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 37 records · Page 2Linked to original sources
PubMed1992

Flow cytometric DNA analysis in Japanese colorectal cancer. A multivariate analysis.

Using Cox's proportional hazard model, we performed a multivariate analysis of survival data from 126 patients who underwent curative resection for colorectal cancer from 1971 to 1985. Flow cytometric DNA measurements were carried out using paraffin wax-embedded tissue blocks. Fifty-four per cent of the cases were found to be aneuploid and 46% were diploid. The depth of invasion of the tumor, nodal status, liver metastasis, peritoneal dissemination, DNA ploidy pattern, histology, macroscopic classification of the tumor, age, sex and site were tested in a survival analysis. The initial seven factors were found to be the significant prognostic variables in a univariate analysis. A multivariate analysis shows liver metastasis, DNA ploidy pattern and peritoneal dissemination as being the significant discriminants of survival (P = 0.0001, 0.0022, 0.0119, in this order). Therefore, nuclear DNA ploidy pattern in colorectal cancer is considered to be an independent prognostic factor.

Adult↗
PubMed2002

Pharmacologic efficacy in gastric variceal rebleeding and survival: including multivariate analysis.

BACKGROUND: Therapy with beta-blocker and nitrate has been reported to improve survival of patients with bleeding esophageal varices and to decrease esophageal rebleeding. However, there is little information available concerning the efficacy of these medications on rebleeding risk and survival in gastric variceal bleeding after initial hemostasis. METHODS: We conducted an open trial to observe the roles of beta-blocker and nitrate in the long-term outcome of bleeding gastric varices. Eighty-three patients were included and evaluated on the basis of age, gender, gastric variceal size, associated esophageal variceal size, Child-Pugh classification, existence of hepatoma and portal vein thrombosis, beta-blocker or nitrate therapy, and follow-up histoacryl injection. Survival analysis and multivariate analysis with the Cox proportional hazards model were performed to evaluate independent risk factors. RESULTS: Larger gastric varices have been shown to be the only risk factor for rebleeding (adjusted odds ratio, 4.50; 95% CI, 1.30-15.59). beta-Blocker and nitrate did not significantly reduce the incidence of rebleeding (adjusted odds ratio, 0.37; 95% CI, 0.08-1.66). Although medical treatment was shown to improve the overall survival by Kaplan-Meier method (p < 0.01), multivariate analysis showed Child-Pugh class B or C and advanced hepatoma with portal vein thrombosis to be the real independent risk factors that influence survival (Child-Pugh class B or C odds ratio, 2.72; 95% CI, 1.53-4.84; portal vein thrombosis odds ratio, 6.99; 95% CI, 2.42-20.16). beta-Blocker and nitrate did not significantly prolong survival independently. CONCLUSIONS: beta-Blocker and nitrate did not decrease the risk of rebleeding and did not improve the overall survival independently. The poor prognosis was correlated with Child-Pugh class B or C, and the advance hepatoma, with portal vein thrombosis.

Adrenergic beta-Antagonists↗
PubMed1986

The correlation of pretreatment transurethral resection of prostatic cancer with tumor dissemination and disease-free survival. A univariate and multivariate analysis.

A multivariate Cox's hazard function analysis was performed on the prognostic variables selected from 240 patients with localized carcinoma of the prostate who received external beam radiotherapy to analyze the association between the method of biopsy and disease-free survival. The patients received 4500 cGy to the pelvis followed by a 2-week treatment rest and then an additional 2000 cGy to the reduced prostatic volume. Median follow-up was 4 years (range, 1-9 years). The 5-year actuarial local control rate was 91%. There was no difference in local control relative to the method of biopsy (needle biopsy or transurethral resection of the prostate [TURP]). The 5-year actuarial risks of metastases were 28% and 48% for the needle biopsy and TURP groups, respectively (P less than 0.01). The 5-year disease-free survival for the needle biopsy and TURP groups were 55% and 42%, respectively (P less than 0.05). This difference maintained statistical significance for Stage C and Grade III tumors. A multivariate analysis demonstrated that "method of biopsy" was the third most powerful variable after serum acid phosphatase level and modified Broders' grade in predicting disease-free survival. Patients who had TURP had an almost twofold higher relative risk of disease progression than those who had needle biopsy. This study established the correlation of the method of biopsy with a lower probability of disease-free survival over and above the information obtained from the clinical stage, histologic grade, presence or absence of symptoms of obstruction, acid phosphatase level, and lymph node status. The association of TURP with tumor dissemination is discussed with reference to the mechanism of metastasis formation and prevention.

