SEARCH · PubMed Health

Results for “Predictive modeling”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 199 records · Page 11Linked to original sources

PubMed1996

A predictive model to estimate the risk of serious bacterial infections in febrile infants.

UNLABELLED: Low risk criteria have been defined to identify febrile infants unlikely to have serious bacterial infection (SBI). Using these criteria approximately 40% of all febrile infants can be defined as being at low risk. Of the remaining infants (60%) only 10%-20% have an SBI. No adequate criteria exist to identify these infants. All infants aged 2 weeks-1 year, presenting during a 1-year-period with rectal temperature > or = 38.0 degrees C to the Sophia Children's Hospital were included in a prospective study. Infants with a history of prematurity, perinatal complications, known underlying disease, antibiotic treatment or vaccination during the preceding 48 h were excluded. Clinical and laboratory variables at presentation were evaluated by a multivariate logistic regression model using SBI as the dependent variable. By using likelihood ratios a predictive model was derived, providing a post test probability of SBI for every individual patient. Of the 138 infants included in the study, 33 (24%) had SBI. Logistic regression analysis defined C-reactive protein (CRP), duration of fever, standardized clinical impression score, a history of diarrhoea and focal signs of infection as independent predictors of SBI. CONCLUSION: CRP, duration of fever, the "standardized clinical impression score", a history of diarrhoea and focal signs of infection were the independent, most powerful predictors of SBI in febrile infants, identified by logistic regression analysis. Although the predictive model is not validated for direct clinical use, it illustrates the clinical potential of the used technique. This technique offers the advantage of assess the probability of SBI in every individual infant. This probability will form the best basis for well-founded decisions in the management of the individual febrile infant.

Bacterial Infections↗

PubMed2025

Proteomic signatures and predictive modeling of cadmium-associated anxiety in middle-aged and elderly populations: an environmental exposure association study.

BACKGROUND: Emerging evidence implicates environmental contaminants such as cadmium (Cd) as modifiable risk factors for anxiety. Despite growing recognition of heavy metal toxicity in neuropsychiatric disorders, the molecular mechanisms linking environmental exposure to anxiety pathogenesis remain poorly understood. METHODS: Based on the established cohort of individuals with cognitive impairment in cadmium-contaminated areas, this cross-sectional association study enrolled 50 middle-aged and elderly hospitalized patients from these regions, adhering to the STROBE guidelines. Blood concentrations of cadmium (Cd), lead (Pb), and mercury (Hg) were analyzed in relation to anxiety severity assessed via the Hamilton Anxiety Rating Scale (HAMA). Plasma proteomic profiling was performed using data-independent acquisition (DIA) quantitative technology with an LC-MS/MS platform (timsTOF Pro, Bruker Daltonics), systematically characterizing 2,531 proteins across all samples. Machine learning techniques, specifically XGBoost and LASSO, were employed to identify biomarkers that were subsequently validated through mediation analysis and animal experiments, allowing for the screening of key protein signatures. Finally, clinical variables were integrated to construct a comprehensive model, which was then thoroughly evaluated. RESULTS: Anxious individuals exhibited significantly higher blood Cd levels than controls (β = 0.50, 95% CI: 0.07-0.93, p < 0.01), with anxiety positively correlating with depression (r = 0.62, p = 0.003) and inversely with ApoE3 genotype prevalence. Proteomics identified 120 differentially expressed proteins in anxious patients, enriched in oxidative phosphorylation and neurodegenerative pathways. CCDC126 emerged as a cadmium-associated biomarker, validated in rat models exposed to Cd. Combining CCDC126, blood Cd, Pb, and hypertension, a clinical prediction model achieved robust discrimination (AUC = 0.80, validation cohort). CONCLUSIONS: This first integrative environmental-proteomic study highlights cadmium's synergistic role in anxiety pathophysiology and psychiatric comorbidity. The predictive model offers translatable potential for early risk stratification, while CCDC126 provides mechanistic insights for targeted interventions in populations exposed to environmental pollutants.

Cadmium↗

PubMed1995

Predictive Models of Lumbar Loadings When Handling Boxes.

