SEARCH · Search PubMed

Results for “External validity”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

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

At least 901 records · Page 50Linked to original sources

PubMed2000

Application of in vitro-in vivo correlations (IVIVC) in setting formulation release specifications.

An in vitro-in vivo correlation (IVIVC) was established for an osmotic controlled-release dosage form by deconvolution using data from an immediate-release treatment as the characteristic response. The established IVIVC was evaluated internally (predicting data used to develop the IVIVC) and externally (predicting data not originally included in developing the IVIVC), and the application of the IVIVC in product development was demonstrated. The estimated in vivo dissolution profile compared favorably with the in vitro drug release profile. Good agreement was demonstrated between the estimated and observed cumulative drug release across the entire time course (level A correlation) with low (<10%) predictive error for both C(max) and AUC(infinity). External validation using lots not initially included in the model development compared very well with the observed in vivo profile, with mean prediction errors less than 10% for both C(max) and AUC(infinity). The proposed method demonstrates a schema for developing IVIVCs using data from biostudies routinely conducted during formulation development. The method should facilitate product optimization and can support setting in vitro dissolution specification.

Algorithms↗

PubMed2005

Monitoring complex media fermentations with near-infrared spectroscopy: comparison of different variable selection methods.

Near-infrared spectroscopy (NIRS) is known to be a suitable technique for rapid fermentation monitoring. Industrial fermentation media are complex, both chemically (ill-defined composition) and physically (multiphase sample matrix), which poses an additional challenge to the development of robust NIRS calibration models. We investigated the use of NIRS for at-line monitoring of the concentration of clavulanic acid during an industrial fermentation. An industrial strain of Streptomyces clavuligerus was cultivated at 200-L scale for the production of clavulanic acid. Partial least squares (PLS) regression was used to develop calibration models between spectral and analytical data. In this work, two different variable selection methods, genetic algorithms (GA) and PLS-bootstrap, were studied and compared with models built using all the spectral variables. Calibration models for clavulanic acid concentration performed well both on internal and external validation. The two variable selection methods improved the predictive ability of the models up to 20%, relative to the calibration model built using the whole spectra.

Algorithms↗

PubMed2005

Predictors of contact difficulty and refusal in a longitudinal study.

BACKGROUND: Attrition presents a serious problem to researchers collecting longitudinal data. Participant loss threatens both the internal and external validity of research findings. This study sought to examine predictors related to contact difficulty and refusals in a longitudinal study. METHOD: Data for this paper came from the Developmental Trends Study, a longitudinal study investigating the development of disruptive behaviour disorders in a sample of 177 clinic-referred boys. Annual follow-up assessments were conducted, ages for the periods examined ranged from 11 to 19 years. The predictor domains during project years 1-4 included demographics, child functioning, parental functioning, parenting skills and participant dispersion. RESULTS: These indicated that participant's older age, low socioeconomic status, the presence of callous and un-emotional behaviours, and having a father with antisocial personality disorder predicted contact difficulty, while participant's older age, living in a rural environment and attention deficit hyperactivity disorder were predictive of refusals. Participant dispersion was not significantly related to either contact difficulty or refusal status. CONCLUSIONS: Investigating a broader range of variables may better allow researchers to identify participants who may be at risk of attrition. Early identification of 'difficult' participants enables development of retention strategies to minimize attrition.

Adolescent↗

PubMed2003

Predictive modeling for the presence of prostate carcinoma using clinical, laboratory, and ultrasound parameters in patients with prostate specific antigen levels < or = 10 ng/mL.

