Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Multivariate Analysis”

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.

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

At least 91 records · Page 5Linked to original sources
PubMed1984

Detection of haemophilia carriers: multivariate analysis compared with discriminant analysis using up to five factor VIII variates.

Multivariate analysis is potentially superior to the linear discriminant analysis which is commonly used to identify carriers of haemophilia. Our aim was to compare these two statistical methods, and to find which factor VIII variates most effectively partitioned carrier and normal subjects. In this study we assayed one- and two-stage factor VIII coagulant activity, factor VIII related antigen by electroimmunoassay and by fluoroimmunoassay, and ristocetin co-factor in 50 normal females and 50 carriers of haemophilia. From the results we calculated multivariate ellipses which circumscribed the normal and the carrier populations, and we displayed these on the monitor of a microcomputer. These ellipses separated the two populations better than linear discriminants calculated on the same data. Multivariate analysis correctly identified 94% of the carriers whereas discriminant analysis correctly identified only 84%. Discriminant analysis gave poorer results because the statistical assumption of equal variance was breached, whereas the assumption of multivariate normality was upheld. Of the five factor VIII variates, two-stage factor VIII coagulant activity and factor VIII related antigen by electroimmunoassay correctly identified the most subjects. Ristocetin co-factor did not improve the diagnostic ability, either when in lieu of or when added to factor VIII related antigen.

Antigens↗
PubMed1995

[Non-significant in univariate but significant in multivariate analysis: a discussion with examples].

Perhaps as a result of higher research standard and advancement in computer technology, the amount and level of statistical analysis required by medical journals become more and more demanding. It is now realized by researchers that univariate analysis alone may not be sufficient, especially for complex data sets. Additional, and sometimes even contradictory, results may be found using multivariate analysis. During the course of data analysis, a common practice is to include in multivariate analysis only those variables that are statistically significant in univariate analysis. Such a habit is risky as some variables not significant in univariate analysis may become significant in multivariate analysis. In this study, we identify, with examples, four possible scenarios in which the above situation could occur: (1) the effect of unbalanced sample size; (2) the influence of missing data; (3) an extremely large within group variation, relative to between group variation; and (4) the presence of interaction. In addition to detailed analysis steps, raw data sets are also available for readers to verify all the results presented. Although we only used the log-rank test and Cox regression for illustration purposes, the underlying concepts can be applied to other multivariate procedures such as the logistic regression and multiple linear regression.

Analysis of Variance↗
PubMed2001

Multivariate analysis of laryngeal fluorescence spectra recorded in vivo.

BACKGROUND AND OBJECTIVE: The potential of using various multivariate analysis methods for classification of fluorescence spectra acquired in vivo from laryngeal tissues in Patients was investigated. STUDY DESIGN/MATERIALS AND METHODS: Autofluorescence spectra were measured on 29 normal tissue sites and 25 laryngeal lesions using 337-nm excitation. Four different multivariate analysis schemes were applied. Laryngeal fluorescence spectra from patients who had been administered delta-aminolevulinic acid (ALA) were obtained using 405-nm excitation and were classified using partial least squares discriminant analysis (PLS-DA). RESULTS: For autofluorescence spectra, logistic regression based on principal component analysis (PCA) or PLS, or PLS-DA all resulted in sensitivities and specificities around 90% for lesion vs. normal. Using ALA and 405-nm excitation gave a sensitivity of 100% and a specificity of 69%. CONCLUSION: Multivariate analysis of fluorescence spectra could allow classification of laryngeal lesions in vivo with high sensitivity and specificity. PLS performs at least as well as PCA, and PLS-DA performs as well as logistic regression techniques on these data.

Adult↗
PubMed1993

Dot-plot comparisons by multivariate analysis (DOCMA): a tool for classifying protein sequences.

A method aimed at classifying protein sequences without resorting to pairwise alignment is presented. Called DOCMA (DOt-plot Comparisons by Multivariate Analysis), it is based on a multivariate analysis of the pairwise dot-plots between all the sequences in the set. The dot-plots are first simplified by considering only the projections of the 'diagonal' segments of similarity onto the axes. From these projections a data matrix is built, in which each column is representative of the comparisons of one given sequence with all the other ones. This data matrix is then transformed into a distance matrix by a chi-squared analysis, from which the coordinates of the sequences in an orthonormal Euclidean space are obtained. The sequences are finally classified by a dynamic clustering procedure followed by a search for strong clusters. Application of this method to protein families such as the globins, the cytochromes c and the aminoacyl-tRNA synthetases shows that it is quite effective in delineating subgroups that contain even distantly related sequences.

