Candidemia Risk Prediction (CanDETEC) Model for Patients With Malignancy: Model Development and Validation in a Single-Center Retrospective Study.
Ontology highlight
ABSTRACT: Background
Appropriate empirical treatment for candidemia is associated with reduced mortality; however, the timely diagnosis of candidemia in patients with sepsis remains poor.Objective
We aimed to use machine learning algorithms to develop and validate a candidemia prediction model for patients with cancer.Methods
We conducted a single-center retrospective study using the cancer registry of a tertiary academic hospital. Adult patients diagnosed with malignancies between January 2010 and December 2018 were included. Our study outcome was the prediction of candidemia events. A stratified undersampling method was used to extract control data for algorithm learning. Multiple models were developed-a combination of 4 variable groups and 5 algorithms (auto-machine learning, deep neural network, gradient boosting, logistic regression, and random forest). The model with the largest area under the receiver operating characteristic curve (AUROC) was selected as the Candida detection (CanDETEC) model after comparing its performance indexes with those of the Candida Score Model.Results
From a total of 273,380 blood cultures from 186,404 registered patients with cancer, we extracted 501 records of candidemia events and 2000 records as control data. Performance among the different models varied (AUROC 0.771- 0.889), with all models demonstrating superior performance to that of the Candida Score (AUROC 0.677). The random forest model performed the best (AUROC 0.889, 95% CI 0.888-0.889); therefore, it was selected as the CanDETEC model.Conclusions
The CanDETEC model predicted candidemia in patients with cancer with high discriminative power. This algorithm could be used for the timely diagnosis and appropriate empirical treatment of candidemia.
SUBMITTER: Yoo J
PROVIDER: S-EPMC8367162 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:BackgroundDermatomyositis accompanied with malignancy is a common poor prognostic factor of dermatomyositis. Thus, the early prediction of the risk of malignancy in patients with dermatomyositis can significantly improve the prognosis of patients. However, the identification of antibodies related to malignancy in dermatomyositis patients has not been widely implemented in clinical practice. Herein, we established a predictive nomogram model for the diagnosis of dermatomyositis associated with malignancy.MethodsWe retrospectively analyzed 240 cases of dermatomyositis patients admitted to Sun Yat-sen Memorial Hospital, Sun Yat-sen University from January 2002 to December 2019. According to the year of admission, the first 70% of the patients were used to establish a training cohort, and the remaining 30% were assigned to the validation cohort. Univariate analysis was performed on all variables, and statistically relevant variables were further included in a multivariate logistic regression analysis to screen for independent predictors. Finally, a nomogram was constructed based on these independent predictors. Bootstrap repeated sampling calculation C-index was used to evaluate the model's calibration, and area under the curve (AUC) was used to evaluate the model discrimination ability.ResultsMultivariate logistic analysis showed that patients older than 50-year-old, dysphagia, refractory itching, and elevated creatine kinase were independent risk factors for dermatomyositis associated with malignancy, while interstitial lung disease was a protective factor. Based on this, we constructed a nomogram using the above-mentioned five factors. The C-index was 0.780 (95% CI [0.690-0.870]) in the training cohort and 0.756 (95% CI [0.618-0.893]) in the validation cohort, while the AUC value was 0.756 (95% CI [0.600-0.833]). Taken together, our nomogram showed good calibration and was effective in predicting which dermatomyositis patients were at a higher risk of developing malignant tumors.