BACKGROUND: The objective of the current study was to develop a model for predicting the presence of prostate carcinoma using clinical, laboratory, and transrectal ultrasound (TRUS) data. METHODS: Data were collected on 1237 referred men with serum prostate specific antigen (PSA) levels < or = 10 ng/mL who underwent an initial prostate biopsy. Variables analyzed included age, race, family history, referral indication(s), prior vasectomy, digital rectal examination (DRE), PSA level, PSA density (PSAD), and TRUS findings. Twenty percent of the data were reserved randomly for study validation. Logistic regression analysis was performed to estimate the relative risk, 95% confidence interval, and P values. RESULTS: Independent predictors of a positive biopsy result included elevated PSAD, abnormal DRE, hypoechoic TRUS finding, and age 75 years or older. Based on these variables, a predictive nomogram was developed. The sensitivity and specificity of the model were 92% and 24%, respectively, in the validation study for which the predictive probability > or = 10% was used to indicate the presence of prostate carcinoma. The area under the receiver operating characteristic curve (AUC) for the model was 73%, which was significantly higher compared with the prediction based on PSA alone (AUC, 62%). If it was validated externally, then application of this model to the biopsy decision could result in a 24% reduction in unnecessary biopsy procedures, with an overall reduction of 20%. CONCLUSIONS: Incorporation of clinical, laboratory, and TRUS data into a prebiopsy nomogram significantly improved the prediction of prostate carcinoma over the use of individual factors alone. Predictive nomograms may serve as an aid to patient counseling regarding prostate biopsy outcome and to reduce the number of unnecessary biopsy procedures.

Adenocarcinoma↗

PubMed2004

Disseminating effective cancer screening interventions.

A large gap exists between the results of research concerning efficacious cancer screening programs and the programs delivered in practice. In this article, the authors discuss issues in, barriers to, and lessons learned regarding the dissemination of interventions. They summarize previous reviews, exemplary studies, and theories regarding the diffusion and dissemination of cancer screening interventions. Six lessons learned address the involvement of key stakeholders, factors influencing diffusion, the need for different types of efficacy and effectiveness studies with greater attention to external validity, replication, the use of theoretical and evaluation models, and the importance of policy infrastructure. In this article, the authors make recommendations for future research and practice, including improving the understanding of the intervention process and changing the types of grants funded and review criteria used. Also needed are an enhanced infrastructure, including policies to support dissemination, and the involvement of researchers, health care administrators, clinicians, and funding organizations in dissemination if the gap between research and practice in cancer screening is to be reduced.

Delivery of Health Care↗

PubMed2026

Age-Associated Four-Gene Prognostic Signature in Breast Cancer.

BACKGROUND: Young-onset breast cancer is associated with inferior disease-free survival (DFS), but the contribution of additional molecular heterogeneity remains unclear. AIMS: To identify an exploratory age-associated gene expression signature linked to recurrence-related outcomes and evaluate its prognostic association. METHODS AND RESULTS: We analyzed clinicopathological and RNA-sequencing data from 821 patients with Stages I-III invasive ductal or lobular carcinoma in The Cancer Genome Atlas, including 142 patients aged ≤ 45 years. Genes associated with both age and DFS were screened, followed by LASSO-Cox and stepwise multivariable Cox regression. A four-gene signature (Sig4: C4orf14 [NOA1], LINC01124, ZNF704, and AGFG2) was identified. Young patients had significantly worse DFS than older patients, whereas overall and disease-specific survival did not differ significantly. After adjustment for clinicopathological factors, young age remained associated with worse DFS. Following inclusion of the continuous Sig4 score, the age association was attenuated and no longer statistically significant, while Sig4 remained independently associated with worse DFS. Sig4-high tumors were enriched for proliferation, cell-cycle, DNA-repair, metabolic, and stress-response pathways. In METABRIC, the fixed TCGA-derived Sig4 score was associated with worse relapse-free survival in the overall cohort but not in patients aged ≤ 45 years. CONCLUSION: Sig4 is an exploratory age-associated four-gene signature with potential general prognostic relevance in breast cancer. Its utility for risk stratification specifically in young-onset breast cancer was not externally validated and requires confirmation in independent prospective cohorts enriched for young patients.

Humans↗

PubMed2006

The effect of managed care on use of health care services: results from two contemporaneous household surveys.

This paper estimates treatment effects of managed care plans on the utilization of health care services using data from two contemporaneous, nationally representative household surveys from the USA. The paper exploits recent advances in simulation-based econometrics to take the endogeneity of enrollment into managed care plans into account and identify the causal relationship between managed care enrollment and utilization. Overall, results from the two surveys are remarkably similar, lending credibility to their external validity and to the econometric model and estimation methods. There is significant evidence of self-selection into managed care plans. After accounting for selection, an individual enrolled in an health maintenance organization (HMO) plan has 2 more visits to a doctor and has 0.1 more visits to the emergency room per year than would the same individual enrolled in a nonmanaged care plan.

