Surgical Outcomes

The Surgical Outcomes Program is a collaboration of surgeons, anesthesiologists, scientists and allied health staff whose goal is to improve the quality of surgical patient care at Mayo Clinic and in the United States as a whole.

Invasive surgical procedures are complex and may lead to complications and adverse events, which affect patients and contribute significantly to overall health care costs.

Using Mayo Clinic institutional data and multiple sources of multicenter and national data, the Surgical Outcomes Program focuses on building evidence to inform the practice of surgery and improve its safety, efficiency and quality as well as the experience of surgical patients.

Example projects

Research leads to guidelines that optimize opioid prescriptions

In the last 15 years, the number of Americans receiving opioid prescriptions and the number of deaths involving overdoses have roughly quadrupled, according to the Centers for Disease Control and Prevention. More than 41 people a day died from prescription opioid overdoses in 2015.

Researchers in the Surgical Outcomes Program looked at prescribing practices from January 2013 to December 2015 for 25 common surgeries at Mayo Clinic campuses in Arizona, Florida and Minnesota. In particular, the researchers examined patients who weren't taking opioids in the 90 days before surgery. Within that group of 5,756 patients, they found that 4 in 5 patients received more opioids than initially recommended by Minnesota state guidelines.

Based on these data, the Mayo Clinic Department of Orthopedic Surgery has transformed its prescribing practices for patients who weren't taking opioids in the 90 days before surgery. The department is developing recommended levels based on surgical procedure and patient need. Many other departments have followed suit with their own guidelines.

Returned to the operating room — Complex patient or complications?

One commonly cited measurement of surgical quality is how often patients are returned to the operating room (ROR) for further surgery soon after their initial surgeries. Some research has suggested that as many as 70% of ROR events are related to surgical complications — issues arising because of the previous surgery.

Researchers used a real-time electronic tool to identify all ROR cases at Mayo Clinic's campus in Rochester, Minnesota, between May 2014 and January 2015. Analysis of the 43,607 operations during this period showed 13% ROR within 45 days of initial operation. Of these, the majority were planned returns (7%) and unrelated returns (4%). Planned returns due to complications were 0.8% and unplanned operations were 2%.

Although post-surgical complications can be linked to quality of care, the researchers believe that a simple ROR rate is not a valid surgical quality indicator for large medical centers, which see the most complex surgical and trauma patients. They recommend use of electronic tools designed specifically to identify ROR and the related reason in order to provide validated real-time ROR data for public reporting and to drive quality improvement efforts.

A simple blood test could lead to longer life for patients with pancreatic cancer

Pancreatic cancer is often fatal. In fact, it's the fourth-leading cause of death in men and women. A simple blood test to determine whether the level of carbohydrate antigen (CA) 19-9 is elevated in patients who have been newly diagnosed with pancreatic cancer can have a profound effect on survival — yet many institutions are not using the test.

Mayo Clinic's research investigated the prevalence of test use across the U.S. and determined the real value of the test for patients. Using the National Cancer Data Base, researchers analyzed outcomes for more than 113,000 patients and showed that less than 25 percent of patients receive the CA 19-9 test. Concurrently, the researchers showed that patients with elevated CA 19-9 levels tended to have worse outcomes than others at the same stage of cancer. They also saw that the elevated tumor marker's negative effect on survival was most pronounced in patients diagnosed at an early stage.

The team then looked at when chemotherapy was given to patients and determined that for patients with elevated levels of the CA 19-9 tumor marker, the best outcomes — that is, longest survival rates — occurred if chemotherapy was given before the surgery. The outcomes of patients with elevated and nonelevated levels then were the same.

While the CA 19-9 test for patients with pancreatic cancer has been a standard part of care at Mayo Clinic for years, the researchers noted that the test isn't widely used across the country. More widespread use of the simple test and understanding the mitigating effect of pre-surgical chemotherapy for these patients could save many lives.

Read the study abstract on PubMed.

Colorectal surgical site infection predictive models: Do they add value or just confusion?

