Among the most common complications that occur after surgery are surgical site infections and postoperative sepsis, cardiovascular complications, respiratory complications (including postoperative pneumonia), and thromboembolic complications. Patients who experience postoperative complications have dramatically increased hospital length of stay, hospital costs, and mortality rates. The Centers for Medicare & Medicaid Services, in collaboration with the Centers for Disease Control and Prevention, has implemented the Surgical Infection Prevention Project to decrease the morbidity and mortality associated with postoperative surgical site infections. More recently, the Surgical Care Improvement Project, a national quality partnership of organizations committed to improving the safety of surgical care, has been announced. This review will provide an update from the Surgical Infection Prevention Project and provide an introduction to the Surgical Care Improvement Project.
There are >30 million major operations performed in hospitals each year in the United States [1]. Despite advances in surgical and anesthesia technique and improvements in perioperative care, variations in outcomes for patients having surgery are well known [2–9]. The incidence of postoperative complications ranges from ∼6% for patients undergoing noncardiac surgery [2, 3] to >30% for patients undergoing high-risk surgery [4, 5]. Among the most common complications after surgery are surgical site infections (SSIs) and postoperative sepsis, cardiovascular complications (including myocardial infarction), respiratory complications (including postoperative pneumonia and failure to wean), and thromboembolic complications [2–6, 8, 9].
Patients who experience a postoperative complication have dramatically increased hospital length of stay, hospital costs, and mortality. On average, the length of stay for patients who have a postoperative complication is 3–11 days longer than the length of stay for patients who do not experience complications [2–4, 6–8, 10]. In a recent study of attributable hospital costs associated with surgical complications, Dimick et al. [2] demonstrated that the increased cost was $1398 per patient for infectious complications, $7789 per patient for cardiovascular complications, $52,466 per patient for respiratory complications, and $1810 per patient for thromboembolic complications. Much of the excess length of stay, charges, and mortality due to patient safety events in the hospital can be attributed to postoperative complications [6]. In a recent analysis of data from the Veterans Health Administration (VA) National Surgical Quality Improvement Project (NSQIP), Khuri et al. [9] demonstrated that, independent of preoperative patient risk, the occurrence of a complication 30 days in duration reduced median patient survival by 69%.
The Centers for Medicare & Medicaid Services (CMS) have implemented a number of initiatives designed to improve the quality of inpatient care for people with Medicare [11]. This article will provide an overview of national efforts to improve outcomes for patients having surgery, including an update from the Surgical Infection Prevention (SIP) Project and an overview of the Surgical Care Improvement Project (SCIP).
Background. In 2002, the CMS, in collaboration with the Centers for Disease Control and Prevention, implemented the National SIP Project [12]. The goal of the project is to decrease the morbidity and mortality associated with postoperative SSI by promoting appropriate selection and timing of prophylactic antimicrobials [13]. The project was based on experience the Centers for Disease Control and Prevention gained from implementation of the National Nosocomial Infection Surveillance System and on efforts by the CMS to improve health care quality through its Medicare Quality Improvement Organizations [14–19].
At project initiation, published guidelines [20–26] for surgical antimicrobial prophylaxis were reviewed, and a panel of experts in surgical infection prevention developed 3 performance measures for national surveillance and quality improvement. These measures included: (1) the proportion of patients who have parenteral antimicrobial prophylaxis initiated within 1 h before incision (within 2 h for vancomycin or fluoroquinolones), (2) the proportion of patients who are given a prophylactic antimicrobial regimen consistent with published guidelines, and (3) the proportion of patients whose prophylactic antimicrobial is discontinued within 24 h after surgery end time. Timing and selection of antimicrobial prophylaxis were selected as measures because of an association with reduced SSI incidence. Duration of prophylaxis was selected, because excessive use of antimicrobials promotes bacterial resistance. The CMS also convened the authors of published North American guidelines for surgical antimicrobial prophylaxis, and consensus recommendations for antimicrobial prophylaxis were published as an advisory statement [13].
An assessment of antimicrobial prophylaxis utilization for a national sample of Medicare patients undergoing 5 types of major surgery during 2001 was undertaken by the CMS [27]. On the basis of medical record review of 34,133 Medicare inpatients undergoing cardiac surgery, vascular surgery, general abdominal colorectal surgery, hip and knee total joint arthroplasty, and abdominal and vaginal hysterectomy, we determined that an antimicrobial dose was administered to 55.7% of patients within 1 h before incision. Antimicrobial agents consistent with published guidelines were administered to 92.6% of patients. Antimicrobials were discontinued within 24 h after surgery end time for only 40.7% of the patients [27].