| S-EPMC8667746 | biostudies-literature
Project description:Preeclampsia (PE) is a major cause of adverse maternal and infant outcomes. Accurate screening of PE is currently the focus of clinical attention. This study aimed to develop a model for predicting PE. A retrospective case-control study was conducted with 916 pregnant women who received care at the Second Hospital of Tianjin Medical University (October 2018 to July 2020). Women were randomly divided into the training (n=680) and testing (n=236) sets based on a ratio of 3:1. Demographic and clinical data of women were collected. In training set, logistic regression (LR), classification tree (CT) model, and random forest (RF) algorithm were used to develop prediction models for PE. Using the testing set was to validate these prediction models. The predictive performance of three models were assessed by the area under the curve (AUC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Of the total 916 women, 237 had PE. The family history of hypertension, pre-pregnancy body mass index (pBMI), blood pressure (BP) ≥130/80 mmHg in early pregnancy, age, chronic hypertension, and duration of hypertension were the predictors of PE. The AUCs for the LR, CT, and RF models were 0.778, 0.850, and 0.871, respectively (all P<0.05 for all pair-wise comparisons). The RF had the best predictive efficiency with sensitivity, specificity, PPV, and NPV of 79.6%, 94.7%, 79.6%, and 94.7%, respectively. The RF model could be a practical screening approach for predicting PE, which is helpful for clinicians to identify high-risk individuals and prevent the occurrence of adverse pregnancy outcomes.
| S-EPMC9761146 | biostudies-literature
Project description:Because existing risk prediction models for lung cancer were developed in white populations, they may not be appropriate for predicting risk among African-Americans. Therefore, a need exists to construct and validate a risk prediction model for lung cancer that is specific to African-Americans. We analyzed data from 491 African-Americans with lung cancer and 497 matched African-American controls to identify specific risks and incorporate them into a multivariable risk model for lung cancer and estimate the 5-year absolute risk of lung cancer. We performed internal and external validations of the risk model using data on additional cases and controls from the same ongoing multiracial/ethnic lung cancer case-control study from which the model-building data were obtained as well as data from two different lung cancer studies in metropolitan Detroit, respectively. We also compared our African-American model with our previously developed risk prediction model for whites. The final risk model included smoking-related variables [smoking status, pack-years smoked, age at smoking cessation (former smokers), and number of years since smoking cessation (former smokers)], self-reported physician diagnoses of chronic obstructive pulmonary disease or hay fever, and exposures to asbestos or wood dusts. Our risk prediction model for African-Americans exhibited good discrimination [75% (95% confidence interval, 0.67-0.82)] for our internal data and moderate discrimination [63% (95% confidence interval, 0.57-0.69)] for the external data group, which is an improvement over the Spitz model for white subjects. Existing lung cancer prediction models may not be appropriate for predicting risk for African-Americans because (a) they were developed using white populations, (b) level of risk is different for risk factors that African-American share with whites, and (c) unique group-specific risk factors exist for African-Americans. This study developed and validated a risk prediction model for lung cancer that is specific to African-Americans and thus more precise in predicting their risks. These findings highlight the importance of conducting further ethnic-specific analyses of disease risk.
| S-EPMC2854402 | biostudies-other
Project description:Breakthrough candidemia (BrC) is a significant problem in immunocompromised patients, particularly those with hematological disorders. To assess the characteristics of BrC in patients with hematologic disease treated with novel antifungal agents, we collected clinical and microbiological information on said patients from 2009 to 2020 in our institution. Forty cases were identified, of which 29 (72.5%) received hematopoietic stem cell transplant (HSCT)-related therapy. At BrC onset, the most administered class of antifungal agents were echinocandins, administered to 70% of patients. Candida guilliermondii complex was the most frequently isolated species (32.5%), followed by C. parapsilosis (30%). These two isolates were echinocandin-susceptible in vitro but had naturally occurring FKS gene polymorphisms that reduced echinocandin susceptibility. Frequent isolation of these echinocandin-reduced-susceptible strains in BrC may be associated with the widespread use of echinocandins. In this study, the 30-day crude mortality rate in the group receiving HSCT-related therapy was significantly higher than in the group not receiving it (55.2% versus 18.2%, P = .0297). Most patients affected by C. guilliermondii complex BrC (92.3%) received HSCT-related therapy and had a 30-day mortality rate of 53.8%; despite treatment administration, 3 of 13 patients had persistent candidemia. Based on our results, C. guilliermondii complex BrC is a potentially fatal condition in patients receiving HSCT-related therapy with echinocandin administration.
