Gathering data on wounds is evolving beyond taking measurements using a ruler. With an eye on developing technologies such as smartphones, remote patient care, and advanced data reporting, the author shows how data analytics may look in the future.
Research has shown that the common protocol for assessing wound size—measuring with a ruler—is often inaccurate and could result in significant deviations when different clinicians measure the same wound.1 The ruler is used to capture length and width measurements and then these values are plugged into a simplistic formula that assumes all wounds are rectangles. In situations when a ruler is unavailable, clinicians eyeball the dimensions of a wound before manual documentation into the electronic medical record (EMR).
Now imagine going to the doctor for a routine checkup, and instead of putting you on a scale, they measure your height and width, plug the values into a formula, and assume the output is a good approximation of your weight. Or even worse, the doctor takes a visual assessment of the patient and approximates the patient’s weight based on “experience.” Such a scenario will be unacceptable given that we have readily available tools that can accurately and objectively measure weight. Likewise, generating accurate and consistent wound data is critical because it forms the basis for evaluating different treatment options and facilitates communication among the clinical team.
At the same period while the wound care community was surviving with subpar methods of measurement, a parallel development in technology was underway: smartphones were becoming ubiquitous and more powerful than ever. A few technologists looked at the situation and thought “there must be a better way.” Soon enough, medtech startups sprung up to tackle this problem of inaccurate data capture in wound care. Using advanced machine learning technologies, tools were developed that allowed for accurate wound area measurements and objective tissue composition analysis.2
The overarching goal was to improve the documentation process and streamline the workflow for clinicians, who have been heavily burdened by clunky EMRs.
What is Data Analysis Like Now?
The potential for digital data entry and storage is dependent on achieving interoperability across different EMRs. Technologies such as the Substitutable Medical Apps, Reusable Technology on Fast Healthcare Interoperability Resources (SMART on FHIR), a software protocol that allows healthcare data to be transmitted across various facilities, have reduced the burden of communication between disparate IT systems. Additionally, advancements in machine learning, data science and mobile technology have led to novel applications in wound care including:
Depth measurements. The latest smartphone cameras enable high-resolution videos that could be leveraged to reconstruct 3D models of wounds and accurately calculate the depth of a wound. In order to utilize this feature, the clinician takes a 5–15 second video of the wound in addition to the standard photo. The software powering this functionality is open-source and constantly improving as independent developers continue to create better algorithms. A practical application of a 3D wound model would be to estimate the amount of skin substitutes for a patient.
Advanced reporting and analytics. A secure central repository of accurate and timely patient data enables the creation of a mobile command center. Facility coordinators will be able to monitor and track the change in relevant operating metrics over time via custom dashboards. This could take the form of monitoring the etiologies of wounds for a particular facility or showing the location of ulcer hot spots for different sections of the body. It would also help the procurement team tailor formulary composition to the relevant patient demographic. Visualizations that make such analysis easily accessible will ultimately facilitate the efficient use of resources and better inventory management.
Clinical decision support. A mobile platform enables clinical guidelines to be readily available to clinicians who do not have access to a wound care expert.3 This is normally the case in rural or understaffed facilities. Given the constantly changing nature of best practices, it can be difficult to keep track of the most recent guidelines. An accessible service that places the most recent information at a clinician’s fingertips facilitates an efficient knowledge transfer and ultimately, better wound outcomes.
Wound care EMR module. Area measurement is not the only relevant factor in wound assessment. Point-of-care devices that measure variables related to bacterial infection and perfusion can now be integrated via application programming interfaces (API) to a mobile application and leapfrog the hassle of independent connections to the hospital’s EMR. Without such integration, each new device will have to negotiate a custom connection individually with a facility’s IT department, a laborious and time-consuming task. The holy grail of technology-enhanced wound assessment would be a single device that can perform all these functions seamlessly.
Electronic data capture. It is a common issue for clinical trials to have at least a two-week lag between data collection and analysis. By employing an electronic means of capturing data, real-time analysis becomes possible, which increases the clinical research coordinator’s ability to respond to adverse events. Patients can also give consent remotely and be randomized into the treatment or control arm automatically. The transition from paper forms to electronic records results in data management efficiencies and potentially mitigates compliance and loss to follow up issues.
Remote patient monitoring. Although this feature existed, it lacked traction and adoption until the COVID-19 pandemic. In a matter of days, the ability to practice telemedicine optimized for wound care patients went from a “want” to a “need.” The way it works is that once a patient is added to a facility, the patient can take photos and measurements from anywhere in the world and the photos are sent through a secure cloud portal to their provider. The provider can send reminders and alerts to patients to ensure timely submission of photos. Two-way communication is possible in real time and this potentially reduces the number of no-shows, eliminates commuting costs and drastically reduces the chances of hospital-acquired infections.