Biopsy, Needle↗
PubMed2001

[Prognostic significance of expression of p53, bcl-2 and bax in squamous epithelial carcinoma of the larynx--a multivariate analysis].

BACKGROUND AND OBJECTIVE: Conventional clinicopathologic parameters do not accurately reflect the clinical outcome of patients with head and neck carcinoma. The establishment of additional prognostic factors that may give insight into the biologic features of a tumor is therefore an essential goal. The present study analyses the expression patterns of p53, bcl-2, and bax with regard to their relationships with conventional tumor parameters and to their prognostic significance in patients with laryngeal squamous cell carcinoma. PATIENTS/METHODS: Paraffin-embedded tissue sections of 88 primary laryngeal squamous cell carcinomas diagnosed and treated between 1986 and 1996 were investigated for p53, bcl-2, and bax protein expression by immunohistochemistry. The mean follow-up time was 45.9 months. RESULTS: Bcl-2 immunoreactivity was positively correlated with an advanced clinical stage, a high T category, regional lymph node metastasis, and a high histological grading. Significant relationship between clinicopathologic parameters and p53 or bax expression were not detectable. The age of the patients, advanced disease, positive bcl-2 expression, and a high level of p53 expression were significantly associated with shortened disease-specific survival in univariate analysis. In multivariate analysis, age, clinical stage, and p53 expression had independent prognostic value. CONCLUSIONS: Although expression of p53 and bcl-2 was found to be clinically relevant in univariate analysis, only p53 but not bcl-2 was an independent predictor of patient outcome. This superiority of p53 in multivariate analysis points to its central role within cell cycle and death regulation, with which it influences two important parameters of tumor progression.

Aged↗
PubMed1992

Predictors of dating violence: a multivariate analysis.

A multivariate approach was used to determine the pattern of predictors associated with engaging in dating violence. Predictors were selected whose relationship to dating violence has been established by earlier research: attitudes toward violence, sex-role attitudes, romantic jealousy, general levels of interpersonal aggression, verbal aggression, and verbal and physical aggression received from one's partner. Participants included 305 introductory psychology student volunteers (227 females and 78 males) who completed a set of scales related to dating relationships. Expecting different patterns of predictors to emerge for men and women, we performed separate multiple regression analyses for each. Of the set of predictors employed, receipt of physical violence from one's partner emerged as the largest predictor of expressed violence for both men and women. In addition, higher scores on attitudes toward violence and verbal aggression, and less traditional sex-role attitudes emerged as significant predictors of expressed violence for men. For women, less accepting attitudes toward violence, more traditional sex-role attitudes, feelings of romantic jealousy, higher general levels of interpersonal aggression, and verbal aggression were predictive of expressed violence. The implications of our findings for future research are discussed.

Adult↗
PubMed1996

Inappropriate use of bivariable analysis to screen risk factors for use in multivariable analysis.

The use of bivariable selection (BVS) for selecting variables to be used in multivariable analysis is inappropriate despite its common usage in medical sciences. In BVS, if the statistical p value of a risk factor in bivariable analysis is greater than an arbitrary value (often p = 0.05), then this factor will not be allowed to compete for inclusion in multivariable analysis. This type of variable selection is inappropriate because the BVS method wrongly rejects potentially important variables when the relationship between an outcome and a risk factor is confounded by any confounder and when this confounder is not properly controlled. This article uses both hypothetical and actual data to show how a nonsignificant risk factor in bivariable analysis may actually be a significant risk factor in multivariable analysis if confounding is properly controlled. Furthermore, problems resulting from the automated forward and stepwise modeling with or without the presence of confounding are also addressed. To avoid these improper procedures and deficiencies, alternatives in performing multivariable analysis, including advantages and disadvantages of the BVS method and automated stepwise modeling, are reviewed and discussed.

Confounding Factors, Epidemiologic↗
PubMed1994

Risk factors related to liver metastasis in colorectal carcinoma: a multivariate analysis of clinicopathologic and immunohistochemical variables.