Back problems resulting from the compression forces on the intervertebral disks during manual material-handling tasks are an important problem affecting workers in various industries. The quantification of these forces using intradiscal pressure or biomechanical modeling is complex, time consuming, and costly, and these methods cannot be readily used in the workplace to estimate loadings on the lower back. The objective of this study was to develop a predictive model that would allow the estimation of lumbar loadings for lifting and lowering boxes using easily measured anthropometric variables and variables related to the task. A dynamic and planar segmental model and a model of internal forces at L5/S1 were used to determine the compression forces on the lower back. Two predictive models, a field model and a laboratory model, were developed to estimate the compression forces when lifting or lowering 3.3 kg to 22.0 kg boxes between heights of 15 cm and 185 cm. Both models were validated by an examination of the residuals. Their predictive performance was also compared, with the laboratory model offering a slightly better prediction than the field model. Thus, these equations represent a practical tool for a better planning of handling tasks in the working environment with the purpose of reducing the back injuries of workers.

biomechanics↗

PubMed2000

A predictive model for neutropenia associated with cancer chemotherapy.

Studies of primary prophylaxis of febrile neutropenia (FN) with recombinant human granulocyte colony-stimulating factor (rHu-G-CSF, filgrastim) administered to all patients starting their initial course of chemotherapy have demonstrated clinical effectiveness and an economic advantage in a wide range of settings. A recent meta-analysis confirmed the ability of filgrastim to reduce the risk of FN and documented infection in a variety of malignancies in both adults and children. The threshold risk for FN at which a cost saving is achieved by using filgrastim is inversely related to the daily cost of the drug and duration of hospitalization. Clinical practice guidelines for the use of filgrastim were developed based on these observations. Recent studies incorporating indirect institutional costs demonstrated that a cost saving can be achieved at substantially lower FN risk thresholds than previously estimated. Despite the demonstrated efficacy of filgrastim in primary prophylaxis, its value may be further increased by appropriately selecting patients and better understanding the importance of sustaining dose intensity in specific malignancies. Clinical prediction models capable of identifying individuals at high risk for neutropenic complications yield further reductions in FN risk thresholds and treatment costs in patients receiving cancer chemotherapy. These models also may be used to evaluate the cost-effectiveness or cost-efficiency of filgrastim. A clinical prediction model recently was presented and validated incorporating both baseline clinical characteristics as well as the results of the first cycle of chemotherapy in patients with early-stage breast cancer. A cost-effectiveness ratio of $34,297/year of life saved was estimated based on dose-response assumptions derived from a previously reported adjuvant breast cancer trial studying the impact of dose reduction on disease-free survival. The cost-effectiveness of filgrastim was evident over a wide range of clinical and cost assumptions. Clinical prediction models permit the rational and cost-effective identification of patients for filgrastim support. Existing clinical practice guidelines should be reevaluated in light of new information available on both the total costs associated with FN as well as the cost-effectiveness of these agents in patients receiving chemotherapy for sensitive and potentially curable malignancies.

Antineoplastic Agents↗

PubMed1993

Predicting survival from in-hospital CPR: meta-analysis and validation of a prediction model.

OBJECTIVE: To better clarify patient factors that predict survival from in-hospital cardiopulmonary resuscitation (CPR), using two methods: 1) meta-analysis and 2) validation of a prediction model, the pre-arrest morbidity (PAM) index. DESIGN: Meta-analysis of previously published studies by standard techniques. Retrospective chart review of validation sample. SETTING: University-affiliated teaching hospital. PATIENTS/PARTICIPANTS: Meta-analytic sample of 21 previous studies from 1965-1989. The validation sample consisted of all patients surviving resuscitation from the authors' hospital during the period September 1986 to January 1991. A matched sample of patients who did not survive from the same time period was used as the comparison group. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: The strongest negative predictors of survival, by meta-analysis, were renal failure (r = 0.088, p < 0.0002), cancer (r = 0.08, p < 0.0002), and age more than 60 years (r = 0.063, p < 0.006). Sepsis (r = 0.046, p < 0.02), recent cerebrovascular accident (CVA) (r = 0.038, p < 0.04), and congestive heart failure (CHF) class III/IV (r = 0.036, p < 0.05) were weaker negative predictors. Presence of acute myocardial infarction (AMI) was a significant positive predictor of survival (r = 0.15, p < 0.0001). The PAM score was highly predictive of survival in a logistic regression model (p < 0.0003, R2 = 9.6%). No patient who survived to discharge had a PAM score higher than 8. CONCLUSION: Meta-analysis reveals that the most significant negative predictors of survival from CPR are renal failure, cancer, and age more than 60 years, while AMI is a significant positive predictor. The PAM index is a useful method of stratifying probability of survival from CPR, especially for those patients with high PAM scores, who have essentially no chance of survival.