| S-EPMC10294639 | biostudies-literature
Project description:BACKGROUND:Using two large datasets from Dorset, we previously reported an internally validated multivariable risk model for predicting the risk of GI malignancy in IDA-the IDIOM score. The aim of this retrospective observational study was to validate the IDIOM model using two independent external datasets. METHODS:The external validation datasets were collected, in a secondary care setting, by different investigators from cohorts in Oxford and Sheffield derived under different circumstances, comprising 1117 and 474 patients with confirmed IDA respectively. The data were anonymised prior to analysis. The predictive performance of the original model was evaluated by estimating measures of calibration, discrimination and clinical utility using the validation datasets. RESULTS:The discrimination of the original model using the external validation data was 70% (95% CI 65, 75) for the Oxford dataset and 70% (95% CI 61, 79) for the Sheffield dataset. The analysis of mean, weak, flexible and across the risk groups' calibration showed no tendency for under or over-estimated risks in the combined validation data. Decision curve analysis demonstrated the clinical value of the IDIOM model with a net benefit that is higher than 'investigate all' and 'investigate no-one' strategies up to a threshold of 18% in the combined validation data, using a risk cut-off of around 1.2% to categorise patients into the very low risk group showed that none of the patients stratified in this risk group proved to have GI cancer on investigation in the validation datasets. CONCLUSION:This external validation exercise has shown promising results for the IDIOM model in predicting the risk of underlying GI malignancy in independent IDA datasets collected in different clinical settings.
| S-EPMC8672477 | biostudies-literature
Project description:Background: Endoscopic primary bile reflux is one of the main diagnostic criteria for bile reflux gastritis (BRG). Presently, the risk factors and prediction models of endoscopic primary bile reflux (EPBR) in gastropathy patients who cannot or will not undergo endoscopy due to contraindications are not clear. Thus, this study aimed to evaluate the risk factors of EPBR and to establish and verify a prediction model. Methods: A series of 844 patients (564 subjects with EPBR and 280 control subjects) were retrospectively selected for this study and divided into a training set (n = 591) and a validation set (n = 253) according to the usual ratio of 70:30% for the subsequent internal validation of the logistic regression model for EPBR. Fifteen parameters that might affect the occurrence of EPBR were collected. Subsequently, univariate and stepwise logistic regression analyses were introduced to reveal the risk factors and the multivariate prediction model. An R package was dedicated to the corresponding internal validation of the EPBR model. Results: The univariate analysis showed that gender, age, smoking, Helicobacter pylori (H. pylori) infection status, metabolic syndrome (MS), non-steroidal anti-inflammatory drugs (NSAIDs) use history, and previous medical histories of chronic liver diseases, cholelithiasis, and erosive gastritis were statistically significant between the two groups (P < 0.05). Multivariate regression described that being a male [OR (95%confidence interval (CI)) = 2.29 (1.50-3.50), P < 0.001], age≥45 years old [OR (95% CI) = 4.24 (2.59-6.96), P < 0.001], H. pylori infection status [OR (95% CI) = 2.34 (1.37-4.01), P = 0.002], MS [OR (95% CI) = 3.14 (1.77-5.54), P < 0.001], NSAIDs use history [OR (95% CI) = 1.87 (1.03-3.40), P = 0.04], cholelithiasis history [OR (95% CI) = 3.95 (2.18-7.18), P < 0.001] and erosive gastritis history [OR (95% CI) = 6.77 (3.73-12.29), P < 0.001] were the risk factors for the occurrence of EPBR. Based on the results of these risk factors, an EPBR prediction model with an adequate calibration and excellent discrimination was established [area under the curve (AUC): 0.839, 95% CI = 0.806-0.872]. Conclusions: Being a male, age ≥ 45 years old, H. pylori infection, histories of MS, NSAIDs use, cholelithiasis, and erosive gastritis appear to be the risk factors for EPBR, and our favorable prediction model might be an option for the prediction of EPBR.