Specimens from 48 consecutive patients undergoing surgery for colorectal carcinoma and having synchronous or metachronous liver metastases (Group 1) and those from 52 consecutive patients who had no evidence of hepatic metastases within at least 5 year after colorectal resection (Group 2) were selected and compared using a multiple logistic regression analysis. A multivariate analysis using a stepwise logistic regression revealed six independent risk factors significantly related to hepatic metastases. In addition, the following logistic regression model was obtained from this analysis. P = exp a/(1 + exp a): a = 3.524(SM-V) + 2.731(Ex-V) + 2.718(E/M) + 2.562(Lo) + 1.858(p53) + 1.941(HIR) - 4.397, where P is the probability of hepatic metastasis given six independent risk factors (E/M, Ex/M ratio; Lo, location; HIR, host inflammatory cell reaction). When the estimated probability "P" in the above logistic regression model is more than 0.55 after an examination of surgical specimens, we must consider adjuvant chemotherapy and closely monitor the patient to ensure early detection of hepatic metastases.

Adenocarcinoma↗
PubMed2006

Multivariate analysis of interactions between phytoplankton biomass and environmental variables in Taihu Lake, China.

Phytoplankton variation in large shallow eutrophic lakes is characterized by high spatial and temporal heterogenity. Understanding the pattern of phytoplankton variation and the relationships between it and environmental variables can contribute to eutrophic lakes management. In this study Taihu Lake, one of the largest eutrophic fresh water lake in China, was taken as study area. The water body of Taihu Lake was divided into five regions viz. Wuli bay (WB), Meilian Bay (MB), West Taihu Lake (WTL), Main Body of Taihu Lake (MBTL) and East Taihu Lake (ETL). Concentrations of chlorophyll-a and the related environmental variables were determined in each region in the period 2000-2003. Factor analysis and multivariate analysis were applied to evaluate the interactions between phytoplankton variation and environmental variables. Results showed that the highest average concentrations of TN, TP and Chl-a were observed in WB, followed in a descending order by MB and WTL, and the lowest concentrations of TN, TP and Chl-a were observed in MBTL and ETL. Chl-a and TP concentrations in most regions (except ETL) declined during the study period. It suggested that to some extent the lake was recovering from eutrophication. However, persistent ascending of TN and NH(4)-N in all five regions indicated the deteriorating of water quality in the study period. Results of multivariate showed that the relationships between phytoplankton biomass and environmental variables varied among regions. TP illustrated itself a controlling role on phytoplankton in WB, MB, WTL and MBTL according to the significant positive relations to phytoplankton biomass in these regions. Nitrogen could be identified as a limiting factor to phytoplankton biomass in ETL in view of the positive correlations between TN and phytoplankton and between NH(4)-N and phytoplankton. Spatial variation of interactions between phytoplankton and environmental parameters suggested proper eutrophication control measures were needed to restore ecological system in each region of Taihu Lake.

Biomass↗
PubMed1986

Predicting acute pelvic inflammatory disease: a multivariate analysis.

A multivariate logistic regression analysis of patient symptoms and signs and laboratory findings associated with the diagnosis of acute pelvic inflammatory disease was performed with use of data from 628 women who were clinically diagnosed as having the disease for the first time at the University of Lund, Sweden. In 414 women (65.9%) acute pelvic inflammatory disease was laparoscopically confirmed. We developed a mathematical model that correctly predicted 87.0% of the cases of acute pelvic inflammatory disease and had an overall correct classification rate of 75.6%. Variables that were good predictors of acute pelvic inflammatory disease were purulent vaginal discharge, erythrocyte sedimentation rate greater than or equal to 15 mm/hr, positive gonorrhea result, adnexal swelling on bimanual examination, and rectal temperature greater than or equal to 38 degrees C. Furthermore, we developed "mixed model I" and "mixed model II," which combine simple clinical parameters and laparoscopy in varying degrees. In mixed model I the sensitivity, specificity, and overall classification values were 93%, 67.2%, and 84.5%; in mixed model II these values were 100%, 67.2%, and 89.2%. Use of relatively simple and reproducible clinical parameters can identify those women who would most benefit from laparoscopy to diagnose acute pelvic inflammatory disease.

Acute Disease↗
PubMed2001

[An exploratory positioning analysis: three-mode multivariate analysis for semantic differential data].

The purpose of this study was to propose an exploratory positioning analysis method based on a three-mode multivariate statistical model for analyzing semantic differential (SD) data. This new method can be easily implemented by using a structural equation modeling (SEM) program such as LISREL and CALIS, because the mean structure and covariance structure of this method are expressed as submodels of SEM. Two necessary conditions for checking identifiability of this model are shown. It is also shown that this model is scale equivariant for SD data.