Age Factors↗

PubMed1996

Prognostic factors and a predictive model of follicular lymphoma: a 25-year study at a single institution in Japan.

The incidence of follicular lymphoma in Japan is far lower than that in western countries, and no large-scale clinicopathologic studies on this neoplasm have been conducted in Japan. We reviewed histopathological specimens from 118 of 135 patients who had been diagnosed as having follicular lymphoma between 1968 and 1993. Prognostic factors influencing survival were analyzed using univariate and multivariate analyses. Factors that were independently significant upon multivariate analysis were incorporated into a predictive model. Ninety-three patients (78.8%) had a confirmed diagnosis of follicular lymphoma. Twenty-one of the remaining 25 patients were categorized as having other lymphoma subtypes, and four patients showed indefinite findings or those suggesting diseases other than lymphoma. Major characteristics of the 93 patients with follicular lymphoma were a median age of 53 years (20-85); 59 males (63%) and 34 females (37%); small cleaved cell type in 33 (35%), mixed cell type in 41 (44%) and large cell type in 19 (20%); stage I/II in 41 (44%) and stage III/IV in 50 (54%). Overall survival was 71% at 5 years, 58% at 10 years, and 43% at 15 years with a median survival of 13.3 years. Multivariate analysis revealed that two variables, age (>60) (P=0.001) and the serum LDH level (>1 x normal value) (P=0.026), were unfavorably significant prognostic factors influencing survival. The predictive model using these two variables identified three risk groups with estimated five-year survival rates of 88.5%, 56.8%, and 31.5%. Age and serum LDH were significant predictors of survival in Japanese patients with follicular lymphoma. Our predictive model may provide a basis for future therapeutic trials against follicular lymphoma in Japan.

Adult↗

PubMed2003

A predictive model for outcome after conservative decompression surgery for lumbar spinal stenosis.

This study was designed to develop predictive models for surgical outcome based on information available prior to lumbar stenosis surgery. Forty patients underwent decompressive laminarthrectomy. Preop and 1-year postop evaluation included Waddell's nonorganic signs, CT scan, Waddell disability index, Oswestry low back pain disability questionnaire, low back outcome score (LBOS), visual analog scale (VAS) for pain intensity, and trunk strength testing. Statistical comparisons of data used adjusted error rates within families of predictors. Mathematical models were developed to predict outcome success using stepwise logistic regression and decision-tree methodologies (chi-squared automatic interaction detection, or CHAID). Successful outcome was defined as improvement in at least three of four criteria: VAS, LBOS, and reductions in claudication and leg pain. Exact logistic regression analysis resulted in a three-predictor model. This model was more accurate in predicting unsuccessful outcome (negative predictive value 75.0%) than in successful outcome (positive predictive value 69.6%). A CHAID model correctly classified 90.1% of successful outcomes (positive predictive value 85.7%, negative predictive value 100%). The use of conservative surgical decompression for lumbar stenosis can be recommended, as it demonstrated a success rate similar to that of more invasive techniques. Given its physiologic and biomechanical advantages, it can be recommended as the surgical method of choice in this indication. Underlying subclinical vascular factors may be involved in the complaints of spinal stenosis patients. Those factors should be investigated more thoroughly, as they may account for some of the failures of surgical relief. The CHAID decision tree appears to be a novel and useful tool for predicting the results of spinal stenosis surgery

Adolescent↗

PubMed1993

Development of predictive models of laboratory animal growth using artificial neural networks.

