Send to

Choose Destination
Ann Thorac Surg. 2019 Mar;107(3):897-902. doi: 10.1016/j.athoracsur.2018.07.059. Epub 2018 Sep 22.

Penetration, Completeness, and Representativeness of The Society of Thoracic Surgeons General Thoracic Surgery Database for Lobectomy.

Author information

Division of Thoracic and Cardiovascular Surgery, Duke University Medical Center, Durham, North Carolina. Electronic address:
Duke Clinical Research Institute, Durham, North Carolina.
Division of Cardiothoracic Surgery, University of California, San Diego, La Jolla, California.
Department of Surgery, Yale University School of Medicine, New Haven, Connecticut.
The Society of Thoracic Surgeons Research Center, Chicago, Illinois.
Baptist MD Anderson Cancer Center, Jacksonville, Florida.
Starr-Wood Cardiothoracic Group, Portland, Oregon.
Division of Thoracic Surgery, Massachusetts General Hospital, Boston, Massachusetts.
Johns Hopkins All Children's Hospital, St. Petersburg, Florida.
Department of Surgery, Emory University School of Medicine, Atlanta, Georgia.



Not all surgeons performing lobectomy in the United States report outcomes to The Society of Thoracic Surgeons General Thoracic Surgery Database (STS GTSD). We examined penetration, completeness, and representativeness of the STS GTSD for lobectomy in the Centers for Medicare and Medicaid Services (CMS) patient population.


The STS GTSD lobectomies from 2002 to 2013 were linked and matched to CMS data using a deterministic matching algorithm. Penetration at center- and patient-level were determined by the number of CMS lobectomy sites and patients, matched to STS GTSD data, divided by the total number of CMS lobectomy sites and patients, respectively. Completeness was defined as the ratio of lobectomies linked to STS GTSD data to the total number of lobectomies. Representativeness was determined by comparing outcomes for patients undergoing lobectomy at matched and unmatched STS GTSD sites.


A total of 9,569 centers were included in the study. Center level penetration steadily increased from 1.2% (10 of 859 sites) in 2002 to 25% (169 of 675 sites) in 2013. Patient-level penetration was highest, 38% (4,177 of 11,018), in 2013. Completeness at GTSD sites varied from 59% to 78% over the study period. Postoperative length of stay was longer for nonparticipants than for STS GTSD surgeons (median 6 versus 5 days, p < 0.001); 30-day mortality was higher for nonparticipants than for STS GTSD participants (3.3% versus 1.6%, p < 0.001).


Participation in the STS GTSD has increased over time, but penetration lags behind that of the other STS National Databases. The STS GTSD participants have superior observed perioperative outcomes for lobectomy compared with nonparticipants. Database participation may reflect high quality care, and ongoing efforts to increase surgeon participation in the STS GTSD should be continued.

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center