Humans↗
PubMed1998

The path analysis approach for the multivariate analysis of infant mortality data.

PURPOSE: This paper reviews the use of the Path Analysis (PA) methodology in health determinants modeling, with special reference to infant mortality modeling. METHODS: A review of the literature on PA applications in the modeling of infant mortality and similar problems is presented, together with a discussion of the conceptual basis of PA and its relation to other multivariate statistical techniques. Important aspects of the technique are discussed: 1) criteria for path formulation; 2) parameter estimation methods; 3) direct, indirect, spurious, and joint effects; and 4) goodness-of-fit and modification indices. RESULTS AND CONCLUSION: The review of the literature suggests that PA represents a methodological improvement regarding multivariate techniques used in modeling some health-related issues. PA allows investigation of more complex models, providing information that could have been previously overlooked, such as how the interrelations among independent variables in a model affect the dependent ones.

Humans↗
PubMed1988

An alternative approach for statistical analysis of kidney transplant data: multivariate analysis of single-center experience.

To avoid the center effect and the possible hidden interactions of multicenter studies, the validity of the Cox Proportional Hazards Model for the analysis of a single-center kidney transplant program was tested, considering 287 renal transplants performed in a 10-year period. The inclusion of type of donor and main immunosuppressive drug as covariates in the model did not violate the proportionality assumption of the Cox model. According to this method, the following covariates were significant in predicting graft survival: cyclosporine, type of donor, good human leukocyte antigen (HLA)-A and HLA-B match (DR data were not considered), highest percentage of reactive antibodies against panel cells, and nephroangiosclerosis as a primary renal disease. Cyclosporine did not significantly improve graft survival in living related donor transplants. Pretransplant blood transfusions, cold ischemia time, and donor ABO blood group were initially significant but dropped out in the step-down procedure. Recipient's age at transplant, cyclosporine, HLA-A and HLA-B match, and nephroangiosclerosis were significant in predicting patient survival. It was concluded that using long-term data of cadaveric and living related renal transplants either on azathioprine or cyclosporine is a valid way to perform multivariate analysis of single-center transplant programs that do not have large samples.

Adult↗
PubMed1993

Results of high-dose thoracic irradiation incorporating beam's eye view display in non-small cell lung cancer: a retrospective multivariate analysis.

PURPOSE: To review the University of Michigan clinical experience in nonsmall cell lung cancer using high-dose thoracic irradiation (> or = 60 Gy) so that a starting dose for our prospective dose-escalation study could be determined. METHODS AND MATERIALS: Eighty-eight consecutive patients diagnosed with medically inoperable or locally advanced, unresectable nonsmall cell lung cancer were identified who were treated with thoracic irradiation alone to a minimum total dose of 60 Gy (uncorrected for lung density). All patients except four (95%) underwent computed tomography scanning for treatment planning that included beam's eye view display for tumor and critical structure localization. All patients were treated with standard fractionation in a continuous course to uncorrected total doses ranging from 60 to 74 Gy (median, 67.6 Gy). RESULTS: The median follow-up exceeds 24 months for all surviving patients (range, 12 to 78 months). The median survival time was 15 months, and the 2- and 3-year overall actuarial survival rates were 37% and 15%, respectively. Survival was significantly different between stage of disease (p = .004) and N-stage (p = .002) by univariate analysis. In a multivariate analysis, stage becomes the only characteristic significantly associated with outcome. The median time to local progression for 86 evaluable patients was 29 months. Stage (p = .0003), T-stage (p = .0095) and N-stage (p = .027) were significantly different with respect to local progression-free survival by univariate analysis. However, only stage was prognostic for local progression-free survival by multivariate analysis. There was no difference between large volume treatment (inclusion of the contralateral hilar and supraclavicular lymph nodes) and small volume treatment (exclusion of these elective nodal sites) with respect to local progression-free survival (p = .507) or survival (p = .520). With regard to dose, there was no significant difference between patients who received > 67.6 Gy and patients who received < or = 67.6 Gy with respect to local progression-free survival (p = .094) or survival (p = .142). Within the Stage III subgroup, local progression-free survival (p = .018) and survival (p = .061) were longer favoring the high-dose group of patients. Despite these doses, disease progression within the irradiated field was the predominant first site of treatment failure. CONCLUSION: This retrospective study has shown that it is feasible to deliver uncorrected tumor doses as high as 70 Gy using standard fractionation in NSCLC with acceptable morbidity. Local control remains a significant problem. These data indicate justification for a starting dose in a prospective radiation dose-escalation study.