Traditional regression analysis of body weight growth curves encounters problems when the data are extremely variable. While transformations are often employed to meet the criteria of the analysis, some transformations are inadequate for normalizing the data. Regression analysis also requires presuppositions regarding the model to be fit and the techniques to be used in the analysis. An alternative approach using artificial neural networks is presented which may be suitable for developing predictive models of growth. Neural networks are simulators of the processes that occur in the biological brain during the learning process. They are trained on the data, developing the necessary algorithms within their internal architecture, and produce a predictive model based on the learned facts. A dataset of Sprague-Dawley rat (Rattus norvegicus) weights is analyzed by both traditional regression analysis and neural network training. Predictions of body weight are made from both models. While both methods produce models that adequately predict the body weights, the neural network model is superior in that it combines accuracy and precision, being less influenced by longitudinal variability in the data. Thus, the neural network provides another tool for researchers to analyze growth curve data.

Algorithms↗

PubMed2002

Comparison of prediction models for adverse outcome in pediatric meningococcal disease using artificial neural network and logistic regression analyses.

The objective of this study was to compare artificial neural network (ANN) and multivariable logistic regression analyses for prediction modeling of adverse outcome in pediatric meningococcal disease. We analyzed a previously constructed database of children younger than 20 years of age with meningococcal disease at four pediatric referral hospitals from 1985-1996. Patients were randomly divided into derivation and validation datasets. Adverse outcome was defined as death or limb amputation. ANN and multivariable logistic regression models were developed using the derivation set, and were tested on the validation set. Eight variables associated with adverse outcome in previous studies of meningococcal disease were considered in both the ANN and logistic regression analyses. Accuracies of these models were then compared. There were 381 patients with meningococcal disease in the database, of whom 50 had adverse outcomes. When applied to the validation data set, the sensitivities for both the ANN and logistic regressions models were 75% and the specificities were both 91%. There were no significant differences in any of the performance parameters between the two models. ANN analysis is an effective tool for developing prediction models for adverse outcome of meningococcal disease in children, and has similar accuracy as logistic regression modeling. With larger, more complete databases, and with advanced ANN algorithms, this technology may become increasingly useful for real-time prediction of patient outcome.

Adolescent↗

PubMed2001

Z score prediction model for assessment of bone mineral content in pediatric diseases.

The objective of this study was to develop an anthropometry-based prediction model for the assessment of bone mineral content (BMC) in children. Dual-energy X-ray absorptiometry (DXA) was used to measure whole-body BMC in a heterogeneous cohort of 982 healthy children, aged 5-18 years, from three ethnic groups (407 European- American [EA], 285 black, and 290 Mexican-American [MA]). The best model was based on log transformations of BMC and height, adjusted for age, gender, and ethnicity. The mean +/- SD for the measured/predicted in ratio was 1.000 +/- 0.017 for the calibration population. The model was verified in a second independent group of 588 healthy children (measured/predicted In ratio = 1.000 +/- 0.018). For clinical use, the ratio values were converted to a standardized Z score scale. The whole-body BMC status of 106 children with various diseases (42 cystic fibrosis [CF], 29 juvenile dermatomyositis [JDM], 15 liver disease [LD], 6 Rett syndrome [RS], and 14 human immunodeficiency virus [HIV]) was evaluated. Thirty-nine patients had Z scores less than -1.5, which suggest low bone mineral mass. Furthermore, 22 of these patients had severe abnormalities as indicated by Z scores less than -2.5. These preliminary findings indicate that the prediction model should prove useful in determining potential bone mineral deficits in individual pediatric patients.

Absorptiometry, Photon↗

PubMed1998

Delirium: comparison of four predictive models in hospitalized critically ill elderly patients.

Delirium, a cognitive and behavioral disorder affecting more than one third of all hospitalized elderly patients, is often misdiagnosed or unrecognized by caregivers, leading to higher patient morbidity and mortality rates. Prediction of the disorder, based on known predisposing and precipitating risk factors, can be used to target susceptible patients for prevention and early intervention. Predictive models need to be evaluated for clinical application and predictive value. Therefore, in this study, four predictive models were applied on a case-by-case basis to an elderly sample of 10 delirious patients and 10 nondelirious patients to determine sensitivity, specificity, and predictive values in a critical care setting. Results indicated six individual significant variables in these models: age, infection, dementia, blood urea nitrogen-to-creatinine ratio, severe illness, and comorbidity. A final multivariate model, derived from all variables, exhibited a sensitivity of 100% and specificity of 90% in predicting delirium in this study. Further studies are needed to substantiate these results. Then, identified risk factors can be incorporated into delirium prevention protocols for use by nurses at the bedside.

