ObjectiveTo explore the selection problem of independent variables and stepwise regression method for multiple logistic regression analysis. MethodsAccording to the data of the case-control investigation for coronary heart disease, age (X1), hypertension history (X2), hypertension family history (X3), smoking (X4), hyperlipidemia history (X5), animal fat intake (X6), weight index (X7), type A personality (X8), and coronary heart disease (CHD, Y) were analyzed by SPSS 18.0 software. The multiple logistic regression analysis was done and the differences of risk factors were compared among 6 kinds stepwise regression variable selection method. ResultsThe univariate analysis showed that no difference was found between CHD group and non-CHD group in age distribution (P=0.116). But the multivariate logistic regression analysis showed that, comparing to population over 65 years old, age was a protective factor on the low age groups (OR< 45=0.100, 0.000 to 0.484, P=0.020; OR45-54=0.051, 0.003 to 0.975, P=0.048). If the age was defined as categorical variable, the risk factors for coronary heart disease were animal fat intake (X6), type A personality (X8), hypertension history (X5) and age (X1), respectively (P < 0.05). If the age was defined as a continuous variable, the effect of age (X1) was not statistically significant (P=0.053). The common risk factors were intake of animal fat (X6) and type a personality (X8) by six kinds method of stepwise variable selection. In addition, the risk factor also included hyperlipidemia history (X5) (forward-condition, forward-LR, forward-wald), hypertension family history (X3), age (X1) (backward-condition, backward-LR) and hypertension history (X2) (backward-wald). ConclusionStepwise regression method should be used to analyze all the variables, including no statistically significant independent variables in univariate analysis. If the categorical variable is regarded as continuous variables, some information may be lost, and even the risk factors may be missed. When the risk factors are not the same by several stepwise regression variable selection method, it should be combined with clinical and epidemiological significance, as well as biological mechanisms and other professional knowledge.
Objective To study the influence factors of surgical site infection (SSI) after hepatobiliary and pancreatic surgery. Methods Fifty patients suffered from SSI after hepatobiliary and pancreatic surgery who treated in Feng,nan District Hospital of Tangshan City from April 2010 and April 2015 were retrospectively collected as observation group, and 102 patients who didn’t suffered from SSI after hepatobiliary and pancreatic surgery at the same time period were retrospectively collected as control group. Then logistic regression was performed to explore the influence factors of SSI. Results Results of univariate analysis showed that, the ratios of patients older than 60 years, combined with cardiovascular and cerebrovascular diseases, had abdominal surgery history, had smoking history, suffered from the increased level of preoperative blood glucose , suffered from preoperative infection, operative time was longer than 180 minutes, American Societyof Anesthesiologists (ASA) score were 3-5, indwelled drainage tube, without dressing changes within 48 hours after surgery, and new injury severity score (NISS) were 2-3 were higher in observation group (P＜0.05). Results of logistic regression analysis showed that, patients had history of abdominal surgery (OR=1.92), without dressing changes within 48 hours after surgery (OR=2.07), and NISS were 2-3 (OR=2.27) had higher incidence of SSI (P＜0.05). Conclusion We should pay more attention on the patient with abdominal surgery history and with NISS of 2-3, and give dressing changes within 48 hours after surgery, to reduce the incidence of SSI.
ObjectiveTo analyze the risk factors of multidrug-resistant organism (MDRO) nosocomial infection, and to provide the scientific basis for the prevention and control of MDRO nosocomial infection.MethodsPatients with MDRO in Chengdu Shangjin Nanfu Hospital from 2014 to 2015 were retrospectively collected. The patients were divided into the MDRO nosocomial infection group and the MDRO non-nosocomial infection group. The MDRO infection/colonization, bacterial strain type, specimens type and distribution characteristics of clinical departments were analyzed. Single factor and multiple factor logistic regression analysis were used to analyze the risk factors of MDRO nosocomial infection.ResultsA total of 357 patients of MDRO infection/colonization were monitored, of which 147 times (144 patients) were with nosocomial infections and 213 times (213 patients) were without nosocomial infections. MDRO nosocomial infection incidence rate/cases incidence rate were 0.18%. A total of 371 MDRO bacterial strains were detected, of which 147 (39.62%) were with nosocomial infection and 224 (60.38%) were without nosocomial infections. The MDRO non-nosocomial infections included 175 strains (47.17%) in community infection and 49 strains (13.12%) in colonization. Carbapenem-resistant Acinetobacter baumannii (52.83%) was the main MDRO strains. Sputum (57.14%) and secretion (35.04%) were main specimens. The top three departments of MDRO nosocomial infection strains were orthopedics (32.65%), ICU (27.89%), neurosurgery (13.61%). ICU [odds ratio (OR)=3.596, 95% confidence interval (CI) (1.124, 11.501), P=0.031], surgical history [OR=2.858, 95%CI (1.061, 7.701), P=0.038], indwelling urinary catheter [OR=3.250, 95%CI (1.025, 10.306), P=0.045], and using three or more antibiotics [OR=4.228, 95%CI (1.488, 12.011), P=0.007] were the independent risk factors of MDRO nosocomial infection.ConclusionEffective infection prevention and control measures should be adopted for the risk factors of MDRO nosocomial infection to reduce the incidence rate of MDRO nosocomial infection.