In addition to ongoing local efforts by Medicare Quality Improvement Organizations to promote improvement on the 3 SIP Project measures, a national collaborative that included 56 hospitals and 43 Medicare Quality Improvement Organizations representing 50 states was sponsored by the CMS to test change ideas for prevention of SSIs and to facilitate the spread of improvement methodologies between hospitals across the country [28]. During the 1-year collaborative, participating hospitals showed improvement with respect to the 3 antimicrobial prophylaxis measures, including antibiotic timing (from a mean 72% performance rate at baseline to a mean 92% performance rate at the end of the collaborative), appropriate antimicrobial agent selection (from 90% to 95%), and duration of antibiotic therapy (from 67% to 85%). In addition, participating hospitals improved with respect to other performance measures designed to reduce SSIs, including maintenance of patient normothermia in the operating room (from a mean 57% performance rate at baseline to a mean 74% performance rate at the end of the collaborative), use of supplemental oxygenation (from 75% to 94%), maintenance of euglycemia (from 46% to 54%), and appropriate hair removal (from 59% to 95%). Hospitals participating in the collaborative reported a mean 27% reduction in their SSI rates [28].
Since project implementation, the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) has adopted the 3 SIP Project performance measures as an ORYX core measure set [29], and the 3 measures have been selected by the Hospital Quality Alliance for public reporting of hospital performance on the Hospital Compare Web site [30].
Ongoing national surveillance. The CMS continues to monitor national performance on the 3 SIP Project performance measures. Based on abstraction of a stratified random sample of ∼125 operations involving Medicare patients per state per quarter, antimicrobial therapy was initiated within 60 min before incision in 69.7% of operations, 92.2% of patients received an antimicrobial consistent with guidelines, and 52.9% of patients had antimicrobial therapy discontinued within 24 h after initiation during the fourth quarter of 2004 (figure 1). In addition, using hospital self-collected and voluntarily reported data on the 3 SIP Project measures submitted to QualityNet Exchange [31], benchmarks for performance can be calculated using the Achievable Benchmarks of Care methodology (figure 2) [32].
National surveillance data for antimicrobial prophylaxis for surgery. Data are based on a stratified random sample of select operations (hysterectomy, colorectal surgery, cardiac surgery, vascular surgery, or hip or knee arthroplasty) from each state. Approximately 125 patients are selected from each state for medical record abstraction each quarter (Q). *Includes vancomycin or fluoroquinolones that were initiated within 120 min before surgery.
Benchmarks for antimicrobial prophylaxis performance measures. Average performance and benchmark performance are based on self-reported performance rates from 1487 hospitals that voluntarily submitted data for 129,704 patients to QualityNet Exchange in the second quarter of 2005 [31]. Benchmarks are calculated using the “achievable benchmarks of care” methodology [32]. *Includes vancomycin or fluoroquinolones that were initiated within 120 min before surgery.
Surveillance for the most commonly used antimicrobials for surgical prophylaxis is summarized in table 1. First- and second-generation cephalosporins remain the preferred agents for prophylaxis. Many patients receive >1 antimicrobial at the time of surgery. Vancomycin was used in 28.2% of cardiac operations, 13.3% of vascular operations, and 10.5% of hip or knee arthroplasties in the fourth quarter of 2004. Those patients who received vancomycin for prophylaxis often received additional antibiotics (data not shown). For example, of the 315 patients who underwent cardiac surgery who received vancomycin for prophylaxis in the fourth quarter of 2004, 131 (41.6%) also received cefazolin, 58 (18.4%) also received cefuroxime, 33 (10.5%) also received gentamicin, and 29 (9.2%) also received levofloxacin.
Project updates. All of the SIP Project measures are reviewed quarterly by a technical advisory panel of clinical experts, along with staff of the JCAHO and CMS. Having been endorsed, the measures are also subject to review by the National Quality Forum. Recently, a number of issues were identified that affected hospital performance rates for the antimicrobial selection measure. After careful consideration, the CMS, along with the JCAHO, temporarily suspended public reporting of hospital performance on appropriate antibiotic selection for surgical prophylaxis and did not post this measure on the Hospital Compare Web site [33]. There were 3 reasons for the temporary suspension of antibiotic selection as a publicly reported measure of hospital quality.
