments during wheelchair propulsion on a computer-controlled wheelchair dynamometer.