Cse.msu.edu



Background Information for MRSA Health Surveillance SystemThis is a two-phase study to address the high rates of infection by a specific resistant staphylococcal bacterium in children who live in the Atlanta area. First, we will look at children who received care from Children’s Healthcare of Atlanta (CHOA) at Egleston, Hughes Spalding, and Scottish Rite hospitals and have a positive culture due to methicillin resistant Staphylococcus aureus (MRSA). We will compare these children with those who have a methicillin sensitive Staphylococcus aureus culture (MSSA). The outcome of the study in this phase consists of three components: (1) We will develop the methodology to extract and integrate the variables of interest from the medical information and laboratory databases at these locations so that comparisons can be made between those children who are infected by the resistant bacteria to those children who are not. (2) We will develop efficient information monitoring techniques to prospectively identify patients at the time of admission to the hospital, who have had a history of resistant Staphylococcal infection. (3) Additionally, we will provide automatic integration of relevant medical and laboratory variables in real-time for these patients. Another result of this phase of investigation is to develop an automated system of determining whether or not patients are infected by a MRSA strain originating from the hospital or from the community at large. This will be based on whether or not these patients become infected within a specified time of their admission to the hospital, e.g., 48 hours, and the specific antimicrobial susceptibility pattern of the bacterium isolated from the patient (phenotype of the bacterium). In the second phase of this research, we plan to geo-code databases on all children who have positive cultures for Staphylococcus aureus from January 2002 through December 2006 from the hospitals and use geographic information system tools to determine which risk factors may contribute to infection by these resistant staphylococcal strains. For example, presence of clustering of patients infected by CA-MRSA will be identified based on the mapping of these patients’ residences. Incidence of infection and other population information to determine certain risk factors can be derived by applying GIS to data obtained from US Census Bureau. This population information can be grouped according to certain boundaries, e.g., block groups, blocks, or census tracts. This research project will serve to link communication and information technologies at three different hospitals in order to have a better understanding of the epidemiology of children infected by CA-MRSA. The results generated by this research will directly and indirectly impact the delivery of healthcare and assist with the development of health policy guidelines as it pertains to infection control and healthcare quality within these three children hospital sites. This research can be seen as a step towards the HSI vision of shaping the next health care transformation (1). Development of a system to combine data and information on children infected with certain resistant bacteria from the three children’s hospitals servicing Atlanta area has the advantage of providing a seamless access of information and may serve as a testbed towards enabling children health care transformation to disease- prevention and health-focused (1). Concretely, this powerful tool can be used to look at the impact and costs associated with hospitalization of these children. Infection control and quality assurance programs can also use the information from such a system to critically look and address some of the health guidelines and specific costs associated with containing the spread of resistant bacteria. This integration of medical information with laboratory information is also useful from a research standpoint as it creates the infrastructure needed to obtain data more efficiently. Although the existing technology at Egleston and Scottish Rite sites provides for identification of patients who have a history of “multiple resistant organisms”, this project will refine that process so that real time identification, specifically of children infected by MRSA, can be linked to a database for collection of relevant medical and laboratory information for prospective data analysis. Currently, there is no integration or method of “flagging,” prospectively or retrospectively, those patients with MRSA at Hughes Spalding. Moreover, there is no direct linkage of information between Hughes Spalding and the other two hospitals. Because many physicians and other healthcare providers will provide services for all three locations, developing a linkage of information systems will facilitate communication and, hence, better delivery of healthcare at all three sites. 