Algorithms↗
PubMed2001

[Analysis of correlation between physical training and autonomic function by using multivariate analysis: establishing an indicator of health].

The purpose of this study is to assess the effects of physical training on autonomic nervous function. Twenty-two healthy subjects(12 males and 10 females) participated in this study. We inquired contents of the exercise for each subject. Continuous ECG was recorded during 320 seconds at supine resting. Respiratory rhythm was fixed at 0.25 Hz(15 time/min). On the basis of spectral analysis of heart rate variability, the autonomic nervous function was assessed by low frequency power(LF: 0.06-0.13 Hz) and high frequency power(HF: 0.16-0.50 Hz). The predominance of sympathetic nervous activity were indicated by the LF/HF and LF-HF. Relationship between physical training and autonomic function was analyzed by using multivariate analysis(partial correlation analysis). As the results, LF power and LF-HF negatively correlated with training time per day, energy expenditure per day, energy expenditure per week, and total energy expenditure, respectively(p < 0.05). HF power positively correlated with energy expenditure per day, energy expenditure per week, and total energy expenditure, respectively(p < 0.05). These results indicate not training time but energy expenditure alters the autonomic nervous function. This index may be useful for evaluating training effect on the promotion of health.

Adult↗
PubMed1998

A multivariate analysis of tumour biological factors predicting response to cytotoxic treatment in advanced breast cancer.

The study was designed to identify factors that could predict response to chemotherapy in breast cancer. A total of 173 patients with measurable or evaluable metastatic breast cancer were enrolled in a randomized trial between November 1987 and January 1991 to receive a monthly dose of 5-fluorouracil (500 mg m(-2)), epirubicin (60 mg m(-2)) and cyclophosphamide (500 mg m(-2)) either administered in four weekly doses or in an every-4-week dose as first-line cytotoxic treatment. In 103 evaluable patients we performed a multivariate analysis of the tumour biological factors, i.e. histological grade, oestrogen receptor (ER), progesterone receptor (PR), S-phase fraction (SPF), ploidy, p53, c-erbB-2, Bcl-2 and Bax expression, which showed significance in the univariate analysis according to treatment response, time to progression (TTP) or overall survival (OS). In the univariate analysis only SPF, grade and the proapoptotic protein Bax correlated with the response to cytotoxic treatment. In the multivariate analysis SPF had the strongest correlation, followed by grade and Bax. In the univariate analysis grade, PR, Bax and Bcl-2 correlated significantly with TTP, whereas in the multivariate analysis only PR showed a statistically significant correlation. In the univariate analysis PR and Bax correlated with OS and both retained its significance in the multivariate analysis. The factors that correlated significantly with the response to cytotoxic treatment in the univariate analysis, i.e. grade, SPF and Bax, seemed to predict independently the response to treatment in the multivariate analysis also. TTP and OS could be predicted partly by the same factors, although the association was quite weak. More studies and new tumour biological factors are needed to identify the group of breast cancer patients who get the most benefit from chemotherapy.

Analysis of Variance↗
PubMed1995

Factors predicting response to chemotherapy and survival in patients with metastatic or recurrent squamous cell cervical carcinoma: a multivariate analysis.

Different regimens of chemotherapy have been proposed with contradictory results in patients with metastatic and recurrent cervical carcinoma: this seems mainly due to the lack of information about prognostic factors. This study was aimed at identifying significant factors predicting response to treatment and survival in 140 patients treated with chemotherapy for advanced, recurrent, or persistent squamous cell cervical carcinoma. Age, performance status, histologic type and grade, previous irradiation, interval from start of primary treatment and from irradiation, site of tumor, and therapeutic regimen were considered as possible predictors of response and survival in multivariate analysis. By multivariate analysis, only performance status and interval from irradiation (> 1 year) were significant in predicting response to treatment, whereas interval from first diagnosis, site of tumor, and response to treatment were significant in predicting survival. None of the polychemotherapy regimens significantly improved survival, despite a fourfold increase in the costs when compared to the least expensive monochemotherapy regimen. Ethical and economical concerns should be considered when proposing aggressive regimens to the patients. Factors such as site of tumor, performance status, and interval from first treatment should be considered as minimal requirements for correct evaluation of newly proposed regimens.