Adolescent↗

PubMed2002

Revealed preference valuation compared to contingent valuation: radon-induced lung cancer prevention.

This paper explores and compares two tools of economic valuation, revealed preference and contingent valuation, with the purpose of ultimately informing the use of two methods of economic evaluation, CEA and CBA. The valuation methods are applied to empirical data for radon-induced lung cancer prevention. However, only the single bound CV and the subjective revealed preference estimates have overlapping confidence intervals, indicating that they do have external validity as assessed by convergent validity. The revealed preference subjective risk valuation was (180 pounds sterling (144 pounds sterling, 247 pounds sterling)) and the single bounded contingent valuation estimate was (269 pounds sterling (201 pounds sterling, 343 pounds sterling)).

Air Pollutants, Radioactive↗

PubMed2003

Impact of comorbidity on the outcome of laryngeal squamous cancer.

BACKGROUND: Comorbidity has been shown to be a determinant in treatment selection and survival in various cancers. We have previously shown that the Adult Comorbidity Evaluation-27 index is applicable in a United Kingdom setting, and the process of comorbidity grading by retrospective notes evaluation is an accurate and reliable process. METHODS: The impact of comorbidity and other factors on survival was examined retrospectively in a cohort of 180 patients with laryngeal squamous cell cancer. RESULTS: This study shows for the first time that the higher comorbidity burden seen in supraglottic cancers in contrast to glottic tumors might account for the poorer prognosis. This study also externally validates the composite comorbidity-TNM staging system described by Piccirillo and shows the composite system to be a better predictor of outcome than TNM staging system alone. CONCLUSIONS: Moderate and severe comorbidity have a greater and statistically significant impact on survival than the TNM staging system.

Adult↗

PubMed2026

Refining the Multivariable Predictive-Prognostic PREDICTR-OPC Model for Survival in Surgical Escalation for Oropharyngeal Squamous Cell Carcinoma.

OBJECTIVES: The PREDICTR-OPC model is the only prognostic classifier for oropharyngeal squamous cell carcinoma (OPSCC) also predictive of surgical outcomes. Of the four biomarkers included, survivin contributes minimally and presents practical limitations. This study aimed to refine and simplify the model by removing survivin, then re-assess its prognostic predictive performance compared to the original. METHODS: This retrospective cohort study analyzed a multi-center training cohort (n = 600) and an external validation cohort (n = 385) of OPSCC patients. Tumor biopsies were stained for p16, high-risk human papillomavirus (HR-HPV) DNA, tumor-infiltrating lymphocytes (TILs), and survivin and independently scored by at least three certified pathologists. Cox proportional hazards models assessed overall survival (OS), comparing three-biomarker (p16, HR-HPV, TILs) and four-biomarker models. Hazard ratios (HRs) for OS were estimated in the validation cohort, adjusting for covariates. Discrimination, calibration, and decision curve analysis (DCA) evaluated performance and clinical utility. RESULTS: Among 985 patients (median age: 57 years), median OS = 8.8 years (95% CI: 6.9-10.5). The three-biomarker model yielded HR = 4.10 (95% CI: 2.41-6.98, p < 0.001) for high- vs. low-risk groups in the validation cohort, comparable to the four-biomarker model (HR = 4.24, p < 0.001). Surgery was associated with improved OS in high-risk (HR = 0.45, p = 0.001) but not low-risk (HR = 0.83, p = 0.72) patients, consistent with the original model. The models performed similarly across all metrics (e.g., Concordance Index: 0.71 vs. 0.72; Brier Score: 0.22 for both) as was model fit (Likelihood Ratio Test: p = 0.066). DCA revealed comparable clinical benefit. CONCLUSION: Removing survivin preserves PREDICTR-OPC's predictive performance, offering a more cost-effective, easier-to-implement tool for OPSCC treatment recommendations.