2. Broader impacts Analysis of the data will provide valuable information regarding the epidemiology of those children who are infected by CA-MRSA. Specifically, the information gained from using GIS will determine what risk factors lead to the acquisition of this particular MRSA which originated from the community. The real time aspect of the project will define the changing epidemiology of this bacterium as it moves from the community into hospitals. This information gained has a significant impact on our healthcare system as we strive to develop infection control policies to contain the transmission of resistant bacteria within the hospital setting. The data can also provide insights as to the costs of treating such bacterial infections in our hospitalized children. Development of this system will also improve the timing in which these children are identified and result in a more efficient delivery of treatment. Perhaps more importantly, development of real time data collection and integration across different informational systems will provide the backdrop needed, so that this technology can be applied to population-based databases involving other emerging infectious disease pathogens. One area of surveillance might include the Active Bacterial Core Surveillance of the Emerging Infections Program Network, a collaboration between CDC and state health departments and universities participating in the Emerging Infections Program Network(2). This collaboration was developed in response to the increasing number of emerging infectious diseases of which CA-MRSA is one of them. Developing the capability to efficiently integrate epidemiologic and clinical data across different surveillance systems is an extremely valuable tool in the battle to decrease the spread of these pathogens within our communities. Although GIS tools have been used in other medical areas, there are few studies using GIS to look at clinical and molecular epidemiology of resistant bacteria in the US. There are major health policies that can be impacted by findings based upon population-based studies using GIS technology. For instance, communities identified to be at high risk for a specific phenotype of MRSA can be targeted to receive services including empiric treatment tailored against that particular circulating phenotype. 3. Background and Significance Antibiotic resistant bacteria are becoming an increasing problem in the United States and worldwide as they cause serious and, often times, life-threatening infections in the face of a limited armaterium of antibiotic agents. Resistant bacteria usually originate from hospitals whereby selection pressure from frequent antibiotic exposure leads to the development of different mechanisms of resistance. However, one resistant bacterium today is believed to have originated from the community and not from the hospital—community-associated methicillin resistant Staphylococcus aureus (CA-MRSA). Nationally, the incidence of CA-MRSA infections is increasing at an alarming rate (3, 4). Morbidity associated with infection caused by this bacterium range from minor skin or soft tissue infections (SSTI) to severe illnesses (pneumonia, septic shock, bone infections) resulting often times in death (3, 5-8). The clinical presentation is believed to be related to virulent factors of CA-MRSA (3). CA-MRSA reflect a particular strain of resistant Staphylococcus aureus, namely, USA clone 300. These strains have a typical antimicrobial susceptibility pattern which is characteristic of the phenotype which originated from the community-- resistance to the antibiotics oxacillin or resistance to both oxacillin and erythromycin. We propose to develop a model to systematically merge existing microbiology laboratory database information with electronic medical record information to identify patients with resistant Staphylococcus aureus. From this centralized database, we begin to differentiate which patients have infection by bacterial strains that originated from the community versus those whose strains are acquired from within the hospitals. There are few studies that have identified which risk factors are associated with infection by CA-MRSA. For example, few studies have systematically addressed whether or not race/ethnicity, younger age, geographic location (i.e., specific neighborhoods), crowding, or socioeconomic status increase the risk for infection. Paramount to improving the healthcare delivery to children who are infected with this resistant bacterium is quick identification of those children who are potentially infected so that proper and correct antibiotic therapy can be given. We propose to use geographic information system tools to identify potential risk factors for the acquisition of this resistant bacterium. Additionally, we propose to develop a system of rapid identification of children who may have these risk factors for resistant bacterial infection at the time they present for healthcare. Identifying those children who have risk factors for infection by CA-MRSA based on area of residence or other demographic information can improve the delivery of preventative or empiric treatment for clinical diseases caused by such resistant bacterial infections. Using two approaches will provide key information regarding the epidemiology of those children who are infected by CA-MRSA. Aims ? To merge databases from laboratory, medical information databases, and electronic medical records and identify children who have been infected with both resistant and non-resistant Staphylococcus aureus through automated information update monitoring and tracking techniques. ? To identify risk factors for CA-MRSA using geographic information system tools and visual and automated data mining methods. Data Collection Assumptions:Data variables of interest. For patients who had Staphylococcus aureus isolated from a clinical culture, the following information will be obtained from the laboratory (See Appendix I for list of laboratory variables): Patient medical record number, body site where culture was obtained, hospital location of patient at the time of specimen collection, date of culture collection, time for culture to become positive, bacterial isolate specimen number and complete antimicrobial susceptibility pattern. Demographic information on all patients from whom Staphylococcus aureus was isolated will be obtained through the Medical Information Department at each children’s hospital. (See Appendix I for list of medical information variables.) In addition, information on patient days, by hospital