Fecal Transplant for Treatment of Toxic Megacolon Associated With Clostridium Difficile Colitis in a Patient With Duchenne Muscular Dystrophy.

Clostridium difficile (C diff) colitis infection is the most common cause of nosocomial infectious diarrhea and the prevalence is increasing worldwide. Toxic megacolon is a severe complication of C diff colitis associated with high mortality. Gastrointestinal (GI) comorbidity and impaired smooth muscle contraction are risk factors for the development of C diff-associated toxic megacolon. We present a case of fulminant C diff colitis with toxic megacolon in a patient with Duchenne muscular dystrophy (DMD) in the intensive care unit. C diff colitis was diagnosed by clinical presentation and positive C diff DNA amplification test (polymerase chain reaction). The impairment of GI tract due to DMD predisposes these patients to severe C diff infection and toxic megacolon, as observed in this case report. For the same reason, the recovery of GI function in these patients can be prolonged. While surgery was conducted for relieving the pressure from toxic megacolon, fecal microbiota transplantation through colonoscopy resulted in successful resolution of the C diff symptoms, although the recovery is prolonged due to DMD.

Empyema due to severe Pasteurella multocida treated with quinolone: a case report.

Pasteurella multocida (PM) is gram-negative coccobacillus that is commonly acquired through a cat scratch or bite. The standard antibiotic of choice for treating PM is penicillin. We present a case of PM empyema in a penicillin allergic 56-year-old female who has a history of contact with domestic cats and is known to have chronic obstructive pulmonary disease and a chronic history of smoking. The patient was treated successfully with Levofloxacin as alternative treatment to penicillin.

[Study on three-dimension spatial variability of regional soil salinity based on spectral indices].

In order to illustrate the three-dimension spatial variability of soil salinity in central China flood area of the Yellow river, integrated soil sampling data and remote sensing data, spectral indices and inverse distance weighting (IDW) method were applied to the estimation and simulation of three-dimension spatial distribution of soil salinity. The study was carried out in typical central China flood area of the Yellow river in Fengqiu County, Henan Province, China. The electrical conductivity of the saturation extract (EC1: 5) of 505 soil samples collected at 101 points was measured. The results indicated that the coefficient of variation of soil salinity at each soil layer is from 0.218 to 0.324 and exhibited the moderate spatial variability. The average of soil electrical conductivity is from 0.121 to 0.154 ds x m(-1). The 2 820 three-dimension spatial scattered data for soil electrical conductivity were taken at soil salinity mapping interpreted by spectral indices and soil electrical conductivity. Three-dimension IDW interpolation showed that a large area of high soil salinity mainly located in the region of Tianran canal and the along of the Yellow river. The shape of the soil salinity profile was downward flowed, revealing soil salinity increasing with depth in whole soil profile and soil salinity accumulated in the subsoil. The accuracy of the predictions was tested using 20 soil sampled points. The root mean square error (RMSE) of calibration for three-dimension distribution of soil salinity showed that the IDW method based on spectral indices was ideal. The research results can provide theoretical foundations to the management and utilization of salt-affected land in China flood area, especially in the Yellow river zone.

Functional genetic variants within the SIRT2 gene promoter in type 2 diabetes mellitus.

AIMS: Type 2 diabetes mellitus (T2D) is a common and complex metabolic diseases caused by interactions between environmental and genetic factors. Genome-wide association studies have identified more than 80 common genetic variants for T2D, which account for only ∼10% of the heritability of T2D cases. SIRT2, a member of NAD(+)-dependent class III deacetylases, is involved in genomic stability, metabolism, inflammation, oxidative stress and autophagy. In maintaining metabolic homeostasis, SIRT2 regulates adipocyte differentiation, fatty acid oxidation, gluconeogenesis, and insulin sensitivity. Thus, we hypothesized that DNA sequence variants (DSVs) in SIRT2 gene promoter may change SIRT2 levels, contributing to T2D. METHODS: SIRT2 gene promoter was genetically and functionally analyzed in large cohorts of T2D patients (n = 365) and ethnic-matched controls (n = 358). RESULTS: A total of 18 DSVs, including 5 SNPs, were identified in this study. Four novel heterozygous DSVs (g.38900912G > T, g.38900561C > T, g.38900359C > T and g.38900237G > A) were identified in four T2D patients, three of which (g.38900912G > T, g.38900359C > T and g.38900237G > A) significantly increased the transcriptional activity of the SIRT2 gene promoter in cultured pancreatic beta cells (P < .01). Seven novel heterozygous DSVs were only found in controls, and one heterozygous deletion DSV and five SNPs were found in both T2D patients and controls, which did not significantly affect SIRT2 gene promoter activity (P > .05). CONCLUSIONS: Our findings suggested that the DSVs may increase SIRT2 gene promoter activity and SIRT2 levels, contributing to T2D development as a risk factor.