Colorectal surgical site infections (SSIs) are modeled and scrutinized, and there is increasing movement in the direction of public reporting. However, external validation of colorectal SSI risk prediction models is lacking, and factors governing colorectal SSI occurrence are complicated and multifactorial. Mayo Clinic researchers hypothesized that existing colorectal SSI prediction models have limited ability to accurately predict colorectal SSI in independent data.

In the researchers' independent institutional data set, published colorectal SSI risk prediction models do not perform well. The researchers found that application of externally developed prediction models to any individual practice must be validated or modified to account for institution specific and case-mix factors. This finding questions the validity of using externally or nationally developed models for "expected" outcomes in individual institutions as well as interhospital comparisons.

Policymakers should be aware of the limited applications of national and institutional SSI models to any particular institution's expected surgical outcomes. The researchers recommend validating those models on Mayo Clinic institutional data before applying predictive models to patient care.

Read the study abstract.

Predictors of surgical site infection after craniotomy for intracranial neoplasms

Patients with intracranial neoplasms are at higher risk of surgical site infections (SSIs) after craniotomies than are patients receiving surgery for other reasons. Mayo Clinic sought to identify patient risk factors associated with the development of SSI and to understand the impact of SSI on resource utilization, including rates of return to the operating room and length of stay.

The researchers analyzed prospectively collected data in the National Surgery Quality Improvement Program (NSQIP) database, which gave this study a larger sample size than has been afforded to other studies investigating SSI. The NSQIP data set is also a surgical registry. As a result, its level of clinical detail is superior to that of administrative data sets and allows researchers to achieve a level of resolution that is closer to that seen in single institution studies.

This study identified patient risk factors that may assist clinical decision-making regarding patient risk stratification, timing of surgery or preoperative antibiotic prophylaxis for patients with intracranial neoplasms undergoing craniotomies. Moreover, this study reports the additional morbidity and resource utilization — including increased rates of return to the operating room and length of stay — associated with SSI and highlights the need for continued efforts to reduce SSI.

These findings may assist clinicians in decisions regarding patient risk stratification, informed consent, timing of surgery or preoperative antibiotic prophylaxis. The study also reports significant additional morbidity associated with SSI, including increased rates of return to the operating room and length of stay. It highlights the importance of SSI as a quality metric and should serve as a motivator for clinicians and researchers to continue exploring methods for reducing SSIs.

Read the article on ScienceDirect.

Real-time tissue analysis during surgery prevents re-operations

A Mayo Clinic study showed that frozen section analysis during breast cancer lumpectomies, which ensures surgeons remove all cancerous tissue while patients are still anesthetized, spares patients the need for a repeat lumpectomy within a month in about 96% of cases at Mayo Clinic in Rochester, Minnesota. Mayo's success rate is higher than the approximately 87% rate nationally.

These findings have significant implications for other types of surgery, patient satisfaction and cost of care. Mayo Clinic remains one of the only U.S. medical centers to perform frozen section analysis, which was pioneered at Mayo more than 100 years ago and is used in a variety of surgeries. Elizabeth B. Habermann, Ph.D., scientific director of the Surgical Outcomes Program, was senior author on the study.

Read the article on ScienceDirect.

Additional Mayo Clinic research identified some of the financial impacts that widespread use of this process would have on health care costs, including potential savings as great as $90 million in the U.S.

Read the article in the Journal of Oncology Practice.

Past project highlights

Read about previous projects in the Surgical Outcomes Program.


The Surgical Outcomes Program team has been instrumental in answering a number of clinical questions, building the evidence base in health care delivery and contributing to the body of scientific knowledge.

The link below returns citations from PubMed, a service of the National Library of Medicine. PubMed is composed of references and abstracts from Medline, life science journals and online books. It includes publications authored by members of the Surgical Outcomes Program team.

Review publications from the Surgical Outcomes Program in the Mayo Clinic Robert D. and Patricia E. Kern Center for the Science of Health Care Delivery.


Elizabeth B. Habermann, Ph.D.