Authorities nationally remain concerned about the high rate of adverse events among hospitalized patients. The famous Institute Of Medicine (IOM) report "To Err is Human: Building a Safer Health System" estimated that there are 44,000- 98,000 annual deaths due to medical errors in the U.S. Today, an increasing adoption of health Information Technology (IT) and event reporting system by U.S. hospitals present a unique opportunity to study the medical errors and their consequences related to patient safety and the hospital costs.

The objectives of this dissertation were to determine the adverse event rates among surgical hospitalizations, explore whether high Body Mass Index (BMI) was associated with higher likelihood of an adverse event and, finally, estimate the hospital costs of a Reported Adverse Event (RAE) among Medicare surgical hospitalizations.

Secondary data related to the adverse events was obtained from a Midwestern medical center's event reporting system while the demographic and clinical information for all the surgical hospitalizations from the four-year (2006-09) study period was extracted from the information warehouse of the same Midwestern medical center. All the data were related to three hospitals of the aforementioned Midwestern medical center. In order to determine relevant associations between the outcome variables and the independent demographic and clinical variables, the first study used binary logistic regression analyses; the second study employed conditional logistic regression; and the third study used multivariate linear regression.

Study one results showed that the average RAE rate was 8.3 per 100 surgical hospitalizations with a high variation across various surgical groups (i.e. RAE rate ranged from 55.7 per 100 surgical hospitalizations to 0.0 per 100 surgical hospitalizations). More than half of the RAEs were related to care management, medications, and invasive procedures. The regression analyses found patient's age and the number of RAEs per hospitalization episode as significant predictors of high severity RAE after adjusting for Charlson's comorbidity index score, year of admission, and surgical groups.

Results of the second study showed a statistically significant negative effect of overweight BMI status of the patient on the likelihood of an RAE. Similarly, the study found no statistically significant relationship between obese, morbidly obese, or underweight BMI and the likelihood of an RAE occurrence.

The third study estimated the average adjusted cost of an RAE to be $4,507 (95% CI: $3,917-$5,186). The top three surgical groups with high total cost of RAEs, during the four-year (2006-09) study period, were operations on valves and septa of heart ($1.43 million), operations on heart and pericardium ($1.14 million), and incision, excision and anastomosis of intestine ($0.90 million). The medication RAEs had the highest average adjusted cost of $5,013, followed by care management related RAEs ($4,970).

In conclusion, the event reporting system data used in combination with patient medical records can help improve quality of care and cost reductions by early detection of adverse events followed by workflow and care coordination improvements.