Multiplicative Multitask Feature Learning.

We investigate a general framework of multiplicative multitask feature learning which decomposes individual task's model parameters into a multiplication of two components. One of the components is used across all tasks and the other component is task-specific. Several previous methods can be proved to be special cases of our framework. We study the theoretical properties of this framework when different regularization conditions are applied to the two decomposed components. We prove that this framework is mathematically equivalent to the widely used multitask feature learning methods that are based on a joint regularization of all model parameters, but with a more general form of regularizers. Further, an analytical formula is derived for the across-task component as related to the task-specific component for all these regularizers, leading to a better understanding of the shrinkage effects of different regularizers. Study of this framework motivates new multitask learning algorithms. We propose two new learning formulations by varying the parameters in the proposed framework. An efficient blockwise coordinate descent algorithm is developed suitable for solving the entire family of formulations with rigorous convergence analysis. Simulation studies have identified the statistical properties of data that would be in favor of the new formulations. Extensive empirical studies on various classification and regression benchmark data sets have revealed the relative advantages of the two new formulations by comparing with the state of the art, which provides instructive insights into the feature learning problem with multiple tasks.

Predicting readmission risk with institution-specific prediction models.

OBJECTIVE: The ability to predict patient readmission risk is extremely valuable for hospitals, especially under the Hospital Readmission Reduction Program of the Center for Medicare and Medicaid Services which went into effect starting October 1, 2012. There is a plethora of work in the literature that deals with developing readmission risk prediction models, but most of them do not have sufficient prediction accuracy to be deployed in a clinical setting, partly because different hospitals may have different characteristics in their patient populations. METHODS AND MATERIALS: We propose a generic framework for institution-specific readmission risk prediction, which takes patient data from a single institution and produces a statistical risk prediction model optimized for that particular institution and, optionally, for a specific condition. This provides great flexibility in model building, and is also able to provide institution-specific insights in its readmitted patient population. We have experimented with classification methods such as support vector machines, and prognosis methods such as the Cox regression. We compared our methods with industry-standard methods such as the LACE model, and showed the proposed framework is not only more flexible but also more effective. RESULTS: We applied our framework to patient data from three hospitals, and obtained some initial results for heart failure (HF), acute myocardial infarction (AMI), pneumonia (PN) patients as well as patients with all conditions. On Hospital 2, the LACE model yielded AUC 0.57, 0.56, 0.53 and 0.55 for AMI, HF, PN and All Cause readmission prediction, respectively, while the proposed model yielded 0.66, 0.65, 0.63, 0.74 for the corresponding conditions, all significantly better than the LACE counterpart. The proposed models that leverage all features at discharge time is more accurate than the models that only leverage features at admission time (0.66 vs. 0.61 for AMI, 0.65 vs. 0.61 for HF, 0.63 vs. 0.56 for PN, 0.74 vs. 0.60 for All Cause). Furthermore, the proposed admission-time models already outperform the performance of LACE, which is a discharge-time model (0.61 vs. 0.57 for AMI, 0.61 vs. 0.56 for HF, 0.56 vs. 0.53 for PN, 0.60 vs. 0.55 for All Cause). Similar conclusions can be drawn from other hospitals as well. The same performance comparison also holds for precision and recall at top-decile predictions. Most of the performance improvements are statistically significant. CONCLUSIONS: The institution-specific readmission risk prediction framework is more flexible and more effective than the one-size-fit-all models like the LACE, sometimes twice and three-time more effective. The admission-time models are able to give early warning signs compared to the discharge-time models, and may be able to help hospital staff intervene early while the patient is still in the hospital.

Matrix metalloproteinases in atherosclerosis: role of nitric oxide, hydrogen sulfide, homocysteine, and polymorphisms.

