Patients Response to Early Switch To Oral:Osteomyelitis Study

1.1. Background Information Osteomyelitis is a common disease associated with significant morbidity and high cost (1). The treatment of osteomyelitis can be challenging requiring prolonged administration of antibiotics and extensive surgical procedures. Even when the infection is treated, the relapse rate is as high as 20% (2). When a bone is infected, the local multiplication of bacteria produces a local inflammatory response with presence of neutrophils and macrophages with areas of microthrombi and avascular necrosis. If a significant area of avascular necrosis develops, a segment of the bone without any blood supply can become separated and form a sequestrum. Since infection of sequestrum occurs in most patients with osteomyelitis, it is considered that in addition to antibiotic treatment, the patient requires surgical intervention for removal of necrotic bone.

1.2. Scientific Rationale There is agreement regarding the minimal duration of antibiotic therapy for patients with osteomyelitis. Since an infected bone may take 3 to 4 weeks to re--- vascularize, the duration of therapy should be a minimum of 4 to 6 weeks of antibiotics. Because different organisms can cause osteomyelitis, the initial antibiotic therapy during hospitalization should include broad---spectrum antibiotics to cover the most likely organisms. As part of the initial management, a bone biopsy is regularly performed to identify the particular etiologic agent. Once the organism and its antimicrobial susceptibilities are known, the spectrum of antibiotic therapy is narrowed, and the antimicrobial therapy continues with an antibiotic that is targeted according to the susceptibility of the identified pathogen. Targeted antibiotic therapy in patients with osteomyelitis is usually performed after 3 to 5 days of broad spectrum antibiotics, since this is the time required by the microbiology department to generate antimicrobial susceptibilities after the bone biopsy is performed. In regard to the route of antibiotic administration, the standard approach is to use an intravenous antibiotic.

1.3. Potential Risks A potential risk for the use of an early switch to oral antibiotics in patients with osteomyelitis is that the blood level achieved with oral antibiotics may not be high enough to attain clinical resolution of the infection.

1.4. Potential Benefits There are several potential benefits of using oral antibiotics instead of intravenous antibiotics. First, avoiding a peripherally inserted central line eliminates the risk of line infection and line---associated deep vein thrombosis. An early switch to oral antibiotics may also facilitate early hospital discharge. A shorter hospital stay will decrease the risk of hospital---associated complications such as hospital---acquired infections. Further, the patients' quality of life may be better without a central line. Finally, the total cost of therapy will be significantly reduced with oral therapy.

1. Methods 2.1. Trial design \& setting This will be a prospective, randomized, unblinded clinical trial to define if the clinical outcomes of patients with osteomyelitis treated with the experimental approach of intravenous antibiotics with an early switch to oral therapy is non---inferior to the current standard approach of intravenous antibiotics for the full duration of therapy.

2.8. Sample Size \& Statistical Analysis The null hypothesis for this study will be as follows: H0: πs --- πe ≤ ---Δ Where πs is the proportion of clinical failures in the intravenous therapy only group, πe is the proportion of clinical failures in the intravenous therapy plus early switch to oral therapy group, and Δ is the non---inferiority margin. The alternative hypothesis will be: HA: πs --- πe > ---Δ We expect that there will be a 20% clinical failure rate for the primary outcome in both the intravenous antibiotic therapy group and the intravenous antibiotic therapy plus early switch to oral antibiotic therapy group. The study will be powered at 80% with Δ of 0.1. A total of 396 patients will be needed to obtain a 95% confidence interval for the difference in failure rates between the two groups that has a lower limit above ---Δ. If the lower limit of this 95% confidence interval for the 10 difference in clinical failure rates between the two arms is above -Δ, non---inferiority will be met. Considering that approximately 15% of patients will be lost during study follow---up, a total of 456 patients will be enroll in the trial to obtain the 396 patients necessary for final analyses.

1. Data Quality Management Plan 3.1. Overview of the Clinical and Translational Research Support Center

The University of Louisville Clinical and Translational Research Support Center (CTRSC) will be responsible for data collection, data quality, and data analysis for this project. The CTRSC (http://www.ctrsc.net) is a multi---disciplinary team comprised of professionals in medicine, public health, statistics, and computer science. The team has considerable experience managing and supporting single---site and multi---center clinical research studies. Specifically, members of the CTRSC will be responsible for:

• Study design
• Development of data collection forms
• Development of study manual 11
• Development of electronic Internet base data entry system
• Providing instruction on use of forms and data entry system
• Development of study database
• Tracking subject enrollment
• Overseeing data transmission
• Providing data management
• Handling data validation
• Protecting confidentiality of data
• Performing feasibility evaluation The data quality team leader for this project will be Dr. Robert Kelley with assistance from Dr. Timothy Wiemken and Dr. Paula Peyrani.

The CTRSC has access to the University of Louisville's high---performance computing cluster, which consists of 312 IBM iDatplex nodes each with two Intel Xeon quad--- core processors for 2496 total cores. The cluster is equipped with a variety of statistical and bioinformatics software including SAS, R, Matlab, ClustalW, and Blast, and C, Fortran, Perl, and Python libraries. In addition, the CTRSC has several iMac and IBM---compatible workstations with several data management and analysis packages installed including R, Matlab, SQL Server 2012, SAS, SPSS, Eclipse, Visual Studio .NET 2010, MySQL Server 5.1, Tableau 8.0, and REDCap.

3.2. Purpose of Data Quality Management Plan

The purpose of this data quality management plan is outline the procedures and processes necessary to:

1. Ensure the data collection and data management for the study are conducted in a manner consistent with University of Louisville standards as well as state and federal regulations.
2. Ensure that data collected are accurate and complete when verified against source documents.
3. Provide approaches for early interception and correction of errors in data collection.
4. Identify areas where specific education and training efforts regarding data collection need to be focused.
5. Outline the tools that will be used to monitor and assess data quality.
6. Outline the Data Quality Management Team meeting schedule for this project. 3.3. Data Capture Primary data collection will be performed by a qualified study coordinator(s) who will abstract subject data from the electronic medical record onto a paper case report form. Once the paper case report form is complete, the study coordinator or other designee will enter the data into a secure, web---based clinical data management system.