Venous leg ulcers are a significant health problem. Very few innovative therapies have been proven to be superior to standard care, which at best can be used to successfully heal less than 60% of patients within 20 weeks of initiating care. While standard care has evolved over the past several years, it is not substantially different than care described by Unna more than 100 years ago. The goal of this dissertation was to evaluate standard care, in order to better understand the biology of chronic wound healing, and improve the effectiveness of this treatment by determining those that are most likely to benefit from this treatment. In that way, future trials of new treatments can focus on individuals who are less likely to heal using traditional care. These dissertation investigations applied two basic epidemiologic techniques called explanatory and prognostic modeling. In addition, methodologic studies using sensitivity analyses and Monte Carlo simulations were conducted to demonstrate the validity of these techniques when applied to this clinical situation. In brief, patients with wounds that are of longer duration (greater than 6 months) and of larger size (more than 5 cm2) were found less likely to do well with standard care. Using two sensitivity analyses, it was shown that the effect of these two wound characteristics is unlikely due to an unmeasured confounder. These two commonly measured clinical wound characteristics can also be used to create a very simple but very effective prognostic model, with an area under the receiver-operating curve (AROC) of 0.87 (95% CI: 0.82, 0.91). Finally, since a bootstrap method was used to calculate the 95% confidence intervals for the AROC and was used to compare the AROCs of different models, Monte Carlo simulations were used to validate and compare it to more commonly used methods. These simulations showed that this method is valid and may be superior to other methods. The results of this dissertation should be of interest to health care providers who treat venous leg ulcers, clinical scientists planning efficacy trials for new therapies for a venous leg ulcer, and epidemiologists interested in explanatory and prognostic modeling.
Supervisor: Brian L. Strom. Thesis (Ph.D. in Epidemiology and Biostatistics) -- University of Pennsylvania, 2000. Includes bibliographical references.