This article is more than 5 years old. Information may no longer be current.
BLOG: Parkland Ortho app offers PE risk assessment for trauma patients
Many readers who were at the recent American Academy of Orthopaedic Surgeons Annual Meeting probably recognized the increase in new technology and mobile apps now available for orthopedic surgeons. We certainly did, and are eager to introduce many of the new apps to you in this blog during the next few months.
One app that stood out to us was the Parkland Ortho, which was presented in a poster session. The Parkland Ortho app is a pulmonary embolism (PE) risk assessment tool for trauma patients. The app itself is very simple, and in fact, only has one screen.
The tool allows the user to select seven patient variables that are available immediately in the emergency room: patient age; pulse rate in emergency room; mode of transportation to the hospital (ambulance or helicopter); motorcycle accident as mechanism of injury; obesity (BMI greater than or equal to 30); need for admission to ICU; and areas of body injured (thorax, abdomen, or lower extremity/pelvis). As the variables are entered, the app returns a risk ratio for the development of a pulmonary embolism in real time based on the collected data.
Background
The idea is simple, but the concept is novel and the tool is well-made. We sat down with Sheena R. Black, MD, Paul C. Chin, MD, PhD, and Adam J. Starr, MD, three orthopedic surgeons from the Department of Orthopaedic Surgery at the University of Texas Southwestern Medical Center to learn more about how they created this free and useful tool.
Orrin I. Franko, MD and Matthew DiPaola, MD: How did you become involved in creating an app for orthopedic surgeons?
Sheena R. Black, MD, Paul C. Chin, MD, PhD, and Adam J. Starr, MD: At our institution, we see a significant amount of trauma that includes high-energy poly-trauma as well as low-energy trauma patients. One of the most devastating and sometimes fatal complications we see is pulmonary embolism (PE) despite early emphasis on thromboprophylactic measures. We recognize others have studied venous thromboembolism (VTE) in trauma patients and have developed risk assessment models and scoring systems to predict who will develop a DVT or a PE. However, the models often include variables not known at the time of admission, such as history of a prior DVT, family history of VTE or presence of cancer. In addition, most of the models and scoring systems are complex and difficult to use.
The goal of our project was to develop a model that was quick and easy to use and utilizes information available immediately in the emergency room to predict which patients would develop a PE. We only sought to look at PE because that is what kills patients.
Franko and DiPaola: Please describe the steps taken to collect the data, validate the outcomes, and establish this clinical tool?
Black, Chin and Starr: The data utilized was from the National Trauma Registry of The American College of Surgeons (NTRACS) at our single level I trauma center from 2000 to 2012. Overall, 44,777 patients were identified, but we had to exclude 6,180 patients due to incomplete data. This left 38,597 patients in our final sample. Registry event codes were used to establish a PE event.
A multivariate binary logistic regression model was developed to predict the likelihood of developing a PE during initial hospitalization. Using a split-half cross validation process, the model was developed using a 50%, randomly selected development sub-sample, and then tested for accuracy using the remaining 50% validation sample. Our model was able to discriminate between predicted and actual PE events with a receiver operating characteristic area under the curve of 0.87.
Franko and DiPaola: Was the creation of a smartphone-based clinical tool always your plan?
Black, Chin and Starr: No, our initial plan was to identify those factors that would predict which trauma patients would develop a PE. In the initial phases, we actually had developed a website that would input the variables and generate a risk ratio. However, as the project progressed and with the increasing usage and ubiquity of cell phones as well as popularity of cell phone applications, we thought that a smart phone application was the next step for this project.
Franko and DiPaola: How did you work together with your hospital in this project?
Black, Chin and Starr: This project was initiated and developed within the Department of Orthopaedic Surgery at Parkland Hospital. However, we worked closely with the Department of General Surgery at Parkland Hospital to obtain the trauma registry data from the NTRACS. In addition, we also worked with the University of Texas at San Antonio Epidemiology Department with Dr. Howard. We also worked closely with the administration at Parkland to promote and host the application on the website for our orthopedic surgery department.
Franko and DiPaola: How did you identify developers for actual software coding? How was the project funded?
Black, Chin and Starr: The initial website developer was identified by our statistician. The first author on the paper (Dr. Black) identified the smart phone application developer as a brother of one of the medical students who was doing an orthopedic surgery rotation at Parkland Hospital.
The project was funded through the Suzanne and Aaron A. Hoffman, MD, Orthopaedic Research Fund. We did not obtain any grants.
Franko and DiPaola: What was the duration of this project?
Black, Chin and Starr: This project was started in June 2013 and is still in the process of submission to a journal for publication. The ultimate goal of this project would be to prospectively validate our work, however, given the low incidence of PE in trauma patients, this would necessitate a multicenter study. As time elapses, we can include additional years of data to strengthen the study.
Franko and DiPaola: How has this app changed any practices for you?
Black, Chin and Starr: This app has changed the awareness in both the orthopedic surgery as well as general surgery departments of the heightened risk for PE in specific trauma patients at our institution. The ease of the application combined with the simplicity of the variables, make it a quick and easy step in the care of these patients, and ultimately may end up saving a patient’s life. As the app has not yet been studied prospectively, it has not yet specifically changed any protocols at our hospital. However, we are planning for a prospective study to assess the app’s effectiveness.
Franko and DiPaola: Do you have additional apps in mind for the future?
Black, Chin and Starr: Not at this time. However the simplicity and streamlined process for creating this app, as well as the ease for others to use the findings of our research and translate it into useful information, will make it more likely for us to be interested in combining a smart phone app with future research projects.
Franko and DiPaola: What advice do you have for other surgeons interested in creating an app?
Black, Chin and Starr: If you don’t know the answer or solution to a problem, then ask those around you. If they don’t know, they may know someone who does.
Life-saving tool
We thank the developers of Parkland Ortho for taking the time to share about their experience. But more importantly, we applaud their efforts in developing a useful, life-saving tool that is easy to use and freely available. We encourage all of our readers to download their app for either iPhone or Android. The links can be found at http://www.parklandhospital.com/phhs/orthopaedic-research.aspx.