Aged↗

PubMed2003

Cost savings for a preferred provider organization population with multi-condition disease management: evaluating program impact using predictive modeling with a control group.

Disease management programs have historically had difficulty demonstrating the financial value of their programs using statistically rigorous methods. This preliminary study utilized a predictive model built from a control group that consisted of individual employer groups whose benefits managers did not purchase disease management. The purpose of this evaluation was to examine the cost savings associated with Health Management Corporation's disease management program for asthma, diabetes, and coronary artery disease. Study and control group members were identified with exactly the same selection criteria as self-insured employer groups (n = 76,194) with preferred provider organization health plans that offered the disease management program. The study group consisted of members (n = 1,009) who were identified as having any of the three conditions and were eligible for the disease management program, regardless of the extent of their participation. The control group included members (n = 2,491) who were identified as having any of the same three conditions. The control group data were used to develop a predictive model designed to calculate expected medical claims costs for the study group. The gross savings of the disease management program was measured by calculating the difference between the expected medical claims costs predicted by the model and the actual medical claims costs for the study group. Preliminary analyses indicate that Health Management Corporation's disease management program produced a return on investment of $2.84: $1.00 and $1.45 gross savings per member per month.

Adolescent↗

PubMed2001

Organ system failures prediction model in intensive care patients with acute renal failure treated with dialysis.

OBJECTIVE: To evaluate the organ system failures hospital mortality predictions in critically ill patients with acute renal failure requiring dialysis. DESIGN: Prospective, cohort study. SETTING: Intensive care units in a tertiary care university hospital in Taiwan. PATIENTS: A total of 112 patients admitted to the intensive care units with acute renal failure who required dialysis from January 1999 through December 1999. INTERVENTIONS: Collection of information necessary to compute the number of failed organs. MEASUREMENTS AND RESULTS: Of the 112 patients studied, 75 were men and 37 were women. The mean age of survivors and non-survivors was 58.59 +/- 19.91 years and 58.76 +/- 19.62 years. The overall mortality rate was 67%. There were no significant differences between survivors and non-survivors in terms of age, gender, or indication for dialysis. The cause of death in the majority of patients was related to organ system failure during the 24 hours immediately preceding the initiation of acute hemodialysis, and carry mortality rates exceeding 83% with the coexistence of four or more failed organs. The area under the organ system failures prediction model receiver operating characteristic curve equaled 0.772 +/- 0.046. CONCLUSION: We conclude that mortality rate for acute renal failure in intensive care unit patients continues to be high. Organ system failures prediction model performed well and simple in its ability to identify patients who die in hospital. Mortality rate increases as number of failed organ increases.

Acute Kidney Injury↗

PubMed2003

Outcome prediction model for severe diffuse brain injuries: development and evaluation.

BACKGROUND: Intensive care resources for the management of severe diffuse brain injury patients (SDBI) are limited. Their optimal use is possible only if we can predict at admission which patients are unlikely to improve. AIMS: To develop a simple and effective model to predict poor outcome in patients with SDBI in order to help guide initial therapy. MATERIAL AND METHODS: The prognostic factors and outcomes of 289 patients with severe diffuse brain injury (GCS 3-8) were analyzed retrospectively. The prognostic factors analyzed were age, mode of injury, GCS at admission, pupillary reaction, horizontal oculocephalic reflex, and CT scan findings. Outcome at 1 month was classified as unfavorable--death or persistent vegetative state, or favorable--improvement with or without some disability. A stepwise linear logistic regression analysis was used to identify the most important predictors of poor outcome. A prediction model (NIMHANS model-NM) was developed using these factors. NM and several currently available outcome prediction models were prospectively applied in a separate group of 26 patients with severe diffuse brain injury managed with a different protocol. RESULTS: The most important predictors of poor outcome were found to be the horizontal oculocephalic reflex, motor score of GCS, and midline shift on CT scan. NM was found to be more sensitive (75%) and specific (67%) than most other models in predicting unfavorable outcome. NM had high false pessimistic results (33%). CONCLUSION: Prediction models cannot be used to guide initial therapy.

Adult↗

PubMed1997

[Value of a predictive model of ambulatory blood pressure integrating physical activity].