Adult↗
PubMed2004

Is seminal vesicle ablation mandatory for all patients undergoing radical prostatectomy? A multivariate analysis on 1283 patients.

OBJECTIVE: With a shift in prostate cancer stage and a majority of patients operated nowadays with PSA levels <10 ng/ml, rates of seminal vesicle (SV) invasion found on radical prostatectomy specimens have decreased as compared to historical data. Since SV-sparing surgery may possibly have an influence on post-operative erectile dysfunction and urinary recovery, we tried to determine which patients could be safely spared SV excision during radical prostatectomy. MATERIAL AND METHODS: We used preoperative data from 1283 patients operated by radical retropubic prostatectomy--777 with serum PSA <10.0 ng/ml--to predict SV invasion on final pathological examination. Variables analyzed included age, digital rectal examination, serum PSA, biopsy Gleason score and percentage of biopsy cores invaded by prostate cancer. Statistical analysis included univariate, multivariate logistic regression analysis and receiver operating characteristic (ROC) curves. RESULTS: Out of 1283 patients, 137 (10.6%) had SV involvement, 41/777 (5.2%) with PSA <10.0 ng/ml, 16.1% in the 10-20 ng/ml range and 26.2% when PSA was >20 ng/ml. Percentage of biopsies affected by prostate cancer and biopsy Gleason score were significant predictors of SV invasion in multivariate analysis, both in the entire population and in the subset of patients with PSA <10.0 ng/ml (p < 0.0001). Probability graphs created for patients with PSA <10 ng/ml indicate a risk of seminal invasion <5% when Gleason score on biopsy is <7 or when the percentage of biopsies affected by cancer is <50%. CONCLUSIONS: Resection of SV might not be "oncologically" necessary in all patients undergoing RP when PSA levels are below 10 ng/ml except when biopsy Gleason score is > or =7 or when more than 50% of prostate biopsy cores show cancer involvement. SV-sparing surgery could be prospectively compared to standard retropubic prostatectomy in selected individuals analyzing potential benefits on erectile function and urinary continence.

Aged↗
PubMed2005

Sentinel node status in melanoma patients is not predictive for overall survival upon multivariate analysis.

Sentinel lymph node biopsy (SLNB) has become a widely accepted standard procedure in the staging of patients with cutaneous melanoma and absence of clinical lymph node metastases, although there is no final proof that SLNB influences overall survival in these patients. This study investigated the accuracy of SLNB and the clinical outcome of patients after a mean follow-up of 22 months. Between 1998 and 2003, SLNB was performed in 309 consecutive patients. Patients with one or more positive sentinel lymph nodes (SLNs) were subjected to selective lymphadenectomy (SL). Survival analyses were performed using the Kaplan-Meier approach. A Cox proportional-hazard analysis was used for univariate and multivariate analysis to explore the effect of variables on survival. Sentinel lymph nodes were identified in 299 of 309 patients (success rate: 96.8%). Of these, 69 (23%) had a positive SLN. The false-negative rate was 9.2%. Recurrence of disease to the regional lymph node basin (3.0%) and to the locoregional skin (2.6%) was rare in SLN-negative patients in contrast to SLN-positive patients (7.2 and 17.4%, respectively). The 3-year overall survival was 93 and 83% for SLN-negative and SLN-positive patients, respectively. Upon multivariate analysis, SLN status (P<0.001), Breslow thickness (P<0.02) and ulceration (P<0.026) were all found to be independent prognostic factors with respect to disease-free survival, whereas Breslow thickness proved to be the only significant factor with respect to overall survival.

Adult↗
PubMed1993

A multivariate analysis of prognostic factors in management of pineal tumor.

