TY - JOUR
T1 - Predictors of Hospital-Acquired Conditions Are Predominately Similar for Spine Surgery and Other Common Elective Surgical Procedures, With Some Key Exceptions
AU - Horn, Samantha R.
AU - Pierce, Katherine E.
AU - Oh, Cheongeun
AU - Segreto, Frank A.
AU - Egers, Max
AU - Bortz, Cole
AU - Vasquez-Montes, Dennis
AU - Lafage, Renaud
AU - Lafage, Virginie
AU - Vira, Shaleen
AU - Steinmetz, Leah
AU - Ge, David H.
AU - Buza, John A.
AU - Moon, John
AU - Diebo, Bassel G.
AU - Alas, Haddy
AU - Brown, Avery E.
AU - Shepard, Nicholas A.
AU - Hassanzadeh, Hamid
AU - Passias, Peter G.
N1 - Publisher Copyright:
© The Author(s) 2019.
PY - 2019/10/1
Y1 - 2019/10/1
N2 - Study Design: Retrospective review of a prospectively collected database. Objective: To predict the occurrence of hospital-acquired conditions (HACs) 30-days postoperatively and to compare predictors of HACs for spine surgery with other common elective surgeries. Methods: Patients ≥18 years undergoing elective spine surgery were identified in the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database from 2005 to 2013. Outcome measures included any HACs: superficial or deep surgical site infection (SSI), venous thromboembolism (VTE), urinary tract infection (UTI). Spine surgery patients were compared with those undergoing other common procedures. Random forest followed by multivariable regression analysis was used to determine risk factors for the occurrence of HACs. Results: A total of 90 551 elective spine surgery patients, of whom 3021 (3.3%) developed at least 1 HAC, 1.4% SSI, 1.3% UTI, and 0.8% VTE. The occurrence of HACs for spine patients was predicted with high accuracy (area under the curve [AUC] 77.7%) with the following variables: female sex, baseline functional status, hypertension, history of transient ischemic attack (TIA), quadriplegia, steroid use, preoperative bleeding disorders, American Society of Anesthesiologists (ASA) class, operating room duration, operative time, and level of residency supervision. Functional status and hypertension were HAC predictors for total knee arthroplasty (TKA), bariatric, and cardiothoracic patients. ASA class and operative time were predictors for most surgery cohorts. History of TIA, preoperative bleeding disorders, and steroid use were less predictive for most other common surgical cohorts. Conclusions: Occurrence of HACs after spine surgery can be predicted with demographic, clinical, and surgical factors. Predictors for HACs in surgical spine patients, also common across other surgical groups, include functional status, hypertension, and operative time. Understanding the baseline patient risks for HACs will allow surgeons to become more effective in their patient selection for surgery.
AB - Study Design: Retrospective review of a prospectively collected database. Objective: To predict the occurrence of hospital-acquired conditions (HACs) 30-days postoperatively and to compare predictors of HACs for spine surgery with other common elective surgeries. Methods: Patients ≥18 years undergoing elective spine surgery were identified in the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database from 2005 to 2013. Outcome measures included any HACs: superficial or deep surgical site infection (SSI), venous thromboembolism (VTE), urinary tract infection (UTI). Spine surgery patients were compared with those undergoing other common procedures. Random forest followed by multivariable regression analysis was used to determine risk factors for the occurrence of HACs. Results: A total of 90 551 elective spine surgery patients, of whom 3021 (3.3%) developed at least 1 HAC, 1.4% SSI, 1.3% UTI, and 0.8% VTE. The occurrence of HACs for spine patients was predicted with high accuracy (area under the curve [AUC] 77.7%) with the following variables: female sex, baseline functional status, hypertension, history of transient ischemic attack (TIA), quadriplegia, steroid use, preoperative bleeding disorders, American Society of Anesthesiologists (ASA) class, operating room duration, operative time, and level of residency supervision. Functional status and hypertension were HAC predictors for total knee arthroplasty (TKA), bariatric, and cardiothoracic patients. ASA class and operative time were predictors for most surgery cohorts. History of TIA, preoperative bleeding disorders, and steroid use were less predictive for most other common surgical cohorts. Conclusions: Occurrence of HACs after spine surgery can be predicted with demographic, clinical, and surgical factors. Predictors for HACs in surgical spine patients, also common across other surgical groups, include functional status, hypertension, and operative time. Understanding the baseline patient risks for HACs will allow surgeons to become more effective in their patient selection for surgery.
KW - complications
KW - hospital-acquired condition (HAC)
KW - spine surgery
UR - http://www.scopus.com/inward/record.url?scp=85073012162&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85073012162&partnerID=8YFLogxK
U2 - 10.1177/2192568219826083
DO - 10.1177/2192568219826083
M3 - Article
C2 - 31552152
AN - SCOPUS:85073012162
SN - 2192-5682
VL - 9
SP - 717
EP - 723
JO - Global Spine Journal
JF - Global Spine Journal
IS - 7
ER -