| [Background] Surgical site infection is the most common healthcare associated infection in surgical patients; which affects the medical quality, increases medical expenses, even lead to death of the patient. The prevention and control of surgical site infection is cost-effective, therefore, it is of great significance to implement effective surgical site infection management. At present, transparency supervision has become a new management tool used to improve the quality of of hospital infection management. The United States has publicly reported surgical site infection indicators, and achieved some success. The effect of public report mainly depends on two factors:1. the selection of indicators for public report; 2. the quality of risk adjustment of the indicators. Therefore, it is the premise of the implementation of transparency supervision to select the surgical site infection indicators appropriate to report publicly and to explore the optimal risk adjustment method.[Purpose] To determine the surgical site infection indicators appropriate to report publicly; to explore the optimal risk adjustment method for surgical site infection indicators and lay the methodological foundation for implementation of transparency supervision of the provider.[Methods]1. The surgical site infection indicators appropriate to report publicly were determined by Delphi method. The design of expert advisory table was based on the literature review and group discussion. The surgical site infection indicators for screening were determined by the literature review, and the criteria for selecting indicators was determined by focus group discussion method. Using Likert 5 scoring method, experts scored the indicators according to selection criteria. Advisory expert group was composed of the experts from the field of epidemiology and statistics, hospital infection management and health management. The indicators were screened by critical value method, and the surgical site infection indicators appropriate to report publicly were determined after two rounds of expert consultation.2. The optimal risk adjustment method of surgical site infection indicators were explored by applying surgical site infection rate of caesarean section as an example. Through the risk adjustment, the influence of the risk adjustment factor to the surgical site infection rate was eliminated and the comparability of the indicator was realized.(1) To determine the risk factors.Systematic review:through the literature retrieval, literature screening, literature quality assessment and determination of risk factors, the risk factors of caesarean section were ascertained. Expert consultation:based on the risk factors determined by systematic review method; experts in field of obstetrics and hospital infection management were consulted, and then the complementary risk factors were added. So the final risk factors collection were completed.(2) To determine the risk adjustment factors.Literature review:viewpoints about which risk factors needed adjustment were collected; Expert consultation:the experts in the field of obstetrics and hospital infection management were consulted, the risk factors which need to be adjusted were ascertained and then the risk adjustment factors were defined. According to the definition, the risk adjustment factors were selected from the risk factors collection.(3) To construct a multifactor risk adjusted model and calculate the standardized infection ratio.All the 2506 cases of caesarean section from 2013 to 2014 of T hospital were collected as the reference data. The risk adjustment models were established respectively by logistic stepwise regression and poisson stepwise regression; the dependent variable was the occurrence of surgical site infection, the independent variables were risk adjustment factors. The c-index was used to compare the predictive power of the logistic regression model and NNIS risk index models. Pearson goodness-of-fit was used to compare the goodness-of-fit of poisson regression model and NNIS risk index models.Standardized infection ratio is the risk adjusted index of surgical site infection rate which eliminates the influence of risk adjustment factors on infection rate, and it is defined as the number of observed infections divided by the number of expected infections which is calculated by risk adjustment model. Standardized infection ratio less than 1 indicates that the performance of hospital infection management is better than the benchmark, while greater than 1 indicates inferior to the benchmark. The standardized infection ratios of each quarter and each year of 2013-2014 of T hospital were calculated for intrahospital comparisons over time. The cases of caesarean section in August 2015 were collected from T hospital, X hospital and S hospital.The standardized infection ratios were calculated for interhospital comparisons.EXCEL2007 was used for data input and processing, SPSS 17.0 and STATA12.0 were used for data analysis.[Results]1. A total of 9 selection criteria including impact, improvability and inclusiveness were developed. According to the criteria, a total of 13 surgical site infection indicators suitable for public report including cesarean section surgical site infection rate, herniorrhaphy surgical site infection rate and adenomammectomy surgical site infection rate were selected. Of the 13 chosen indicators,9 indicators were procedure-specific surgical site infection rate, accounting for 69.2% of the total indicators.2. By system review,28 risk factors were identified for cesarean section surgical site infection,2 risk factors were added from expert consultation and then 30 risk factors were determined. The risk factors reflecting the patient’s susceptibility to infection which is difficult to intervention were defined as risk adjustment factors. According to this definition,10 risk adjustment factors were screened, accounting for one-third of the total number of risk factors, namely BMI, ASA score, diabetes, hypertension, with hepatitis B or C, amniotic cavity infection, premature rupture of membranes, multiple pregnancy, scar uterus and age.3. The two risk adjusted model constructed respectively by logistic regression and poisson regression included the same variables (BMI and ASA score), standardized infection ratios calculated by the two model were highly consistent. The C-index of the new logistic regression risk adjustment model is higher than that of the NNIS risk index model (0.63 and 0.60, respectively); the Pearson goodness-of-fit of the new poisson regression risk adjustment model is higher than that of the NNIS risk index model(p=0.95 and p=0.85, respectively). The procedure duration does not conform to the definition of risk adjustment factor. The SIR of T hospital in 2013 and in 2014 were 1.4 and 0.9, respectively. In the 8 quarter from 2013 to 2014 of T hospital, the SIR of 4 quarters is less than 1, while 4 quarters greater than 1. The SIR of T hospital, X hospital and S hospital in August 2015 were 1.0,1.3 and 1.1.[Conclusions]Specific-procedure surgical site infection rate is suitable for public report. Systematic review combined with expert consultation is scientific and feasible for determining the risk factors collection. The risk factors reflecting the patient’s susceptibility to infection which is difficult to intervention were defined as risk adjustment factors. Accordingly, the risk adjusted index eliminates the influence of the patient’s susceptibility on infection rate and can reflect the real performance level of hospital infection management. The new risk adjustment model constructed by logistic regression and poisson regression are better than the NNIS risk index model. The two methods have their own advantages, and they are highly consistent with the results. As the surgical site infection rate decreases, poisson regression will be more dominant. The standardized infection ratio is comparable, and then can be used for intrahospital comparisons over time, and interhospital comparisons.[Innovation and Limitation]1. From the perspective of transparency characteristics; a total of 9 selection criteria including impact, improvability and inclusiveness were developed. According to the criteria, a total of 13 surgical site infection indicators suitable for public report including cesarean section surgical site infection rate, herniorrhaphy surgical site infection rate and adenomammectomy surgical site infection rate were selected.2. From the perspective of transparent supervision of the provider, the risk adjustment method of public indicators based on comparability and the standardized infection ratio of pulic indicators were determined.Limitation:Only one hospital’s data were used to construct the risk adjustment model. Although it is reasonable to construct risk adjustment model using the data of last 2-3 years from a benchmark hospital. However, the ideal reference data for constructing risk adjustment model is the data of the last 2-3 years from multiple hospitals. Only cesarean section surgical site infection rate was risk adjusted in this study, the applicability of risk adjustment method which was proposed in this study need to be further verified. |
PDF Full Text Request