PURPOSE: A multivariate analysis of prognostic factors of treatment outcome of pineal tumor. METHODS AND MATERIALS: From February 1979 to June 1987, 25 patients with primary pineal tumors were treated in our department. Patients were treated with either AECL Co-60 unit or 10 MV linear accelerator to the primary tumor with an adequate margin or to the whole brain (median dose of 36 Gy) with or without a cone-down boost of 10 to 20 Gy. Craniospinal irradiation was performed in two patients with positive CSF cytology. Minimum follow-up was 40 months. Patients were further stratified according to tumor type. Group I consisted of seven patients with pineal germinoma. Group II included nine patients with nongerminoma, and Group III represented nine patients treated without a histological verification but clinical diagnosis. RESULTS: The relapse-free survival (RFS) of Group I patients was 100% and 86% at 2 and 5 years, respectively. Relapse-free survival was 55% and 21% at 2 and 5 years, respectively, for Group II patients. Six of 9 patients in Group II died of disease due to either local recurrence or tumor seeding. Eight of 9 patients in Group III remain no evidence of disease, and RFS was 89% at 2 and 5 years. Multivariate analysis revealed that tumor histology is the only significant prognosticator. Age, gender, type of surgical procedure, RT field, and tumor dose were not. Cox's regression model also failed to demonstrate a significant correlation of tumor seeding with the type of surgery. CONCLUSION: The type of tumors in the pineal region dictates the treatment outcome. Definitive radiation therapy is effective in controlling germinoma, whereas a more aggressive approach is needed to improve local control for nongerminoma. For a localized pineal lesion, we advocate that treatment can be tailored to the primary tumor with adequate margins. However, for locally advanced tumors whole brain or craniospinal irradiation should be considered. No definitive correlation between type of surgery and the probability of tumor seeding was identified.

Actuarial Analysis↗
PubMed1991

Multivariate analysis of metastasis risk in laryngeal carcinoma. I. Tumor factors.

A multivariate analysis with a logistic regression model was performed in 181 cases of patients affected by laryngeal cancer, in order to evaluate the histologic tumor factors in predicting metastatic risk. The patterns used were: grading, localization, extension to the extralaryngeal structures, mode of growth. Nodal metastases are less influenced by the localization and the differentiation grade of the tumor. Cancer extended to the extralaryngeal tissues and with "infiltrating" type of growth are significantly correlated with nodal involvement. Multivariate analysis seems to be the best method to identify the exact value of each histological parameter used in prognosis of laryngeal carcinoma.

Carcinoma↗
PubMed2003

Multivariate analysis of neuronal interactions in the generalized partial least squares framework: simulations and empirical studies.

Identification of spatiotemporal interactions within/between neuron populations is critical for detection and characterization of large-scale neuronal interactions underlying perception, cognition, and behavior. Univariate analysis has been employed successfully in many neuroimaging studies. However, univariate analysis does not explicitly test for interactions between distributed areas of activity and is not sensitive to distributed responses across the brain. Multivariate analysis can explicitly test for multiple statistical models, including the designed paradigm, and allows for spatial and temporal model detection. Here, we investigate multivariate analysis approaches that take into consideration the 4D (time and space) covariance structure of the data. Principal component analysis (PCA) and independent component analysis (ICA) are two popular multivariate approaches with distinct mathematical constraints. Common difficulties in using these two different decompositions include the following: classification of the revealed components (task-related signal versus noise), overall signal-to-noise sensitivity, and the relatively low computational efficiency (multivariate analysis requires the entire raw data set and more time for model identification analysis). Using both Monte Carlo simulations and empirical data, we derived and tested the generalized partial least squares (gPLS) framework, which can incorporate both PCA and ICA decompositions with computational efficiency. The gPLS method explicitly incorporates the experimental design to simplify the identification of characteristic spatiotemporal patterns. We performed parametric modeling studies of a blocked-design experiment under various conditions, including background noise distribution, sampling rate, and hemodynamic response delay. We used a randomized grouping approach to manipulate the degrees of freedom of PCA and ICA in gPLS to characterize both paradigm coherent and transient brain responses. Simulation data suggest that in the gPLS framework, PCA mostly outperforms ICA as measured by the receiver operating curves (ROCs) in SNR from 0.01 to 100, the hemodynamic response delays from 0 to 3 TR in fMRI, background noise models of Guassian, sub-Gaussian, and super-Gaussian distributions and the number of observations from 5, 10, to 20 in each block of a six-block experiment. Further, due to selective averaging, the gPLS method performs robustly in low signal-to-noise ratio (<1) experiments. We also tested PCA and ICA using PLS in a simulated event-related fMRI data to show their similar detection. Finally, we tested our gPLS approach on empirical fMRI motor data. Using the randomized grouping method, we are able to identify both transient responses and consistent paradigm/model coherent components in the 10-epoch block design motor fMRI experiment. Overall, studies of synthetic and empirical data suggest that PLS analysis, using PCA decomposition, provides a stable and powerful tool for exploration of fMRI/behavior data.