Atherosclerosis is an inflammatory process that involves activation of matrix metalloproteinases (MMPs); MMPs degrade collagen and allow for smooth-muscle cell migration within a vessel. Moreover, this begets an accumulation of other cellular material, resulting in occlusion of the vessel and ischemic events to tissues in need of nutrients. Homocysteine has been shown to activate MMPs via an increase in oxidative stress and acting as a signaling molecule on receptors like the peroxisome proliferator activated receptor-γ and N-methyl-D-aspartate receptor. Nitric oxide has been shown to be beneficial in some cases of deactivating MMPs. However, in other cases, it has been shown to be harmful. Further studies are warranted on the scenarios that are beneficial versus destructive. Hydrogen sulfide (H2S) has been shown to decrease MMP activities in all cases in the literature by acting as an antioxidant and vasodilator. Various MMP-knockout and gene-silencing models have been used to determine the function of the many different MMPs. This has allowed us to discern the role that each MMP has in promoting or alleviating pathological conditions. Furthermore, there has been some study into the MMP polymorphisms that exist in the population. The purpose of this review is to examine the role of MMPs and their polymorphisms on the development of atherosclerosis, with emphasis placed on pathways that involve nitric oxide, hydrogen sulfide, and homocysteine.

Regional and physician specialty-associated variations in the medical management of atherosclerotic renal-artery stenosis.

For people enrolled in Cardiovascular Outcomes in Renal Atherosclerotic Lesions (CORAL), we sought to examine whether variation exists in the baseline medical therapy of different geographic regions and if any variations in prescribing patterns were associated with physician specialty. Patients were grouped by location within the United States (US) and outside the US (OUS), which includes Canada, South America, Europe, South Africa, New Zealand, and Australia. When comparing US to OUS, participants in the US took fewer anti-hypertensive medications (1.9 ± 1.5 vs. 2.4 ± 1.4; P < .001) and were less likely to be treated with an angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker (46% vs. 62%; P < .001), calcium channel antagonist (37% vs. 58%; P < .001), and statin (64% vs. 75%; P < .05). In CORAL, the identification of variations in baseline medical therapy suggests that substantial opportunities exist to improve the medical management of patients with atherosclerotic renal-artery stenosis.

Caecum perforation after renal transplantation: a case report and review of literature.

Gastrointestinal (GI) complication used to be the second most common complication in renal transplant patients after infection (Bardaxoglou et al. in Transpl Int 6(3):148-152, 1993). Review of transplant registry reveals that GI complication is no longer the second most common type of complication after renal transplant, but that it is still a common cause of significant amount of deaths in renal transplant recipients (De Bartolomeis et al. in Transpl Proc 37(6):2504-2506, 2005). In a study of 1,515 adults with severe GI complication after renal transplant, Sarkio et al. (Transpl Int 17(9):505-510, 2004) reported that gastroduodenal ulcers followed by colon perforation were the two biggest groups of GI complications during the first year after renal transplantation. Colonic perforation is estimated to occur in about 1 % of all cases of renal transplant patients, and it does predispose to potentially fatal complication. About 50 % of all colonic perforation is due to complication of acute inflammation of diverticular disease (Bardaxoglou et al. in Transpl Int 6(3):148-152, 1993; Guice et al. in Am J Surg 138(1):43-48, 1979; Koneru et al. in Arch Surg 125(5):610-613, 1990; Coccolini et al. in Transpl Proc 41(4):1189-1190, 2009). This is particularly so because these patients were previously exposed to uremia before transplantation which alters their protein metabolism hence interfering with tissue healing there after (Carson et al. in Ann Surg 188(1):109-113, 1978). GI complications including colon perforation after renal transplantation have effect on a patient's long-term survival (Gil-Vernet et al. in Transpl Proc 39(7):2190-2193, 2007). Despite this, the role of renal transplantation medication compared to anatomic anomaly in GI complication has been equivocal.

Another cause of chest pain: Staphylococcus aureus sternal osteomyelitis in an otherwise healthy adult.