1. Increasing prevalence of both health care—associated methicillin-resistant Staphylococcus aureus (MRSA) and community-acquired MRSA. Several published guidelines for surgical prophylaxis recommend use of vancomycin for prophylaxis for some operations performed in hospitals with a “high rate” of infection due to MRSA or methicillin-resistant Staphylococcus epidermidis [22–24]; however, there is no guidance on what constitutes a “high rate” of MRSA infection. Although the Healthcare Infection Control Practices Advisory Committee has previously stated in their guideline “the routine use of vancomycin in antimicrobial prophylaxis is not recommended for any kind of operation” [23, p. 260], some high-risk patients would benefit from use of vancomycin for prophylaxis [34].
2. National shortages of antibiotics recommended for prophylaxis for patients undergoing general abdominal colorectal surgery. AstraZeneca discontinued the production of cefotetan, and no other manufacturers produce this antibiotic [35]. In addition, there were national shortages of cefoxitin because of the inability of the manufacturers to meet demand for the antibiotic. Consequently, cefazolin used in combination with metronidazole was, in many institutions, the only guideline-recommended parenteral alternative for surgical prophylaxis for patients undergoing colorectal surgery.
3. Conflicting antibiotic recommendations for prevention of endocarditis in patients undergoing surgery who have coexisting valvular heart disease. Surgical antimicrobial prophylaxis guidelines do not address the need for additional antibiotics to prevent endocarditis in patients at risk, and the American Heart Association guidelines for prevention of endocarditis do not address surgical antimicrobial prophylaxis [36].
In November 2005, the CMS reconvened the authors of published guidelines for surgical antimicrobial prophylaxis to address the 3 reasons for suspension of the antibiotic selection performance measure. Key updates to the antibiotic selection performance measure are summarized in table 2. For colorectal surgery and hysterectomy, the authors of the guideline recommended adding ampicillin-sulbactam to the list of acceptable antibiotics.
Updated consensus recommendations of the Surgical Infection Prevention Guideline Writers Workgroup.
The guideline authors also discussed the use of vancomycin for surgical prophylaxis. Although the number of SSIs nationally has declined, the proportion of SSIs due to MRSA has increased [37]. There is no published guidance on when vancomycin should be used for prophylaxis, and few studies provide evidence on which to base a recommendation. There was a general consensus that hospitals should not have policies supporting routine use of vancomycin (i.e., that some accounting of patient risk factors for MRSA infection should be entertained before the decision to use vancomycin is made). Nationally, gram-negative organisms account for 20%–30% of SSIs reported to the National Nosocomial Infection Surveillance System for cardiac surgery and total joint arthroplasty, and vancomycin provides no protection against these infections [37]. There are few data indicating that providing prophylaxis with vancomycin alone will result in fewer SSIs (although there may be fewer SSIs due to MRSA), and the number of cases due to certain infections may be increased [38]. Because of the uncertainty about the use of vancomycin for surgical prophylaxis, the performance measure for antibiotic selection has been modified to allow the use of vancomycin as an acceptable antibiotic for patients undergoing cardiac, vascular, or orthopedic surgery. In the absence of a documented β-lactam allergy, physician documentation of the reason for use of vancomycin will be required for the case to pass the performance measure and to allow for national surveillance of indications for use. In addition, the CMS is currently tracking use of vancomycin at the hospital level and has documented wide variation in use among hospitals. As a function of its public health responsibility, the CMS may explore targeted assessments of hospitals with high rates of vancomycin use to determine if they have documented the rationale or created policies regarding use of vancomycin. Although some hospitals use vancomycin for prophylaxis because of high “institutional prevalence” of MRSA infections, use of vancomycin because of institutional prevalence should be based on surveillance of organisms specifically causing SSIs.
There is little evidence that antibiotic prophylaxis prevents endocarditis. Leadership of the SIP Project met with representatives of the American Heart Association, who are currently updating guidelines for endocarditis prophylaxis. If the clinician decides to provide endocarditis prophylaxis for a patient having surgery, a drug that will inhibit growth of Enterococcus species should be used. Because ampicillin-sulbactam has been added as an acceptable antibiotic for patients undergoing colorectal surgery and hysterectomy, no additional modifications to the measure were necessary.