Algorithms↗
PubMed1995

Surgical mortality and morbidity in malignant obstructive jaundice: a prospective multivariate analysis.

OBJECTIVE: To construct prognostic scores using multivariate analysis for morbidity and mortality in jaundiced patients with malignant biliary obstruction. DESIGN: Prospective study. SETTING: 16 university and 12 general hospitals affiliated to the French Association for Surgical Research. MAIN OUTCOME MEASURES: Results of application of severity indexes for mortality and morbidity constructed from 17 variables. That for mortality was: 0.0497 x age + 0.9219 x American Society of Anesthesiologists' (ASA) grade + 0.0037 x serum bilirubin concentration minus 0.0239 x prothrombin time + 0.0001 x white cell count minus 5.593. That for morbidity was: minus 0.7499 x ASA grade + 0.0294 x prothrombin time + 1.4220 x cause (0 = carcinoma of bile duct, 1 = pancreatic cancer) minus 1.5080 x operation (0 = bypass, 1 = resection) minus 1.537. RESULTS: The scores correctly predicted mortality in 77% and morbidity in 65% (infective morbidity in 73%). CONCLUSIONS: We recommend that when the mortality index is negative operation should be the treatment of choice, and when it is positive the patient should be advised to have non-surgical palliative treatment. When the morbidity index is negative the risk of complications is high, and when it is positive the risk is low. The application of these indexes allows for better choice of patients suitable for operative treatment of malignant biliary obstruction.

Adult↗
PubMed2003

Multivariable analysis: a primer for readers of medical research.

Many clinical readers, especially those uncomfortable with mathematics, treat published multivariable models as a black box, accepting the author's explanation of the results. However, multivariable analysis can be understood without undue concern for the underlying mathematics. This paper reviews the basics of multivariable analysis, including what multivariable models are, why they are used, what types exist, what assumptions underlie them, how they should be interpreted, and how they can be evaluated. A deeper understanding of multivariable models enables readers to decide for themselves how much weight to give to the results of published analyses.

Biomedical Research↗
PubMed1994

Multivariable analysis of DNA ploidy, p53, and HER-2/neu as prognostic factors in endometrial cancer.

BACKGROUND: Several molecular-genetic alterations in endometrial cancers, including aneuploidy and aberrant expression of p53 and HER-2/neu, have been associated with poor prognosis. To determine the importance of molecular-genetic factors relative to more traditional surgical-pathologic prognostic factors, a multivariable analysis was performed. METHODS: Immunohistochemical staining for p53, HER-2/neu, estrogen receptor, progesterone receptor, and epidermal growth factor receptor was performed on frozen sections from 100 primary endometrial cancers. DNA ploidy was determined using computerized image analysis of Feulgen-stained touch preparations. In addition, information regarding surgical-pathologic features of the cancers was obtained. Univariable analysis was performed followed by multivariable analysis using Cox's proportional hazards model to identify variables predictive of poor prognosis. RESULTS: With univariable analysis, race, histologic type, stage, grade, myometrial invasion, estrogen receptor, progesterone receptor, ploidy, p53 and HER-2/neu were predictive of the presence of persistent or recurrent disease. In the multivariable analysis, only International Federation of Gynecology and Obstetrics stage (P = 0.005), grade (P = 0.005), myometrial invasion (P = 0.024), and ploidy (P = 0.028) were significant. CONCLUSIONS: Among molecular-genetic prognostic factors, DNA ploidy was the most strongly predictive of persistent or recurrent disease.

Aged↗
PubMed1989

A multivariate analysis of prognostic factors in adult Hodgkin's disease.