Chest pain requires a detailed differential diagnosis with good history-taking skills to differentiate between cardiogenic and noncardiogenic causes. Moreover, when other symptoms such as fever and elevated white blood cell count are involved, it may be necessary to consider causes that include infectious sources. A 53-year-old female with no significant past medical history returned to the hospital with recurrent complaints of chest pain that was constant, substernal, reproducible, and exacerbated with inspiration and expiration. The chest pain was thought to be noncardiogenic, as electrocardiography did not demonstrate changes, and cardiac enzymes were found to be negative for signs of ischemia. The patient's blood cultures were analyzed from a previous admission and were shown to be positive for Staphylococcus aureus. The patient was started empirically on vancomycin, which was later switched to ceftriaxone as the bacteria were more sensitive to this antibiotic. A transthoracic echocardiogram did not demonstrate any vegetation or signs of endocarditis. There was a small right pleural effusion discovered on X-ray. Therefore, computed tomography as well as magnetic resonance imaging of the chest were performed, and showed osteomyelitis of the chest. The patient was continued on intravenous ceftriaxone for a total of 6 weeks. Tests for HIV, hepatitis A, B, and C were all found to be negative. The patient had no history of childhood illness, recurrent infections, or previous trauma to the chest, and had had no recent respiratory infections, pneumonia, or any underlying lung condition. Hence, her condition was thought to be a case of primary sternal osteomyelitis without known cause.

Recurrent myocardial infarctions in a young football player secondary to thrombophilia, associated with elevated factor VIII activity.

Myocardial infarction (MI) due to coronary atherosclerosis in young adults is uncommon; rare causes such as cocaine abuse, arterial dissection, and thromboembolism should be considered. A 21-year-old football player, and otherwise healthy African American man, developed chest pain during exercise while bench-pressing 400 lbs. Acute MI was diagnosed based on physical examination, electrocardiography findings, and elevated cardiac enzymes. Coronary arteriography showed a thrombus occluding the proximal left anterior descending artery (LAD). Aggressive antiplatelet therapy with aspirin, clopidogrel, and eptifibatide was pursued, in addition to standard post-MI care. This led to the successful resolution of symptoms and dissolution of the thrombus, demonstrated by repeat coronary arteriography. Five months later, he presented with similar symptoms during exercise after lifting heavy weights, and was found to have another acute MI. Coronary arteriography again showed a thrombus occluding the LAD. No evidence of coronary artery dissection or vasospasm was found. Only mild atherosclerotic plaque burden was observed on both occasions by intravascular ultrasound. A bare metal stent was placed at the site as it was thought this site had acted as a nidus for small plaque rupture and thrombus formation. Elevated serum factor VIII activity at 205% (reference range 60%-140%) was found, a rare cause of hypercoagulability. Further workup revealed a patent foramen ovale during a Valsalva maneuver by transesophageal echocardiography. Both events occurred during weight lifting, which can transiently increase right heart pressure in a similar way to the Valsalva maneuver. In light of all the findings, we concluded that an exercise-related increase in factor VIII activity led to coronary arterial thrombosis in the presence of a small ruptured plaque. Alternatively, venous clots may have traversed the patent foramen ovale and occluded the LAD. In addition to continuing aggressive risk factor modification, anticoagulation therapy with warfarin was initiated with close follow-up.

Categorizing medications from unstructured clinical notes.

One of the important pieces of information in a patient's clinical record is the information about their medications. Besides administering information, it also consists of the category of the medication i.e. whether the patient was taking these medications at Home, were administered in the Emergency Department, during course of stay or on discharge etc. Unfortunately, much of this information is presently embedded in unstructured clinical notes e.g. in ER records, History & Physical documents etc. This information is required for adherence to quality and regulatory guidelines or for retrospective analysis e.g. CMS reporting. It is a manually intensive process to extract such information. This paper explains in detail a statistical NLP system developed to extract such information. We have trained a Maximum Entropy Markov model to categorize instances of medication names into previously defined categories. The system was tested on a variety of clinical notes from different institutions and we achieved an average accuracy of 91.3%.

A new, accurate predictive model for incident hypertension.

OBJECTIVE: Data mining represents an alternative approach to identify new predictors of multifactorial diseases. This work aimed at building an accurate predictive model for incident hypertension using data mining procedures. METHODS: The primary study population consisted of 1605 normotensive individuals aged 20-79 years with 5-year follow-up from the population-based study, that is the Study of Health in Pomerania (SHIP). The initial set was randomly split into a training and a testing set. We used a probabilistic graphical model applying a Bayesian network to create a predictive model for incident hypertension and compared the predictive performance with the established Framingham risk score for hypertension. Finally, the model was validated in 2887 participants from INTER99, a Danish community-based intervention study. RESULTS: In the training set of SHIP data, the Bayesian network used a small subset of relevant baseline features including age, mean arterial pressure, rs16998073, serum glucose and urinary albumin concentrations. Furthermore, we detected relevant interactions between age and serum glucose as well as between rs16998073 and urinary albumin concentrations [area under the receiver operating characteristic (AUC 0.76)]. The model was confirmed in the SHIP validation set (AUC 0.78) and externally replicated in INTER99 (AUC 0.77). Compared to the established Framingham risk score for hypertension, the predictive performance of the new model was similar in the SHIP validation set and moderately better in INTER99. CONCLUSION: Data mining procedures identified a predictive model for incident hypertension, which included innovative and easy-to-measure variables. The findings promise great applicability in screening settings and clinical practice.

