Professor of Surgery
Section of Cardiac & Thoracic Surgery
Head, Thoracic Surgery Service

The University of Chicago
5841 S. Maryland Avenue, MC5040
Chicago, IL 60637
Phone: (773) 702-2500
Fax: (773) 702-2642

Research in Thoracic Surgery

Clinical research in Thoracic Surgery consists primarily of outcomes research, and is performed in collaboration with the University of Chicago Cancer Research Center, Department of Health Studies, Department of Radiology and the Section of Hematology/Oncology.

Mark K. Ferguson, MD, focuses on improving patient selection for major thoracic surgery, analyzing short-term and long-term outcomes after lung resection and esophagectomy, and in managing risk associated with thoracic surgery.  Current protocols and investigations include:

Use of imputation techniques for estimating missing variables in large datasets

Large administrative (and some research) databases are frequently found to be missing values for variables that might be important in the prediction of risk and outcomes.  There are a variety of imputation techniques that can be used to estimate values for these missing variables.  Dr. Ferguson and his colleagues employ the use of classification and regression tree (CART) methods to identify predictive variables, which are used to impute the missing values.  We then fit a logistic regression model for each outcome of interest using CART variables and any covariates that are of interest clinically. Subsequent non-parsimonious regression models are created incorporating all main effects and significant two-way interactions among the covariates.  These techniques describe variables associated with outcomes much more completely than traditional risk modeling methods.

Development of scoring systems for prediction of mortality and pulmonary complications after major lung resection

The thoracic group has developed one of the most accurate scoring systems for prediction of complications after major lung surgery. This system, known as the EVAD system, is based on values for spirometry, diffusing capacity and patient age.  This model is being assessed for its utility in predicting surgical outcomes in larger administrative databases.  It is also being evaluated for its utility in predicting long-term survival after major resection for lung cancer.

Risk modeling in thoracic surgery

The ability to predict risk in thoracic surgery is vital to outcomes research, enabling investigators to appropriately stratify patients by risk as well as by more traditional parameters such as cancer stage, extent of surgery, and age.  The group has recently completed a multi-institutional study of risk of unexpected ICU admission after lung resection.   ICU readmission is an important marker of postoperative mortality and increased cost of hospitalization.  In this study, they identified risk factors for this complication and developed a scoring system that stratifies patients into low, average or high risk for ICU readmission.  This can help surgeons plan for appropriate postoperative ICU admission and may help with decisions about when patients should be transferred from the ICU postoperatively.

Assessment of physician prediction of operative risk

Inherent in risk prediction and the development of scoring systems for estimating risk is the absence of physician judgment in risk assessment.  There is no doubt that physician estimates of risk differ substantially from risk scores.  Whether one is more accurate than another, or whether they would be more accurate when formally combined, is unknown.  They are evaluating physician estimates of risk and comparing them to formal risk scores for major lung resection using clinical vignettes for 100 patients.  Additionally, they are comparing estimates of risk provided by experienced surgeons to those offered by surgeons in training to determine the influence of experience on physician risk estimates.

Evaluation of quality of life and mood in the elderly after major lung resection

The diagnosis of cancer can have a profoundly negative impact on mood and quality of life.  In elderly patients in whom treatment options are somewhat limited, these effects may be more pronounced.  Altered mood can influence recovery after surgery and subsequent quality of life long term.  They are evaluating mood and quality of life using several assessment tools in our lung resection patients to determine the relationship of quality of life and mood to age and to measurable physical factors.  Correlations among these will help identify patients who are at risk for poor quality of life postoperatively.

Correlation of pro-inflammatory and anti-inflammatory gene status and outcomes after major lung resection

Complications after major lung surgery are common and have been associated with activation of inflammatory mediators.  Dr. Ferguson and his colleagues are evaluating alleles of pro-inflammatory and anti-inflammatory genes to determine whether a specific genotype is associated with an increased incidence of complications attributable to inflammatory mediators.  Subsequent work will investigate protein expression by these genes, providing a means for assessing the phenotype of the inflammatory response in this patient group.  Such efforts hopefully will culminate in the ability to identify patients at high risk for serious postoperative complications and intervene with risk reduction measures.

Feasibility of minimally invasive esophagectomy for cancer

Esophagectomy for cancer is among the more high risk operations routinely performed.  Many such operations require two large incisions and extensive dissections in two body cavities to enable removal of the affect organ.  The long operation and extensive surgical trauma predispose patients to complications.  It is thought by some that use of minimally invasive techniques for esophagectomy, which reduce surgical trauma through use of smaller incisions, may decrease the risk of postoperative complications.  The feasibility of this approach has been documented in a few centers with extensive experience using the technique.  This multi-institutional trial is designed to determine feasibility of using this technique among surgeons with interest and skill in esophagectomy.

Utility of diffusing capacity in predicting complications after major lung resection in patients without chronic obstructive pulmonary disease

In the late 1950s, the University of Chicago was among the first to document decreases in diffusing capacity of the lung, a measure of oxygen exchange capacity, in patients undergoing major lung resection.  In the late 1980s we demonstrated that diffusing capacity was, at the time, the single strongest predictor of morbidity and mortality after major lung resection. Current research focuses on the utility of measuring diffusing capacity in patients without emphysema as a means for stratifying risk of major lung resection.  Our preliminary data indicate that risks associated with impaired diffusing capacity are similar in frequency and severity regardless of the degree of underlying emphysema.  This suggests that diffusing capacity should be measured routinely, even in otherwise healthy patients, as a means for risk stratification.