A multivariate analysis of the prognostic factors was carried out with a Cox model on 167 patients with Stage I-IV Hodgkin's disease. The following indicators were prospectively registered: Sex, age, clinical stage, systemic symptoms, histological type, number of involved areas, nodal size, mediastinal thoracic ratio (MTR), localization of mediastinal involvement according to the aortic arcus, hilar involvement, extent of radiotherapy, and extent and mode of chemotherapy. A linear logistic analysis showed that most of the indicators were interrelated. This emphasizes the necessity of a multivariate analysis in order to assess the independent influence of each of them. The three main prognostic indicators for complete remission achievement were age, clinical stage and involvement of lung hili. The three main prognostic indicators for disease-free survival were sex, size of mediastinal involvement (MTR) and use of alternating chemotherapy COPP/ABVD in the initial treatment strategy. The most important prognostic factor for survival was clinical stage followed by hilar involvement, age, and nodal size. The coefficients estimated by Cox regression analysis can be used to combine the important prognostic factors into a single index for each patient at presentation. The prognostic index for each patient would allow the identification of cases with poor prognosis who might benefit from additional initial treatment.

Antineoplastic Combined Chemotherapy Protocols↗
PubMed1990

Improvement of diagnosis in the non-invasive assessment of coronary artery disease: enhanced evaluation of quantitative exercise 201thallium imaging by multivariate analysis.

STUDY OBJECTIVE: The aim was to evaluate discriminant analysis, performed in patients without prior myocardial infarction, in enhancing the diagnostic value of quantitative exercise 201thallium scintigraphy. DESIGN: All clinical, electrocardiographic, and scintigraphic variables were first subjected to a univariate analysis. Afterwards a discriminant analysis was done. PATIENTS: 135 patients (104 male) were studied. Age was 24-70 years, mean 55 years. MEASUREMENTS AND MAIN RESULTS: Two discriminant analyses were done. In the first analysis, the ability to detect the presence of coronary artery disease was tested. Significant variables were: (1) history of angina, (2) sex, (3) quantitative analysis of 201thallium scintigraphy, (4) age, (5) ischaemic ST response, (6) angina during the test, and (7) the pressure-rate product. The sensitivity, specificity, and accuracy of classification using the discriminant function were 91%, 87%, and 90%, respectively. The sensitivity was higher than when using only visual interpretation (sensitivity 70%; p less than 0.0002) or quantitative interpretation (sensitivity = 66%; p less than 0.0001) of thallium scans, without significant loss of specificity (p less than 0.5488; p less than 0.6875). In the second analysis, a discriminant function was calculated to detect multivessel disease. Five input variables were selected: (1) number of vessels with stenosis predicted by quantitative analysis, (2) number of vessels with stenosis predicted by visual analysis, (3) ischaemic ST response, (4) sex, (5) angina during the test. Multivariate analysis showed an increase in sensitivity when compared with visual interpretation (78% v 55%; p less than 0.0043) and quantitative analysis (66%; p less than 0.0156). Using the classification, the discriminant function was more accurate than visual analysis (81% v 69%) or quantitative analysis (77%). CONCLUSIONS: The results show that multivariate analysis of non-invasive test results in quantitative thallium exercise testing allows convenient use for clinical purposes with improved results.

Adult↗
PubMed2005

The time-related changes of the importance of prognostic factors in breast cancer. A sequential multivariate analysis of 1423 Japanese patients.

The value of individual prognostic factors may change dependent on the length of the follow-up period, if some factors have their greatest prognostic potential immediately after operation. It is not clear how long these factors keep their prognostic relevance. We retrospectively examined data on 1423 surgically treated Japanese patients with primary breast cancer between 1983 and 2002. Survival analysis was done starting at 2.5-yearly intervals after operation and follow-up in the first analysis started at the time of the operation. The changing importance of the prognostic factors during different follow-up periods was investigated by univariate and multivariate analysis. Based on multivariate analysis, tumor size retained its prognostic value even up to 7.5 years after operation, whereas the age, vascular involvement, ER and PgR showed a changing influence on prognosis dependent on the length of the follow-up period. The prognosis of patients some years after operation is necessarily different from the initial prognosis established after operation. Detecting the changing importance of prognostic factors could provide new biological insights that might otherwise be missed, and may help determine the most appropriate clinical use of various factors.

Adult↗