Leveraging rich annotations to improve learning of medical concepts from clinical free text.

Information extraction from clinical free text is one of the key elements in medical informatics research. In this paper we propose a general framework to improve learning-based information extraction systems with the help of rich annotations (i.e., annotators provide the medical assertion as well as evidences that support the assertion). A special graphical interface was developed to facilitate the annotation process, and we show how to implement this framework with a state-of-the-art context-based question answering system. Empirical studies demonstrate that with about 10% longer annotation time, we can significantly improve the accuracy of the system. An approach to provide supporting evidence for test documents is also briefly discussed with promising preliminary results.

Sparse classification for computer aided diagnosis using learned dictionaries.

Classification is one of the core problems in computer-aided cancer diagnosis (CAD) via medical image interpretation. High detection sensitivity with reasonably low false positive (FP) rate is essential for any CAD system to be accepted as a valuable or even indispensable tool in radiologists' workflow. In this paper, we propose a novel classification framework based on sparse representation. It first builds an overcomplete dictionary of atoms for each class via K-SVD learning, then classification is formulated as sparse coding which can be solved efficiently. This representation naturally generalizes for both binary and multiwise classification problems, and can be used as a standalone classifier or integrated with an existing decision system. Our method is extensively validated in CAD systems for both colorectal polyp and lung nodule detection, using hospital scale, multi-site clinical datasets. The results show that we achieve superior classification performance than existing state-of-the-arts, using support vector machine (SVM) and its variants, boosting, logistic regression, relevance vector machine (RVM), or kappa-nearest neighbor (KNN).

Development and validation of a prognostic model using blood biomarker information for prediction of survival of non-small-cell lung cancer patients treated with combined chemotherapy and radiation or radiotherapy alone (NCT00181519, NCT00573040, and NCT00572325).

PURPOSE: Currently, prediction of survival for non-small-cell lung cancer patients treated with (chemo)radiotherapy is mainly based on clinical factors. The hypothesis of this prospective study was that blood biomarkers related to hypoxia, inflammation, and tumor load would have an added prognostic value for predicting survival. METHODS AND MATERIALS: Clinical data and blood samples were collected prospectively (NCT00181519, NCT00573040, and NCT00572325) from 106 inoperable non-small-cell lung cancer patients (Stages I-IIIB), treated with curative intent with radiotherapy alone or combined with chemotherapy. Blood biomarkers, including lactate dehydrogenase, C-reactive protein, osteopontin, carbonic anhydrase IX, interleukin (IL) 6, IL-8, carcinoembryonic antigen (CEA), and cytokeratin fragment 21-1, were measured. A multivariate model, built on a large patient population (N = 322) and externally validated, was used as a baseline model. An extended model was created by selecting additional biomarkers. The model's performance was expressed as the area under the curve (AUC) of the receiver operating characteristic and assessed by use of leave-one-out cross validation as well as a validation cohort (n = 52). RESULTS: The baseline model consisted of gender, World Health Organization performance status, forced expiratory volume, number of positive lymph node stations, and gross tumor volume and yielded an AUC of 0.72. The extended model included two additional blood biomarkers (CEA and IL-6) and resulted in a leave-one-out AUC of 0.81. The performance of the extended model was significantly better than the clinical model (p = 0.004). The AUC on the validation cohort was 0.66 and 0.76, respectively. CONCLUSIONS: The performance of the prognostic model for survival improved markedly by adding two blood biomarkers: CEA and IL-6.

Automated identification of medical concepts and assertions in medical text.

This paper describes a machine learning, text processing approach that allows the extraction of key medical information from unstructured text in Electronic Medical Records. The approach utilizes a novel text representation that shares the simplicity of the widely used bag-of-words representation, but can also represent some form of semantic information in the text. The large dimensionality of this type of learning models is controlled by the use of a ℓ(1) regularization to favor parsimonious models. Experimental results demonstrate the accuracy of the approach in extracting medical assertions that can be associated to polarity and relevance detection.