| S-EPMC8631358 | biostudies-literature
Project description:BackgroundPulmonary infections can imitate pulmonary neoplasms. Pulmonary tuberculosis (TB) is a typical example of an infection that mimics cancer and results in unexpected exposure of healthcare workers to TB. A large number of patients with suspected lung malignancy are referred to cancer centers, although the epidemiology of the final diagnosis is unclear in Japan. This study aimed to determine the frequency and nature of pulmonary infections that imitate malignancy among patients with presumed lung cancer that is subsequently diagnosed as a pulmonary infection based on bronchoscopy findings. We also aimed to identify the prevalence of formerly undiagnosed pulmonary tuberculosis that could pose an occupational risk to healthcare workers.MethodsThis single-center retrospective cross-sectional study included patients with suspected pulmonary malignancy who underwent bronchoscopy at a tertiary care cancer center in Japan between April 2017 and March 2020. Electronic medical records of the bronchoscopy database were reviewed to identify the final diagnoses recorded by physicians.ResultsAmong the 460 patients enrolled in the present study, 362 (78.7%) and 8 (1.7%) had primary or metastatic pulmonary lesions and benign lesions, respectively. Sixty-six patients (14.3%) had nonspecific findings or other pulmonary diseases. Infection was confirmed in 24 patients (5.2%). Mycobacterial infections (n = 16) were the most frequent infectious disease; four patients had TB and 12 had nontuberculous mycobacterial infections.ConclusionsDespite the rare occurrence of TB in patients with suspected lung malignancy, healthcare workers should remain vigilant regarding the possibility of TB to prevent occupational exposure during invasive procedures such as routine bronchoscopy.
| S-EPMC7796787 | biostudies-literature
Project description:Recently, modified REACH-B (mREACH-B) risk prediction model for hepatocellular carcinoma (HCC) development was proposed. We validated the accuracy of the mREACH-B model and compared its accuracy with those of other prediction models. Between 2006 and 2012, 1,241 patients with chronic hepatitis B (CHB) were recruited. All patients underwent transient elastography at enrollment. The median age of the study population (840 males, 401 females) was 49 years. The median PAGE-B, LSM-HCC, and mREACH-B values were 10, 10, and 8, respectively. Among patients without cirrhosis (n = 940, 75.7%), the median REACH-B value was 9. During the follow-up period (median 77.4 months), 66 (5.3%) and 83 (6.7%) patients developed HCC and liver-related events (LRE), respectively. Higher liver stiffness (LS) independently predicted HCC (hazard ratio [HR] = 1.047) and LRE development (HR = 1.047) (all P < 0.05). The mREACH-B significantly predicted HCC (AUC = 0.824 at 3-year and 0.750 at 5-year) and LRE development (AUC = 0.782 at 3-year and 0.739 at 5-year) (all P < 0.001) and it performed similarly or significantly better than the PAGE-B and LSM-HCC (AUC = 0.715-0.809 at 3-year and 0.719-0.742 at 5-year for HCC; AUC = 0.704-0.777 at 3-year and 0.721-0.735 at 5-year for LRE). Among patients without cirrhosis, mREACH-B predicted HCC (AUC = 0.803 vs. 0.654-0.816 at 3-year and 0.684 vs. 0.639-0.738 at 5-year) and LRE development (AUC = 0.734 vs. 0.619-0.789 at 3-year and 0.674 vs. 0.626-0.729 at 5-year) similarly to PAGE-B, REACH-B, and LSM-HCC. mREACH-B appropriately predicted HCC and LRE development in patients with CHB and showed similar or superior accuracy to those of PAGE-B, REACH-B, and LSM-HCC.