Exploratory data analysis. It is not uncommon for wound care product manufacturers to have data from previous trials sitting in different repositories across the organization. With a well-defined data structure made possible by digital health technologies, it is now easier to leverage independent studies to construct a meta-study analysis on clinical trial versus real world effectiveness via cohort matching. These virtual trials can inform post-market analysis, clinical trial design and new product development.
Health economics and outcomes research (HEOR). Early results have shown a significant reduction in hospital-acquired pressure injuries for facilities that engage in photo documentation of wounds. This is important because hospitals in the worst quartile pay a penalty fee to the Centers for Medicare and Medicaid Services (CMS). By maintaining a strict protocol around patient assessment upon admission and documenting changes in skin integrity, clinicians are better equipped to address developing pressure injuries before the condition gets worse.
What Will Data Analysis Look Like in the Future?
Despite the advancements that have been made so far, the future progress of data analytics in wound care is reliant on further simplifying documentation tasks and making insights more accessible for clinicians. Some of the features on the horizon include:
Predictive analytics. Forecasting the time-to-heal for chronic wounds has been the holy grail of wound care. We are just now getting to the point where patient and wound profiles are expansive enough to create reasonable time-to-heal predictions. This will affect wound care by allowing for the efficient management of patients across the cost–time spectrum; balancing between providing affordable and effective care.
Voice dictation. AI assistants like Siri on the iPhone demonstrate how far we’ve come in voice technology. Suffice it to say that this could be leveraged in a wound care setting by allowing clinicians to document wound visits hands-free. This ameliorates the laborious task of inputting unstructured information into the EHR.
Formulary barcode scanning. A major motivation for documentation in healthcare revolves around reimbursement. An ability to accurately track what products are being used in treating patients would increase real world efficacy studies. Most smartphone cameras are able to read the barcode or QR code on products. By pairing this technology with a formulary list, clinicians could adopt a kiosk-like “scan-and-go” workflow in order to easily track the use of advanced wound care products. This further improves health economics and outcomes research by directly tying effective outcomes to products.
Real world data analysis. The digitization of wound care data makes it easier to implement complex calculations involving novel analytical methods. A case in point is the calculation of wound healing rates. Two common and accessible methods involve looking at the change in absolute or percentage area over time. However, these are susceptible to bias depending on the initial wound size. A more robust methodology, the baseline adjusted healing rate (BAHR), has been shown to be more suitable for certain applications and etiologies.4 The BAHR uses wound perimeter as one of its inputs and this has been a major barrier to broader adoption in the past because it is more difficult to capture wound perimeter with a ruler or perform the necessary calculations manually. Digital measurements eliminate that barrier by easily automating the capture of wound perimeter information and the associated complex calculations.
Virtual trials. A common criticism of wound care trials is that they do not represent the real-world patient population. To address this issue, researchers can work with electronic medical records to define patient cohorts from real world data that best represent their target population. Advanced statistical analysis can be used to create a probabilistic understanding of the efficacy of old therapies for new indications and the best targets for new therapies.
Tobe Madu is a Biomedical Engineer focused on data analytics and product strategy at Tissue Analytics (TA). TA develops software solutions that use artificial intelligence to automatically and objectively extract high quality data from input and images. Before joining TA, Tobe was a medical device analyst at GlobalData, a Canadian market research and consultancy firm where he specialized in developing market forecast models and authoring in-depth market research reports on industry dynamics. Tobe received his BS in Biomedical Engineering from the Johns Hopkins University and his MSc in Mechanical Engineering from the University of Toronto.
1. Langemo D, Anderson J, Hanson D, et al. Measuring wound length, width, and area: which technique? Adv Skin Wound Care. 2008; 21(1):42–45.
2. Budman, Joshua, et al. Design of a smartphone application for automated wound measurements for home care. Iproceedings. 2015; 1(1):e16.
3. Jordan S, McSwiggan J, Parker J, Halas GA, Friesen M. An mHealth app for decision-making support in Wound Dressing Selection (WounDS): protocol for a user-centered feasibility study. JMIR Res Protoc. 2018;7(4):e108.
4. Tallman, Phillip, et al. Initial rate of healing predicts complete healing of venous ulcers. Arch Dermatol. 1997; 133(10):1231-1234.