The importance of patient characteristics for the prediction of radiation-induced lung toxicity.

PURPOSE: Extensive research has led to the identification of numerous dosimetric parameters as well as patient characteristics, associated with lung toxicity, but their clinical usefulness remains largely unknown. We investigated the predictive value of patient characteristics in combination with established dosimetric parameters. PATIENTS AND METHODS: Data from 438 lung cancer patients treated with (chemo)radiation were used. Lung toxicity was scored using the Common Toxicity Criteria version 3.0. A multivariate model as well as two single parameter models, including either V(20) or MLD, was built. Performance of the models was expressed as the AUC (Area Under the Curve). RESULTS: The mean MLD was 13.5 Gy (SD 4.5 Gy), while the mean V(20) was 21.0% (SD 7.3%). Univariate models with V(20) or MLD both yielded an AUC of 0.47. The final multivariate model, which included WHO-performance status, smoking status, forced expiratory volume (FEV(1)), age and MLD, yielded an AUC of 0.62 (95% CI: 0.55-0.69). CONCLUSIONS: Within the range of radiation doses used in our clinic, dosimetric parameters play a less important role than patient characteristics for the prediction of lung toxicity. Future research should focus more on patient-related factors, as opposed to dosimetric parameters, in order to identify patients at high risk for developing radiation-induced lung toxicity more accurately.

Dyspnea evolution after high-dose radiotherapy in patients with non-small cell lung cancer.

PURPOSE: To determine what the influence is of dyspnea (CTCAE3.0) before high-dose radiotherapy (RT) on the incidence and severity of subsequent lung toxicity in patients with non-small cell lung cancer (NSCLC). METHODS: In 197 patients with stage I-III NSCLC maximal dyspnea scores (CTCAE3.0) were obtained prospectively at three time periods: before RT, the first 6 months post-RT and 6-9 months post-RT. Only patients who were clinically progression-free 12 months or more after RT were included, thus minimizing dyspnea due to tumor progression. Time-trends of dyspnea as a function of baseline dyspnea were investigated and correlated with gender, age, chemotherapy, mean lung dose (MLD), lung function parameters (FeV1 and DLCO), stage, PTV dose, overall treatment time and smoking habits. RESULTS: The proportion developing less, the same or more dyspnea 6-9 months post-treatment according to their baseline dyspnea scores was: Grade 0: none, 82.9%, 17.1%; Grade 1: 21.2%, 51.9%, 26.9%; Grade 2: 27.3%, 54.5%, 18.2%, respectively. Only age was associated with increased dyspnea after RT. CONCLUSIONS: Patients with dyspnea before therapy have a realistic chance to improve after high-dose radiotherapy. Reporting only dyspnea at one time-point post-RT is insufficient to determine radiation-induced dyspnea.

Development and external validation of prognostic model for 2-year survival of non-small-cell lung cancer patients treated with chemoradiotherapy.

PURPOSE: Radiotherapy, combined with chemotherapy, is the treatment of choice for a large group of non-small-cell lung cancer (NSCLC) patients. Recent developments in the treatment of these patients have led to improved survival. However, the clinical TNM stage is highly inaccurate for the prediction of survival, and alternatives are lacking. The objective of this study was to develop and validate a prediction model for survival of NSCLC patients, treated with chemoradiotherapy. PATIENTS AND METHODS: The clinical data from 377 consecutive inoperable NSCLC patients, Stage I-IIIB, treated radically with chemoradiotherapy were collected. A prognostic model for 2-year survival was developed, using 2-norm support vector machines. The performance of the model was expressed as the area under the curve of the receiver operating characteristic and assessed using leave-one-out cross-validation, as well as two external data sets. RESULTS: The final multivariate model consisted of gender, World Health Organization performance status, forced expiratory volume in 1 s, number of positive lymph node stations, and gross tumor volume. The area under the curve, assessed by leave-one-out cross-validation, was 0.74, and application of the model to the external data sets yielded an area under the curve of 0.75 and 0.76. A high- and low-risk group could be clearly identified using a risk score based on the model. CONCLUSION: The multivariate model performed very well and was able to accurately predict the 2-year survival of NSCLC patients treated with chemoradiotherapy. The model could support clinicians in the treatment decision-making process.