| S-EPMC5762585 | biostudies-literature
Project description:IntroductionCandidemia results in substantial morbidity and mortality, especially if initial antifungal therapy is delayed or is inappropriate; however, candidemia is difficult to diagnose because of its nonspecific presentation.MethodsTo develop a risk score for identifying hospitalized patients with candidemia, we performed a retrospective analysis of a large database of 176 acute-care hospitals in the United States. We studied 64,019 patients with bloodstream infection (BSI) on presentation from 2000 through 2005 (derivation cohort) and 24,685 from 2006 to 2007 (validation cohort). We used recursive partitioning (RPART) to identify the best discriminators for Candida as the cause of BSI. We compared three sets of models (equal-weight, unequal-weight, vs full model with additional variables from logistic regression model) for sensitivity analysis.ResultsThe RPART identified 6 variables as the best discriminators: age < 65 years, temperature <or= 98 degrees F or severe altered mental status, cachexia, previous hospitalization within 30 days, admitted from other healthcare facility, and need for mechanical ventilation. The prevalence for patients presented with 0 through 6 risk factors in the derivation cohort was 28.7%, 38.8%, 21.8%, 8.3%, 2.1%, 0.3%, and < 0.1% respectively. The corresponding candidemia rates were 0.4% (69/18,355), 0.8% (196/24,811), 1.6% (229/13,984), 3.2% (168/5,330), 4.2% (58/1,371), 9.6% (15/157), and 27.3% (3/11) respectively (P < 0.0001). Findings were similar in the validation cohort (P < 0.0001). The equal-weight risk score model, which signed 1 point to each risk factor, yielded good discrimination in both cohorts with areas under the receiver operating curve (AUROCs) of 0.70 versus 0.71 (derivation versus validation). AUROC values were similar for the unequal-weight model, which signed different weight to each risk factor based on multivariable logistic regression coefficient, (AUROCs, 0.70-0.72). Both equal-weight and unequal-weight models were well calibrated (all Hosmer-Lemshow P > 0.10, indicating predicted and observed candidemia rates did not differ significant across the 7 risk stratus). The full model with 16 risk factors had slightly higher AUROCs (0.74 versus 0.73 for derivation versus validation); however, 7 variables were no longer significant in the recalibrated model for the validation cohort, indicating that the additional items did not materially enhance the model.ConclusionsA simple equal-weight risk score differentiated patients' risk for candidemia in a graded fashion upon hospital presentation.
| S-EPMC2784380 | biostudies-literature
Project description:PurposeTo assess the risk factors and outcomes associated with fungemia caused by the six most commonly occurring Candida species in patients with and without malignancies.MethodsAnalysis of the episodes of fungemia due to common Candida species in adults, based on an active hospital-based surveillance program (Paris area, France, 2002 to 2014).ResultsOf the 3417 patients (3666 isolates), 1164 (34.1%) had a solid tumor (45.7% digestive tract) and 586 (17.1%) a hematological malignancy (41.8% lymphoma, 33.5% acute leukemia). The hematology patients were significantly younger, more often pre-exposed to antifungals, more often infected by C. tropicalis, C. krusei, or C. kefyr, and more often treated in the first instance with an echinocandin. Compared with inpatients who were not in ICU at the time of fungemia, those in ICU were less frequently infected by C. parapsilosis (p < 0.02), had more recent surgery (p < 0.03), and died more frequently before day 8 and day 30 (p < 0.0001). An increase in crude mortality over time in ICU was observed only in oncology patients (p < 0.04). For all patients, lack of prescription of antifungals despite knowledge of positive blood culture increased the risk of death. The odds of being infected by a given Candida species compared with C. albicans were uneven regarding age, gender, type of malignancy, hospitalization in ICU, central venous catheter, HIV status, intravenous drug addiction, and previous exposure to antifungal drugs. Compared with C. albicans, C. glabrata (OR = 0.69 [0.54-0.89]) and C. parapsilosis (OR = 0.49 [0.35-0.67]) were associated with a decreased risk of death by day 8 and day 30.ConclusionThe clinical context of underlying malignancy and hospitalization in ICU may be relevant to the initial management of candidemia.
| S-EPMC5384959 | biostudies-literature