Finally, the Society of Thoracic Surgeons recently published a practice guideline for duration of antibiotic therapy for patients receiving cardiac surgery [39]. The authors concluded that, although there is some evidence that single-dose prophylaxis or 24-h prophylaxis may be as effective as 48-h prophylaxis, few studies have directly compared 24 h of prophylaxis with 48 h of prophylaxis. Because prolonged prophylaxis in cardiac surgery has been associated with an increased risk of infection with drug-resistant organisms, they concluded that prophylaxis should not be administered for >48 h [40]. Effective for cardiac surgery patients discharged on or after 1 January 2006, the antibiotic duration performance measure has been modified to allow for up to 48 h of prophylaxis after the end of surgery.
Background. In April 2003, representatives of the CMS and the Centers for Disease Control and Prevention, together with representatives of the VA, the American College of Surgeons, the American Society of Anesthesiologists, the Agency for Healthcare Research and Quality, the American Hospital Association, and the Institute for Healthcare Improvement, met to consider the possibility of aligning efforts aimed at reducing surgical complications and mortality. That meeting resulted in the development of the SCIP [41]. The SCIP is a national quality partnership of organizations committed to improving the safety of surgical care through the reduction of postoperative complications (table 3). In addition to the 10 organizations that constitute the steering committee, >30 other organizations have committed to be supporting partners for the project [41]. The SCIP steering committee has established a national goal to reduce preventable surgical morbidity and mortality by 25% by 2010.
Improvement in the 3 SIP Project antibiotic measures, the success of the national SIP Project collaborative, and the tremendous improvement in surgical morbidity and mortality with the VA implementation of NSQIP served as important models for the development of the SCIP. NSQIP is a validated, outcome-based, risk-adjusted program for measurement and reporting of surgical care in VA hospitals [42, 43]. Through ongoing collection of patient preoperative, intraoperative, and 30-day outcome data, along with periodic comparative reports that include risk-adjusted ratios of observed to expected outcome rates, VA hospitals have used the information to significantly improve patient outcomes from surgery. Since NSQIP was first piloted in 1991, the 30-day mortality rate after major surgery in the VA has decreased by 31%, and the 30-day morbidity rate has decreased by 45% [42]. On the basis of this experience, the VA has partnered with the American College of Surgeons to roll out a private sector version of NSQIP [43]. After an initial 3-center pilot program in academic medical centers, the Agency for Healthcare Research and Quality funded the Patient Safety in Surgery Study that allowed NSQIP participation by 14 large, non-VA academic medical centers [43, 44]. Although originally entirely based on collection and reporting of risk-adjusted outcomes, both the VA and American College of Surgeons NSQIPs have plans to incorporate the SCIP process of care measures into their data collection tools [43].
SCIP performance measures. The SCIP partnership is focusing measurement of quality on 4 broad areas in which the incidence and cost of complications in surgery is high and there is a significant opportunity for prevention. The proposed processes of care and outcome measures for SCIP are summarized in tables 4 and 5.
Proposed process of care performance measures for the National Surgical Care Improvement Project.
1. Prevention of SSIs. SSIs account for 14%–16% of all hospital-acquired infections and are common complications of surgery, occurring in 2%–5% of patients after clean extra-abdominal operations and in up to 20% of patients undergoing intra-abdominal operations [27]. In addition to the 3 measures of antimicrobial prophylaxis, additional process measures focus on the control of blood glucose postoperatively in cardiac surgery patients [45, 46], proper hair removal [47, 48], and maintenance of normothermia in patients undergoing colorectal surgery [49].
2. Prevention of venous thromboembolism. In the absence of prophylaxis, deep vein thrombosis occurs in ∼25% of all major operations, and pulmonary embolism occurs in ∼7%. More than 50% of major orthopedic operations are complicated by deep vein thrombosis, and up to 30% are complicated by pulmonary embolism without prophylaxis. Despite the publication of evidence-based guidelines [50–52] for venous thromboembolism prophylaxis, multiple medical record audits have demonstrated underuse of prophylaxis [53, 54]. New process measures for the SCIP focus on the ordering and administration of appropriate forms of venous thromboembolism prophylaxis on the basis of the type of operation being performed [52].
3. Prevention of adverse cardiac events. Adverse cardiac events, including acute myocardial infarction, sudden cardiac death, and pulmonary edema, occur in 2%–5% of patients undergoing noncardiac surgery and in as many as 30% of patients undergoing vascular surgery. Perioperative cardiac events are associated with a mortality rate of nearly 60% per event, prolonged hospitalization, and higher costs [55–57]. A new process measure for the SCIP focuses on administration of β-blockers to those patients who have required β-blockers chronically for conditions such as angina, hypertension, or arrhythmias [58].
