An easy-to-use, three-dimensional, computer graphics program is bringing a new level of accuracy, consistency and standardization to the evaluation of burn patients, which should result in more precise treatment plans and better evaluation of new therapies. The program will be presented October 16th at the American College of Surgeons' Clinical Conferene in Chicago.
The software, developed by a team of researchers at the University of Chicago Hospitals' Burn Center, replaces the standard two-dimensional hand-drawn charts of a patient's wounds with a morphable 3-D computer body image. Using a mouse or graphic tablet, instead of pencil on paper, the nurse or physician can adjust the diagram to match the contours of the patient's body, chart the extent and depth of the burn wounds as seen from any angle, compute the percentage of total-body-surface area burned, which facilitates treatment.
"The computer program is more accurate and far more consistent than the standard system for determining burn surface area, especially for moderate burns, where precise information can make the most difference," said team leader Raphael Lee, M.D., Ph.D., professor of surgery and medical director of the Burn Center at the University of Chicago Hospitals.
Accurate assessment is a crucial early step in treatment planning. The size and depth of burn wounds are the most important predictors of clinical outcome. The percentage of body surface area affected is used to calculate the patient's fluid and nutritional needs-which can be enormous for those with severe burns. The initial assessment is also used as a benchmark to monitor a patient's progress and as a research tool to compare effects of different treatments.
But burn centers have long had to rely on the doctor's pencil drawings on paper charts, known as Lund-Browder diagrams, which show a standard male or female body, child or adult, from the front and back. Rough percentages for each body part are listed: for example 13 percent for the entire trunk or back, 9.5 percent for one arm, 7 percent for an adult's head or 11 percent for a child's.
Reliance on these two-dimensional charts results in wide variation in assessment of identical injuries by different professionals. Burns near the sides are less apparent on the charts and are often underestimated, while those right in front can be overemphasized. And a patient's body rarely mirrors the idealized forms on the standard charts.
The computer, using software originally developed for architects, allows the burn team to begin assessment on a frame, assembled from 10,000 tiny triangles, that closely resembles the patient. After keying in sex, height and weight, the physician can manually adjust the resulting image to pull out a bigger abdomen, for example, or shrink the shoulders to match the burned patient's physique.
The burns are then drawn directly onto the rotatable 3-D computer diagram with a resolution of 0.01%. On the display, different colors indicate different wound depths: yellow for superficial, red for deep-partial thickness or brown for full-thickness burns. Then the program computes the percentage of body surface area affected as well as fluid and nutritional requirements.
Users can factor in other injuries or treatments that affect the patient's metabolism, such as smoke inhalation or placement on a ventilator. The computer program automatically adjusts for these variables and then calculates nutritional requirements. The computer is also able to zoom. Body parts can be displayed separately or magnified for accuracy. Wound diagrams can be updated and compared as treatment progresses. Skin grafts, biological dressings and donor sites (which become partial-thickness wounds), can be included in the diagram.
Studies comparing the computer with standard burn assessment found that the computer is much more reliable and consistent than the standard system, particularly for larger burns.
In an initial trial, using a mannequin painted with burn wounds, six nurse or physician observers rated the wounds using Lund-Browder charts and the computer. The computer was significantly more accurate and, for large burns, produced one fifth the amount of variation between observers.
Subsequent tests with real patients in the University of Chicago Burn Unit have confirmed the computer's accuracy as well as the willingness of nurses and physicians to use the program despite the time constraints of the clinical setting. The program requires few keyboard commands. Most functions are controlled by a mouse and simple pull-down menus.
"Accuracy is crucial for treating the patient," said Lee, "but consistency is essential for conducting research and communicating new findings. We hope that by gathering better information at the beginning of treatment, we can improve our ability to evaluate outcomes and perhaps speed the development of new therapies."
The research was sponsored by the Electric Power Research Institute, which will distribute the software to burn centers.
Also involved in developing and testing the software were programmers David Tuch (now in graduate school at M.I.T.), Patrick Jacobsen and Gregory Kicska; burn-unit fellows Mahesh Mankani, M.D. (now in private practice in Washington, D.C.), and William Brownlee, M.D. (now at Cook County Hospital); burn-unit nurses Tina Tinnin, R.N., M.S.N., Alison Boddie, R.N., B.S.N., and Annemarie O'Connor, R.N., B.S.N.