ObjectivesBased on the historical data of inpatients, a logistic regression model was established. It aimed to identify the influencing factors of patient's admission scheduling decisions and compare them with the actual scheduling rules, so as to discover the differences and deficiencies.MethodsWe extracted data of outpatients and inpatients in Department of Respiration in West China Hospital of Sichuan University from January 1st, 2016 to December 31st, 2016, and standardized the original dataset. We established the binary multivariate logistic regression model through R software and ‘glm’ package.ResultsThe analysis of multi-factor logistic regression showed that the effect of the five variables (type of medical insurance, time of registration, waiting time, type of disease and admission priority) on patient schedule was statistically significant.ConclusionsThe logistic regression model constructed in this study has a good effect on patient planning, which is helpful to provide decision support for admission schedule through identification factors.
Objective To investigate the status of deciduous caries and early childhood caries among 3-5 year-old children of Uyghur and Chinese in Urumqi, and to explore the correlative factors of early childhood caries. Methods According to the criteria recommended by the Third National Oral Health Investigation, and Oral Health Surveys: Basic Methods of World Health Organization, the deciduous caries of 474 Urghur and Chinese children aged from three to five in nine kindergartens were clinically examined. Data were collected by questionnaire from their parents, and the result waw analyzed using Logistic regression analysis. Results The result of logistic regression analysis showed that the variables including nationality, frequency of drinking milk, eating cookie or drinking sweet beverage before sleep, brushing teeth with help, and educational background of the mother were closely related to the incidence of infantile caries. Conclusion The nationality, frequency of drinking milk, eating cookie or drinking sweet beverage before sleep, brushing teeth with help, and educational background of the mother are correlative factors of early childhood caries. Necessary methods for prevention of deciduous caries must be taken into consideration as early as possible, and the bilingual propaganda for preventing and treating caries should be also highly emphasized.
Objective To analyze the factors associated with the adoption of targeted therapy in patients with human epidermal growth factor receptor 2 (HER2)-positive breast cancer and to generate evidence to inform decision-making on public security policy regarding innovative anticancer medicines for the benefit of patients. Methods The study population comprised female patients diagnosed with HER2-positive breast cancer and treated at Fujian Cancer Hospital from 2014 to 2020. The patients were eligible for targeted therapy. The demographic and sociological characteristics and clinical information of patients were extracted from the hospital information system. We performed binary logistic regression analysis of factors associated with the adoption of targeted therapy in patients with HER2-positive breast cancer. We also divided the participants into two groups according to their tumor stage for subgroup analysis. Results A total of 1 041 female patients with HER2-positive breast cancer were included, among them, 803 received targeted therapy. In September 2017, molecular-targeted medicines for HER2-positive breast cancer began to be included in the local basic health insurance program. Only 282 (35.1%) patients adopted targeted therapy before September 2017, after which this number increased to 521 (64.9%). Among the patients who adopted targeted therapy, most were formally employed (45.8%) and enrollees of the urban employee health insurance program (66.0%). Among those who did not adopt targeted therapy, most were unemployed (42.4%) and enrollees of the resident health insurance program (50.0%). Binary logistic regression analysis revealed that patient occupation, gene expression of estrogen receptor, tumor stage, surgery or not, radiotherapy or not, and undergoing treatment before or after September 2017 were correlated with the adoption of targeted therapy (P<0.05). Conclusions Inclusion of targeted medicines for HER2-positive breast cancer in the health insurance program substantially increased the overall administration of these therapies. Individual affordability is a critical factor associated with the application of targeted therapy in eligible patients. Future policies should enhance the public security of patients with a relatively weak ability to pay and provide insurance coverage for innovative anti-cancer medicines.
Objective To explore the positive rate of emphysema in groups under Low-dose CT screening, then take the regression analysis on related risk factors for emphysema. Methods A total of 1 175 volunteers involved in low-dose CT screening and completing the questionnaire were collected and taken the CT quantitative assessment for emphysema, then the positive rate of emphysema was calculated. Questionnaire data were collected and non-conditional Logistic regression was used to analyze the factors in the questionnaire. Results Ninety-seven cases of emphysema had been detected in 1 175 volunteers, and the positive rate was 8.26%. The positive rate for the males and the females was 9.90% (71/717) and 5.68% (26/458), respectively. Three risk factors (smoking, second-hand smoking, history of chronic bronchitis) were screened out by Logistic regression. Conclusions According to the results of the regression analysis, smoking, second-hand smoking and history of chronic bronchitis are main risk factors for emphysema. Some effective measures could be made against emphysema in high risk population. In that way the morbidity and perniciousness of emphysema could be reduced.