Humans↗

PubMed2002

Cognitive impairment in depressed outpatients as measured with the Dementia Checklist: a simple method for primary care and in field research.

The Dementia Checklist is a 12-item dementia rating scale for physicians who, for whatever reason, cannot be specifically trained. It addresses symptoms of cognitive decline that can easily be identified, and that are typical for different stages of cognitive impairment. This allows an easy classification of the severity of dementia. In a first study, the dementia checklist was used in 937 geriatric outpatients who were treated by neuropsychiatrists for depression. All items contribute to the accuracy of measurement (Cronbach's alpha = 0.84). Differences in cognitive impairment depending on age (chi 2 = 51.7; p < or = 0.001) and depression (chi 2 = 47.6; p < or = 0.001) indicate external validity of the dementia checklist and 5.7% of the outpatients were rated as demented. The Dementia Checklist provides a very economical and easy-to-use assessment of cognitive decline.

Age Factors↗

PubMed1986

The case study: a viable approach to clinical research.

The case study, although associated with several misconceptions which compromise its effective utilization, is especially appropriate for investigation of clinical nursing problems. A literature review of the application of this research approach in other practice-based disciplines was completed. Results indicate that the case study permits focusing upon the individual in an intensive, longitudinal perspective, and facilitates bridging of the research-practice gap. Internal and external validity, considered especially problematic in case studies, can be maximized by a variety of techniques.

Humans↗

PubMed1990

Quality of reporting in diabetes patient education research: 1954-1986.

The quality of reported research representing a number of disciplines involved in diabetes patient education was investigated in this study. A rigorous literature search identified 47 studies reported between 1954 and 1986 that met inclusion criteria; 29 were published studies, 18 were unpublished. Quality of the research was measured by Duffy's Research Appraisal Checklist (RAC) and also by coding each study as to the number of threats to internal and external validity present. Overall quality as measured by the RAC ranged from 34 to 95 on a 100-point scale. A statistically significant relationship was found between publication date of the research report and quality, indicating improvement in quality during recent years. Quality of published versus unpublished research reports was not found to differ significantly. Recommendations for improving methodological rigor of future studies include: (a) the use of more rigorous designs, particularly those involving control groups such as randomized clinical trials; (b) reporting of more complete data in research reports; and (c) monitoring of the quality of future studies.

Clinical Nursing Research↗

PubMed1994

Gender and ethnic stereotyping and narcotic analgesic administration.

The purpose of this study was to investigate whether nurses provide different amounts of narcotic analgesics to male and female patients, and different amounts to white and ethnic minority patients. A retrospective survey was conducted with the medical records of 101 male and 79 female uncomplicated adult appendectomy patients, 40 of whom were ethnic minority members. Narcotic analgesic doses for the entire postoperative period were converted to equianalgesic doses comparable to intramuscular morphine. Male patients received significantly larger initial doses than female patients. There was no gender difference in the total dose received postoperatively. White patients received significantly more total postoperative narcotic analgesics than ethnic minority patients. The gender difference provides modest external validation for prior experimental results. The ethnic difference suggests that irrelevant cues may be used in nurses' medication decisions.

Adolescent↗

PubMed2006

Quantitative proteomic profiling of pancreatic cancer juice.

Pancreatic juice is an exceptionally rich source of cancer-specific proteins shed from cancerous ductal cells into the pancreatic juice. Quantitative proteomic analysis of the proteins specific to pancreatic cancer juice has not previously been reported. We used isotope-code affinity tag (ICAT) technology and MS/MS to perform quantitative protein profiling of pancreatic juice from pancreatic cancer patients and normal controls. ICAT technology coupled with MS/MS allows the systematic study of the proteome and measures the protein abundance in pancreatic juice with the potential for development of biomarkers. A total of 105 proteins were identified and quantified in the pancreatic juice from a pancreatic cancer patient, of which 30 proteins showed abundance changes of at least twofold in pancreatic cancer juice compared to normal controls. Many of these proteins have been externally validated. This is the first comprehensive study of the pancreatic juice proteome by quantitative global protein profiling, and the study reveals numerous proteins that are shown for the first time to be associated with pancreatic cancer, providing candidates for diagnostic biomarkers. One of the identified proteins, insulin-like growth factor binding protein-2 was further validated by Western blotting to be elevated in pancreatic cancer juice and overexpressed in pancreatic cancer tissue.