unit, will be obtained for hospital administrative databases. Merging of medical information with laboratory information systems (See Appendix II for Timeline for Proposal). It is anticipated that this phase of the proposal will be completed during the first quarter of the funding period. At Scottish Rite and Egleston Hospitals. All clinical culture results from the period January 2002 through December 2006 will be obtained by downloading information from computer systems used by Egleston and Scottish Rite campuses. Duplicate isolates will be excluded. This information will be translated into a format by which GIS technology can be applied. At Hughes Spalding Hospital. Similarly, at Hughes Spalding, all clinical culture results from the same period will be obtained. Methodology to extract and merge data from laboratory and medical information systems at Hughes Spalding Hospital will involve the following: The design of a consolidated database that will combine laboratory database of Hughes Spalding Children’s Hospital with its medical information database in an effective manner. This requires the design of a consolidated database schema such that the consolidated database can be used efficiently for several purposes: (1) timely collecting and integrating of medical data and laboratory data of those patients who have a positive culture; (2) timely detecting and flagging of those patients who have a history of methicillin resistant Staphylococcus aureus (MRSA) infection; (3) timely tracking of all patients and the interested variables that present with positive culture for MRSA; and (4) providing efficient translation of consolidated data into a format by which GIS tools can be applied. The concrete methods for combining and linking both laboratory and medical information systems at Hughes Spalding Hospital include (1) techniques for creating and maintaining a clinical cultures positive database for Staphylococcus aureus from 1/2002 to 12/2006; (2) techniques for dynamic joins of laboratory and medical information of the patients in real time with user-friendly graphical interfaces for pediatric doctors and researchers to use; and (3) techniques for automated extraction of interesting data items from the consolidated database for performing statistical analysis and data mining. Data Distribution Requirements:Establishing a database prospectively at all three hospitals. Active surveillance for Staphylococcus aureus infections will be performed on all patients with positive cultures. This phase of the project will be started during the second quarter of the funding period and will be ongoing thereafter. This collection will occur in real-time and involve the following measures: Timely and efficient tracking of significant updates on patients with positive clinical cultures, including the list of variables we have identified (see Appendix I). An example of such monitoring services is to track patients whose cultures became positive within 48 hours of admission. Providing end-user level facilities, such as GUI, visualization tools, data mining tools, to allow pediatricians and pediatric researchers to analyze patients’ data and perform statistical data analysis in real time. Providing system-level facilities to support complex triggers for monitoring significant updates of laboratory and clinical information. The engineering and development of a real-time database update monitoring system on top of the consolidated database, which not only will link laboratory database of Hughes Spalding Children’s Hospital to its medical information database automatically in real time, but also will provide efficient and reliable methods and techniques for identifying, tracking and alerting in real time the significant updates on both laboratory data and clinical data of patients of interest. Geo-coding databases of all children infected with Staphylococcus aureus based on antibiotic resistance pattern. We will geo-code all patients with positive Staphylococcus aureus cultures for the study period and will begin this process during the second half of the funding period. This geo-coded dataset will then be sorted based on patients’ antibiotic resistance pattern as defined in Appendix I. Specifically, we will group patients into those with CA-MRSA phenotype, non CA-MRSA phenotype and MSSA case controls. Using ArcView 9.1 (Environmental Systems Research Institute (ESRI), Redlands, California) and statistical software (e.g., SPSS, SAS), we will determine if certain demographic traits are associated with risk for infection for the respective groups. Population information to look at possible risk factors including housing conditions (e.g., overcrowding), race, and household income will come from census data. For example, Fridkin et al suggested in their population-based study the risks in Atlanta associated with increased incidence of infection included being younger than two years or black (9). Similarly, risks associated with either infection or colonization by CA-MRSA in studies of adults include history of incarceration or HIV infection(10). Spatial clustering of isolates will also be assessed to determine whether specific strains of resistant MRSA occur in specific communities. Analysis plan. Data will be collected both retrospectively and prospectively in order to obtain complete data for the time period of January 2002 through December 2006. The data will be used to determine monthly rates of infection or colonization due to CA-MRSA. The incidences and rates of CA-MRSA, HA-MRSA and nosocomial-MRSA will be determined. The denominator for each of these determinations is based on the total number of cultures positive for Staphylococcus aureus. We will also determine the