Objective To analysis correlation factors for preoperative sudden death of patients with type A aortic dissection in order to determine clinical management strategy.?Methods?We retrospectively analyzed clinical data of 52 patients with type A aortic dissection who were admitted in Department of Cardiothoracic Surgery of the Affiliated Drum Tower Hospital of Nanjing University Medical School from January 2003 to January 2010. According to the presence of preoperative death, all the patients were divided into two groups, 9 patients in the preoperative sudden death （PSD）group including 7 males and 2 females with their mean age of 52.0±12.1 years；43 patients in the control group including 31 males and 12 females with their mean age of 51.5±10.9 years. Univariate and multivariate logistic regression analysis were used for analysis of preoperative factors related to sudden death.?Results?Univariate analysis result showed 7 candidate variables：body mass index （BMI, Wald χ2=2.150, P=0.143）, time of onset （Wald χ2=2.711, P= 0.100）, total cholesterol （TC, Wald χ2=1.444, P=0.230）, low density lipoprotein cholesterol （L-C, Wald χ2=1.341, P=0.247）, aortic insufficiency （AI, Wald χ2=2.093, P=0.148）, aortic sinus involvement （Wald χ2=3.386, P=0.066）and false lumen thrombosis （Wald χ2=7.743, P=0.005）. Multivariate logistic regression analysis showed that BMI （Wald χ2=4.215, P=0.040, OR=1.558）and aortic sinus involvement （Wald χ2=4.592, P=0.032, OR=171.166 ）were preoperative risk factors for sudden death, and thrombosed false lumen （Wald χ2=5.097, P=0.024, OR=0.011）was preoperative protective factor for sudden death.?Conclusion?Type A aortic dissection patients with large BMI and/or aortic sinus involvement should receive operation more urgently than others and patients with thrombosed false lumen may have relatively low risk of preoperative sudden death.
ObjectiveTo establish logistic regression analysis model to evaluate the diagnostic efficacy of breast imaging report and data system (BI-RADS) ultrasound signs in forecasting malignant risk of breast lesions. MethodUltrasound graphic materials of 1 660 breast lesions diagnosed during January to September 2011 were retrospectively studied and standardized by BI-RADS. Pathology results were regarded as gold standard reference. Ultrasound signs with significant efficacy after single-factor logistic regression were evaluated in multi-factor logistic regression model to predict the malignant risk of breast lesions. ResultsEighteen ultrasound signs of breast lesions on BI-RADS were included in the final regression model. Among them, Cooper ligaments stretch, echogenic halo, skin thickening, axillary lymph node abnormalities, structural distortions and speculation had high OR values of 30 or more and had higher specificity than 90%. The diagnosis values of regressions model were high, with a sensitivity of 84.5%, specificity 95.5% and accuracy 91.4%. The area under ROC curve was 0.964 and prediction accuracy was 91.0%. ConclusionsThe logistic regression model based on BI-RADS ultrasound signs of breast lesions has high diagnostic values in detecting breast cancer.
Objective To analyze and screen the risk factors of both immunohistochemistry and pathology for lung cancer lymphatic metastasis, and to build a mathematical model for preliminary evaluation. Methods By conducting retrospective studies, the information of lung cancer patients in the General Hospital of Air Force from 2009 to 2011 were collected. Both single and multiple unconditional logistic regression analyses were applied to screen total 27 possible factors for lymphatic metastasis. After the factors with statistical significance were selected, the relevant mathematical model was built and then evaluated by means of receiver operating characteristic (ROC) analysis. Results A total of 216 patients were included. The single analyses on 27 possible factors showed significant differences in the following 10 factors: pathological grade (P=0.00), age (P=0.00), tumor types (P=0.01), nm23 (P=0.00), GSTII (P=0.01), TTF1 (P=0.01), MRP (P=0.01), CK14 (P=0.02), CD56 (P=0.02), and EGFR (P=0.03). The multiple factors unconditional logistic regression analyses on those 10 risk factors screened 4 relevant factors as follows: pathological grade (OR=2.34), age (OR=1.02), nm23 (OR=1.66), and EGFR (OR=1.47). Then a mathematical diagnostic model was established based on those 4 identified risk factors, and the result of ROC analysis showed it could improve the diagnostic sensitivity and specificity compared with the single factor mathematical diagnostic model. Conclusion Pathological grade, age, nm23, and EGFR are related with lung cancer lymphatic metastasis, and all of them are the risk factors which have higher adjuvant diagnostic value for lung cancer lymphatic metastasis.