4. Prevention of respiratory complications. Ventilator-associated pneumonia occurs in 9%–27% of all intubated patients. Up to 90% of health care—associated pneumonia cases that occur in the intensive care unit occur during mechanical ventilation [59]. In addition, mechanically ventilated patients are at increased risk of stress ulcer disease and gastrointestinal bleeding. Process measures that address elevation of the head of the bed, provision of stress ulcer disease prophylaxis, and use of ventilator weaning protocols to reduce the duration of mechanical ventilation have been proposed. However, implementation of these SCIP measures has been deferred pending National Quality Forum evaluation of the strength of evidence and because the JCAHO has developed a similar set of performance measures that would apply to a much broader population of intensive care unit patients [60].
In addition to the process of care measures, there is strong commitment by the SCIP steering committee to measure outcomes of care (table 5). Although the NSQIP provides an outstanding model for measuring surgical outcomes, there is recognition by the steering committee that many hospitals do not currently have the infrastructure, such as a dedicated clinical nurse reviewer, in place to capture the necessary clinical data and data regarding 30-day events to appropriately measure risk-adjusted outcomes. Several outcome measures have been proposed, with the recognition that these would only be useful for a hospital's internal quality improvement efforts and would not be suitable for public accountability. Global or operation-specific measures of outcomes, such as 30-day readmission and 30-day mortality rates, will be pursued.
The implementation of SCIP measures will be staged, beginning with process measures for the prevention of infectious and thromboembolic complications and followed by cardiac and respiratory processes of care. Outcome-measure data collection, solely for internal institutional quality improvement, is projected to begin in 2007.
There is a strong national commitment to measure processes and outcomes of care for surgery in the United States. Through a broad national partnership, hospitals across the nation will be encouraged to participate in activities to reduce the complications of surgery.
We would like to thank the following participants in the November 2005 meeting of the Surgical Infection Prevention Guideline Writers Workgroup meeting: Dr. Laura Prokuski (University of Wisconsin, Madison; American Academy of Orthopaedic Surgery); Dr. Vanessa Dalton (University of Michigan, Ann Arbor; American College of Obstetricians and Gynecologists); Dr. Christopher Daly (Duquesne University, Pittsburgh, PA; American College of Surgeons); Dr. Matthew Levison (Drexel University, Philadelphia, PA; American Heart Association); Dr. Neil Hyman (University of Vermont, South Burlington; American Society for Colon and Rectal Surgeons); Dr. Keith M. Olsen (University of Nebraska Medical Center, Omaha; American Society of Health-System Pharmacists); Audrey Adams (Montefiore Medical Center, Bronx, NY; Association for Professionals in Infection Control and Epidemiology); Dr. John Jernigan (Centers for Disease Control and Prevention, Atlanta, GA); Lisa Lang, Dr. Sheila Roman, and Dr. David Hunt (Centers for Medicare & Medicaid Services, Baltimore, MD); Dr. E. Patchen Dellinger (University of Washington, Seattle; Healthcare Infection Control Practices Advisory Committee and Infectious Diseases Society of America); Dr. Paul Auwaerter (Johns Hopkins University, Baltimore, MD; Johns Hopkins Antibiotic Guide); Sharon Sprenger (Oakbrook Terrace, IL; Joint Commission on Accreditation of Healthcare Organizations); Dale W. Bratzler, Karina Carr, Wanda Johnson, and Joanie McPhetridge (Oklahoma Foundation for Medical Quality, Oklahoma City; National Surgical Infection Prevention Quality Improvement Organization Support Center); Dr. Steven Gordon (Cleveland Clinic Foundation, Cleveland, OH; Society for Healthcare Epidemiology of America); Dr. Richard Engelman (Baystate Medical Center, Springfield, MA; Society of Thoracic Surgeons); Dr. Gianna Zuccotti (Weill Medical College of Cornell University, New York, NY; The Medical Letter).
In addition, we would like to thank Dr. Allen Ma (Oklahoma Foundation for Medical Quality, Oklahoma City) for his statistical analyses.
Financial support. Contract number 500-02-OK-03, funded by the Centers for Medicare & Medicaid Services, an agency of the U.S. Department of Health and Human Services (SCIPARTC-001-QIOSC-OK-0106).
Potential conflicts of interest. D.W.B. and D.R.H.: no conflicts.
The content of this publication does not necessarily reflect the views of policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government. The authors assume full responsibility for the accuracy and completeness of the ideas presented.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy