The Study Of Hospital Infection Risk Factors And Predict Models

Background:Hospital infection,also known as hospital acquired infection.The incidence of hospital infection are difference between hospitals in China because of the unbalanced development.Hospital infection can cause huge losses to patients and hospitals,it has become an important public health problem,which seriously restricts the medical service quality improved.Object:To investigate the distribution regulation of the hospital infection in a hospital,and discuss the risk factors for the hospital infection.Build the risk appaisal model for individual and trend-season model to provide reference for the hospital infection control.According the antibiotic resistance,to direct the clinical medication and to control the occurence of the hospital infection and bacterial resistance.Method:1.Search through the literature by developing search strategy,obtaining the studies of the hospital infection risk factors,extracting data,and take the system review and meta-analysis.2.According to the results of the system review,retrospective investigate the information of inpatients and hospital infection,compare the incidence and risk factor by department classification,to find the risk factors of hospital infection.3.Using artificial neural network model,to establish the individual risk appraisal model for the hospital infection.Software using SPSS18.0 Neural Network module.4.Using Excel and SPSS18.0,to establish the incidence prediction model for the hospital infection based on the trend-season model.5.Identificate the pathogens and antibiotic resistance by the DL-96 bacterial reagent system and DL-96 board producted by the Zhuhai Deere biological engineering co.LTD.Result:1.There are 11 studies have been included,and 12 risk factors into model,including age(≥60 years),hospital stays(≥4 weeks), malignancy tumor,the trachea pipe cutting or intubation,surgery,immunosuppressants using,interventional examination,hormone using,respirator using,puncture or incision drainage,antibiotic using,central venous catheteration,their pooled OR and 95%CI were respectively:1.053(95%CI, 1.013-1.095)、1.527(95%CI, 1.214-1.920)、2.252(95%CI, 1.758-2.885)、2.495(95%CI, 1.888-3.297)、2.785(95%CI, 2.021-3.839)、2.843(95%CI, 1.815-4.455)、2.876(95%CI, 2.165-3.821)、2.895(95%CI, 2.137-3.921)、4.370(95%CI, 2.479-7.703)、4.544(95%CI, 2.408-8.575)、4.587(95%CI, 1.624-12.956)、5.489(95%CI, 3.479-8.661).2.The situation of the hospital infection:The hospital infection rate was not significant difference between January 2013 and June 2015(P=0.738).The average hospital infection rate is 2.56%.The intensive care unit,the cadre’s ward and the department of hematology were the top three departments.Among the internal medicine ward,the department of hematology has the highest prevelance of hospital infection.Among the surgical ward,the department of orthopaedics and general surgery were the highest.Among the other department ward,the highest was oncology department.The top three location of the hospital infection were:lower respiratory tract,urinary tract and upper respiratory tract.Univariate analysis hospital infection showed that hospital infection was associated with age,hospital stay and disease type(P<0.05) among every ward.Among the surgical ward,hospital infection rate was also associated with season and urinary catheter using(P<0.05).Among the cadre’s ward,hospital infection rate was also associated with gender(P<0.05).Among the other department ward,hospital infection rate was also associated with gender and breathing machine using(P<0.05).In multi-factor analysis,age and hospital stay had a signifficant effect on the hospital infection(P<0.05).Blood disease(OR=2.454)was the risk factor of the hospital infection among internal medicine ward. Second quarter(OR=0.696)was the protective factor among the surgical ward. Tumor(OR=1.630), neurological disease(OR=1.349),respiratory disease(OR=1.801) and digestive disease(OR=1.693) was the risk factors among the other department ward.3.The neural network prediction model for the hospital infection among the internal medicine ward,surgical ward and other department ward,the area under the ROC curve were greater than 0.8,the cadre’s ward was 0.756.4. The trend-season model for hospital infection incidence predict the third quarter of 2015 is 2.51%,actual incidence was 2.43%,the relative error is small,prediction effect is good.5.Main pathogenic bacteria distribution and drug resistance:(1)the clinical specimens was mainly in the sputum.Pseudomonas aeruginosa was the first separate pathogen in 2013 and 2014.(2)the highest percentages of antibiotics for pseudomons aeruginosa was levofloxacin in 2013 and 2014.(3)the highest percentage of antibiotics for acinetobacter baumannii were amikacin and piperacillin/tazobactam in 2013 and 2014.(4)the highest percentage of antibiotics for klebsiella pneumoniae were cefotaxime and compound sulfamethoxazole in 2013 and 2014.(5)the highest percentage of antibiotics for escherichia coli was piperacillin in 2013 and 2014.(6)the highest percentage of antibiotics for staphylococcus aureus was penicillin in 2013 and 2014.Conclusion:All kinds of strains were isolated with high resistance in 2013 to 2014,formed a serious challenge for clinical work.Although the average hospital infect incidence was 2.56%,lower than the level of first-class hospital infection incidence,but the main department incidence of hospital infection is still serious.According to the isolate types,drug resistance and the incidence of hospital infection and risk factor of main department in the two years,to improve the use of antibiotics,strictly abide by the sterile system,try our best to control the incidence of hospital infection. Neural network model and senson-trend model are established to help complete disease control and prevention through individual and entirely prediction,so as to achieve the purpose of effectively reduce the incidence of hospital infection.