Amino Acid Sequence↗

PubMed2006

Near-infrared spectroscopic monitoring of biomass, glucose, ethanol and protein content in a high cell density baker's yeast fed-batch bioprocess.

The use of at-line NIRS to monitor a high cell density fed-batch baker's yeast bioprocess was investigated. Quantification of the key analytes (biomass, ethanol and glucose) and the product quality indicator (percentage protein content) was studied. Biomass was quantitatively modelled using whole matrix samples (as was percentage protein content). The dominance of the whole matrix spectrum by biomass, and its associated light scattering effects, were overcome by use of filtrate samples and adapted (semi-synthetic) filtrate samples, which allowed successful ethanol and glucose modelling, respectively. Calibrations were rigorously challenged via external validation with large sample sets relative to the calibration sample size, ensuring model robustness and potential practical utility. The standard errors of calibration for biomass, glucose, ethanol and total intracellular protein were (g/l) 1.79, 0.19, 0.79 and 0.91, respectively, comparable to those of the primary assays. The calibration strategies necessary to generate quantitative models for this range of analytes in such a complex high cell density bioprocess fluid are discussed.

Biomass↗

PubMed1999

Pitfalls in the categorization of behaviour: a comparison of dolphin whistle classification methods.

The categorization of behaviour patterns into separate classes is crucial to the study of animal behaviour. Traditionally researchers have classified behaviour patterns through careful observation by eye. Recently this method has been increasingly replaced by computer methods. While the definition and fine scale analysis that can be achieved with computers is desirable, only a few studies have actually looked at how these methods perform in comparison with human observation. I compared the classification of bottlenose dolphin, Tursiops truncatus, whistles by human observers with the performance of three computer methods: (1) a method developed by McCowan (1995, Ethology, 100, 177-193); (2) a comparison of cross-correlation coefficients using hierarchical cluster analysis; and (3) a comparison of average difference in frequency along two whistle contours also using hierarchical cluster analysis. The whistle sample consisted of 104 randomly chosen whistles from a group of four captive bottlenose dolphins recorded both during periods when one was separate from the rest of the group and while they all swam in the same pool. The sample contained five individual-specific signature whistles and several nonsignature whistles. Five human observers, without knowledge of the recording context, were more likely than the computer methods to identify signature whistles that were used only while an animal was isolated from the rest of the group. I discuss the limitations of methods commonly used for pattern recognition in communication studies. The discrepancies between methods show how crucial it is to obtain an external validation of the behaviour classes used in studies of animal behaviour. Copyright 1999 The Association for the Study of Animal Behaviour.

Journal Article↗

PubMedDate not supplied

A comparison of methods of recruitment to a health promotion program for university seniors.

BACKGROUND: Participant recruitment is an understudied part of health research. It can be a limiting factor and can affect the external validity of a study. This study compares two recruitment methods, active and passive, used to recruit university seniors into a health promotion study. METHODS: During active recruitment, 3,787 seniors were telephoned and asked to participate. During passive recruitment, 5,644 seniors were mailed literature and asked to respond if interested. RESULTS: During active recruitment, 1,680 h were spent on the phone, 341 participants were measured, and 161 entered the study, at a cost of $79 per participant. During passive recruitment, 970 h were spent on the phone, 238 participants were measured, and 178 entered the study, at a cost of $45 per participant. The active method had a higher recruitment rate (9% vs 4%) and attrition rate (53% vs 25%) than the passive method. CONCLUSIONS: In conclusion, neither recruitment method was ideal. Those recruited with the passive method reported more physical activity and had lower blood pressures, suggesting a self-selection bias. Active recruitment produced a more representative sample and a higher recruitment rate. Passive recruitment was less expensive and led to a lower attrition rate. Conclusions are limited because these are uncontrolled comparisons, not a controlled study.

Adolescent↗