incidence and rate of patients who presented to the emergency department for evaluation and treatment of SSTI. The denominator for this will be based on the total monthly number of patients seen in the emergency department. Finally, we will determine rates of infection or colonization from MRSA isolates that are CA-MRSA phenotype but cultures were obtained and became positive after 48 hours of admission. These results will be compared to other MRSA phenotypes, which were collected from patients after 48 hours of admission. Comparisons of clinical diagnoses, length of hospital stay, and demographic information will be made between those patients who had cultures positive for CA-MRSA by phenotype and those who culture for HA-MRSA. Additionally, similar information will be obtained from a similar group of patients with cultures positive for MSSA. The laboratory and medical information variables will be merged into a single database, which will be analyzed using SPSS version 11 (SPSS, Inc.) and SAS V.8 (SAS Institute, Cary, NC). Prevalence and overall incidence for positive cultures due to CA-MRSA, non-CA-MRSA, MSSA per year will be determined. Rates of CA-MRSA, MSSA, Nosocomial-MRSA and HA-MRSA skin and soft tissue infections will be determined and analyzed using time series analyses. Study protocol will be submitted for approval by Georgia Institute of Technology, Children’s Healthcare of Atlanta, Grady Health System, and Morehouse School of Medicine Institutional Review Boards.Potential Use of Findings Results from this study will inform the healthcare community about incidence of CA-MRSA and identify specific risk factors for infection. This information will be necessary in order for us to develop specific health policy guidelines within communities at high risk for infection. Recommendations on which antibiotics should be selected for empiric therapy can be based on the information gained from this study. For example, if a patient presents for a skin infection and has one of the identified risk factors, treatment with antimicrobial agents targeting CA-MRSA can be made. This will minimize the overuse and misuse of broad spectrum antibiotics. The information gained by implementing this proposal will help develop specific infection control measures which should be adopted by hospitals and other healthcare settings in order to prevent the spread of this resistant bacterium. Developing and refining the methodology for this project has significance in applying the technology towards other population based datasets. For example, the Centers for Disease Control and Prevention have established an active ongoing surveillance system of specific harmful pathogens through its Active Bacterial Core Surveillance. These are population based surveillance systems from across the US. Seamless integration of the various information systems can simplify the process of determining which pathogens pose the greatest infectious disease threat as well as provide a way to systematically look for risks associated with infection by these pathogens. Appendix I: Definitions and Variables Study period: January 1, 2002 through December 31, 2006 Definitions used for study: CA-MRSA-Community-associated methicillin resistant Staphylococcus aureus. Staphylococcus aureus isolates which show either resistance to the antibiotic oxacillin alone or resistance to both oxacillin and erythromycin and therefore, are phenotypically characteristic of methicillin resistant Staphylococcus aureus isolates which originated from the community. HA- MRSA –Healthcare-associated methicillin resistant Staphylococcus aureus. Staphylococcus aureus isolates which show resistance to oxacillin, erythromycin, and other anti-staphylococcal antimicrobial agents, i.e., isolate phenotype is not CA-MRSA. Nosocomial MRSA- Isolates identified by cultures obtained greater than 48 hours following hospital admission and so, will include phenotypes consistent with both CA-MRSA and HA-MRSA. MSSA-Methicillin susceptible Staphylococcus aureus. Staphylococcus aureus isolates which are susceptible to oxacillin. Laboratory variables for each patient identified to have a laboratory culture positive for either MSSA or MRSA 1. Unique medical record number for each patient identified to have a positive culture 2. Hospital location of patient identified to have a positive culture 3. Unique isolate specimen number corresponding to positive culture 4. Body Site or specimen source for positive culture 5. Date positive culture collected 6. Date collected culture turned positive 7. Antibiogram for staphylococcal isolate 8. Any previous cultures positive for MSSA or MRSA Medical records variables for each patient identified to have a laboratory culture positive for either MSSA or MRSA: 1. Unique medical record number for each patient identified to have a positive culture 2. Date of birth for each patient identified to have a positive culture 3. Date of Admission corresponding to date of positive culture 4. Date of Discharge corresponding to date of positive culture 5. All other previous dates of admission and discharges corresponding to patient identified to have a positive culture 6. ICD diagnoses corresponding to date of positive culture 7. Patient Address and Zip code at the time of positive culture 8. Race/ethnicity 9. Payer status Emergency service variables for Scottish Rite and Egleston Emergency locations 1. Overall number of visits for emergency services for soft and skin tissue infections will be determined based on ICD codes 2. Overall number of visits for emergency services per year 3. Number of incision and drainage of abscesses done per year based on procedure codes ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download