Nonconditional multiplicity logistic regression model analysis showed that hospitalization day(χ~2=29.418,P<0.001),invaded operations in hospital(χ~2=19.807,P=0.000),the use of hormone(χ~2=(127.96),P=0.000),the variation antibiotics(χ~2=95.121,P=0.000) were risk factors.

Nonconditional multiplicity logistic regression model analysis showed that age(χ 2=14.120,P=0.000),hospitalization day(χ 2=11.532,P=0.001) and invaded operations in hospital(χ 2=6.827,P=0.009) were risk factors. The prophylactic use of probiotics was an independent protective factor(χ 2=12.943,P=0.000).

Univariate non-conditional logistic regression model analysis showed that cigarette smoking was significantly associated with BV.Odds ratio (OR) for smokers was 3.273 (95% confidence interval (CI)1.663~6.442,P=0.001).

variable analysis and unconditional logistic regression model analysis indicate that cat is one of the risk factors of HFRS infection, the OR of the risk factor is 5.07 (P<0.01).

Multiplicity logistic regression model analysis showed time of recovery to EEG and duration of seizures were the risk factors of secondary epilepsy after viral encephalitis (P < 0.05).

Cox proportional hazard regression model analysis confirmed that MSI was a very important prognostic indicator for NSCLC patients (χ 2=15.4126,P= 0.0001 ), and lymph node metastasis had some prognostic significance in NSCLC patients (χ 2=6.6108,P= 0.0101 ).

6. By Kaplan-Meier survival analysis and Cox regression model analysis, VEGF-C expression and infiltration depth are found to be independent factors for patients' prognosis.

Conclusion The new method is effective and feasible for diagnosis and treatment of multivariable multicollinearity in the logistic regression model analysis.

Results Conditional Logistic regression model analysis showed:some factors were found to have statistical association with coronary heart disease(CHD),such as history of hyperlipemia(OR=24.9),mental work(OR=10.9),history of hypertension(OR= 7.2 ). A type character(OR=3.6),eating sweet foods(OR=2.5) and salted foods(OR= 1.9 ).

Multivariate logistic regression model analysis showed that suicide impulsion and psychiatric disorders were significantly associated with COMT Val/Val, their OR values were 0.052 (95%CI: 0.006-0.437), 2.917 (95%CI: 1.097-7.760).

Result: The multiple logistic regression model analysis showed that the significant factors concerned with HPV were cervical erosion, abnormal uterine bleeding, TCuIUD and vaginitis, the OR of which was 2.686, 4.538, 4.263 and 3.414 respectively .

4. By Cox regression model analysis, the expression of E-cad and the infiltration depth were found to be the independent factors for prognosis of gastric cancer patients.

In this paper, regression model analysis of data shows that United States cities have more fires and more fire fighters per capita but fewer fire fighters per fire than do foreign cities.

Diabetes remains an independent risk factor for these events even in a multivariate logistic regression model analysis.

Around 8?mg/week of MTX was effective in half of the RA patients in the current study, and male sex was the only factor associated with a good response to MTX from a multivariate regression model analysis.

Regression model analysis revealed that low body weight (>amp;lt;44?kg) was associated with an 8.3-fold (95% CI, 3.7-18.9) and a 6.8-fold (95% CI, 3.0-15.6) risk of having low aBMD at the spine and hip, respectively.

Methods: An analysis on recurrence-related factors was made by Cox's proportional hazards model analysis and logistic multiple linear regression model analysis in 212 patients with transitional cell carcinoma treated surgically from 1995～2001.

A case-control study of stomach cancer was carried out in Changle county.There were 181 matched pairs,the patients were verified histologically from 1985 to 1988,the controls were matched with the patients by age,sex and area of residence.The contents interviewed included:dietary factors,smoking and drinking factors and genetic factors etc.Th logistic regression model analysis showed that eating too much moldy food,having bad eating habits were risk factors,wheras taking fresh vegetables,bean products...

A case-control study of stomach cancer was carried out in Changle county.There were 181 matched pairs,the patients were verified histologically from 1985 to 1988,the controls were matched with the patients by age,sex and area of residence.The contents interviewed included:dietary factors,smoking and drinking factors and genetic factors etc.Th logistic regression model analysis showed that eating too much moldy food,having bad eating habits were risk factors,wheras taking fresh vegetables,bean products were protecting factors in stomach cancer.In addition,familiar congregation was more observed in stomach cancer pstients than that In the controls.

A case-control study method was used to investigate the nsk factors of Hemorrhagic Fever with Renal Syndrome (HFRS) in the suburbs of Xi an from October 1987 to January 1989. The results of single variable analysis and unconditional Logistic regression model analysis indicate that the main risk factors of HFRS infection are skin damage, farmwork, food contaminated with excrement and, or urine of rodents, oral mucosa damage and consumption of unhcated leftovers. By unconditional Logistic regression model...

A case-control study method was used to investigate the nsk factors of Hemorrhagic Fever with Renal Syndrome (HFRS) in the suburbs of Xi an from October 1987 to January 1989. The results of single variable analysis and unconditional Logistic regression model analysis indicate that the main risk factors of HFRS infection are skin damage, farmwork, food contaminated with excrement and, or urine of rodents, oral mucosa damage and consumption of unhcated leftovers. By unconditional Logistic regression model analysis, the ORs of the risk factors mentioned above are 2.80, 2.06, 2.01, 2.33 arid 2.68 respectively, and the selected variablity have statistical significance. The results of stratified analysis show that the two variables of food contaminated with excrement and/or urine of rodents, and oral mucosa damage may be co-determinants. It is suggested that the main routes of transmission of HFRS in the endemic area may be skin damage in the course of farmwork, and peroral infection.

A case-control study of hemorrhagic fever with renal syndrome (HFRS) at the suburbs of Xi'an was performed from October 1988 to January 1989. The results of single .variable analysis and unconditional logistic regression model analysis indicate that cat is one of the risk factors of HFRS infection, the OR of the risk factor is 5.07 (P<0.01). The role of cat in transmission of HFRS shouldn't be underestimated.