Adult↗
PubMed1998

[Statistical analysis of oropharyngeal squamous cell carcinoma--multivariate analysis of prognostic factors and evaluation of therapeutic modalities].

Ninety-one cases of oropharyngeal squamous cell carcinoma initially treated at Keio University Hospital between July 1981 and June 1996 were reviewed retrospectively. There were 83 males and 8 females, aged from 29 to 83 years old, with an average age of 62.7. The primary lesion was located in the lateral wall in 52 patients (57.1%), the superior wall in 23 (25.3%), the anterior wall in 14 (15.4%) and the posterior wall in 2 (2.2%). Double cancer was detected in 21 patients (23.1%). The patients were divided into two groups according to the initial main treatment of the primary lesion without regard to chemotherapy: 72 patients (79.1%) who received curative radiotherapy with or without salvage surgery, and 14 patients (15.4%) who underwent curative surgery with or without preoperative and/or postoperative radiation. The remaining 5 patients were treated by chemotherapy alone. Prior to the above treatments 50 patients (54.9%) received neoadjuvant chemotherapy (NAC). Survival distributions were estimated by the Kaplan-Meier method as univariate analysis, and compared by the generalized Wilcoxon test. The overall five-year cumulative survival rate was 55.6%. The five-year survival rates according to stage (UICC classification, 1987) were as follows: stage I (11 cases), 70.7%; stage II (12 cases), 63. 6%; stage III (30 cases), 52.3%; and stage IV (38 cases), 52.5%. Significant clinicopathological variables that influenced survival were: (1) T stage (p = 0.0075); (2) age (p = 0.0274); and (3) location of primary lesion (p = 0.0400). The results of multivariate analysis by Cox's proportional hazards model identified T stage as a significant independent prognostic factor. Evaluation of the therapeutic modalities led to the following conclusions. (1) Differences in the initial treatments of the primary lesion were not reflected in the outcome. (2) Salvage surgery for residual or recurrent tumor contributed to improving the survival. The superior wall type, in particular, seemed to be a good indication for salvage surgery. (3) Although the limitations of radiotherapy are not defined clearly, we have to determine the indications for radical resection of tumors resistant to radiotherapy with reconstruction. (4) The response rate of NAC reached 85.4%, but there were no significant differences in survival between the group that underwent NAC and the other group in any other subset analyses. (5) Among the patients who underwent NAC, the responder (CR + PR) group showed a better five-year survival rate (61.3%) than the non-responder (NC + PD) group (42.9%), but the difference was not significant.

Adult↗
PubMed1992

Correlates of family planning acceptance: a multivariate analysis.

A multivariate generalized linear model is developed to analyze the correlates of family planning acceptance in India using data from the 1980 National Fertility and Mortality Survey, undertaken in Maharashtra. "Variables such as caste, education, number of living sons, [and] number of living children have been included in the analysis. The application of the model revealed that the existing number of children/son has a closer association with acceptance of contraception rather than the education or caste variables."

Asia↗
PubMed2002

Classification of fermentation performance by multivariate analysis based on mean hypothesis testing.

Multivariate analysis, such as principal component analysis and artificial autoassociative neural networks, is currently extensively applied to feature capturing, physiological state recognition, fault detection and bioprocess control. However, it is not clear which process variable should be selected as an important input for multivariate analysis to analyze physiological conditions and/or bioprocess performance a priori. An efficacious method to select more informative process variables from the repository of historical data is highly desired. In this study, we focused on a premodeling step. Mean hypothesis testing (MHT) was used to select appropriate variables for multivariate analysis. Fermentation data sets were classified into two classes "good" and "bad" according to the MHT results. The results showed that selecting discriminating process variables from the historical database by MHT enhanced the overall effectiveness of multivariate analysis prior to principal component analysis and artificial autoassociative neural network model creation.

Journal Article↗
PubMed1993

[Man among the primates. His position after multivariate analysis of his mandible].

Multivariate analysis are applied to 31 ensembles (species of Primates or human breeds), in 31 mandibular linear parameters. A reciprocal averaging show that first factorial plane reproduce the Simpson classification, simplified in 4 groups, with 63.3% of whole inertia. Discriminant analysis calculated with 13 parameters, situates these 4 groups in 3 planes. The D2 Mahalanobis generalised distance is calculated between all groups. The most contributive parameters to generalised distances are inter-molar width and breadth between foramen mentales.

Animals↗