OBJECTIVE: To determine how much of the variations of blood pressure during a 24 hour period could be accounted for by a change in activity and establish a predictive model. MATERIALS AND METHODS: Twenty three healthy subjects (mean age 25 +/- 2 years) were studied. The BP, heart rate (HR), and time of measure (T) were recorded by ambulatory BP monitoring using Spacelabs (4 measures per hour). At each measure the subject noted in a diary the degree of activity on a six level semi-quantitative scale. DATA ANALYSIS: A model was constructed using an analysis of covariance. Different parameters were added in succession to reach a model of the type P: P0 + A + beta + (HR-HR0) + H, were P = predicted systolic pressure, P0 = mean systolic BP over the 24 hours. A variation in systolic BP for activity level, beta = the slope of the regression between systolic BP and HR during activity A, and HR0 the mean HR during this activity. RESULTS: 1) In order to test the model, the values measured in one subject were compared to the predicted values from the model in 22 others. The procedure was then repeated for the other subjects. This common model predicted 41 +/- 21% of fluctuations in BP of the subject analysed with a range of 0 to 66%. 2) In order to refine the individual model two subjects were explored 7 times over 24 h of non consecutive days. The measures of the last recording were compared to the predicted values from the application of the model to the six preceding recordings. The model then predicted 81% and 66% of the BP values of the test day. The mean of the 24 hour individual difference over a one hour period between the measures and its predicted value by the model was 0.13 +/- 4.8 mmHg, and -0.75 +/- 7.7 mmHg. CONCLUSION: This study expresses in a quantitative fashion the importance of the level of activity in the evaluation of the level of ambulatory BP. The introduction of this method of quantification and analysis seems logical in therapeutic trial. The difference in the predictions by the model for some subjects poses the problem of uniform coding of activities and that of the recognition of other events such as stress and dreaming in sleep.

Adult↗

PubMed2001

Use of real time leukaemia data to validate model predictions based on analyses and computer simulations.

Predictions arising out of a mathematical model that describes the expansion of leukaemia from a diffusion-orientated perspective are critiqued and validated by employing available real time data. Based on agreements found between model predictions and the data, but mindful of the limitations it presents, it is concluded that the model could be used to describe the dynamics of normal and abnormal cells in leukaemia. It is suggested that further studies of the behaviour of certain normal cell types in contrast to abnormal cells during leukaemic development could engender additional insights into leukaemia and its treatment.

Cell Death↗

PubMed1981

Lactate after exercise in man: IV. Physiological observations and model predictions.

Following earlier papers that established the mathematical form of the time dependence of lactate concentrations during recovery from several types of exercise, and that set up a two-compartment model predicting the same time dependences, the present work applies the model to obtain parameters of specific physiological processes. Satisfactory agreement between predictions of the model and our experiment and literature data is obtained in the cases were comparisons can be made, as in the muscular lactate time evolution measured from biopsy samples, in blood flows through the active muscle at the end of exercise or at rest and their evolution during recovery, as well as in the volume of the active muscle compartment. The model prediction that lactate efflux from the muscles to the blood can reduce to zero during recovery is verified experimentally.

Arteries↗

PubMed1999

Within-channel cues in comodulation masking release (CMR): experiments and model predictions using a modulation-filterbank model.

Experiments and model calculations were performed to study the influence of within-channel cues versus across-channel cues in comodulation masking release (CMR). A class of CMR experiments is considered that are characterized by a single (unmodulated or modulated) bandpass noise masker with variable bandwidth centered at the signal frequency. A modulation-filterbank model suggested by Dau et al. [J. Acoust. Soc. Am. 102, 2892-2905 (1997)] was employed to quantitatively predict the experimental data. Effects of varying masker bandwidth, center frequency, modulator bandwidth, modulator type, and signal duration on CMR were examined. In addition, the effect of band limiting the noise before or after modulation was shown to influence the CMR in the same way as a systematic variation of the modulation depth. It is demonstrated that a single-channel analysis, which analyzes only the information from one peripheral channel, quantitatively accounts for the CMR in most cases, indicating that an across-channel process is generally not necessary for simulating results from this class of CMR experiments. True across-channel processes may be found in another class of CMR experiments.

Adult↗