- Email: [email protected]
Outcome Management in Cardiac Surgery Using the Society of Thoracic Surgeons National Database
Outcome Management in Cardiac Surgery Using the Society of Thoracic Surgeons National Database LINDA S. HALPIN, MSN, RN; BRET E. GALLARDO, MIS, CIS; ALAN M. SPEIR, MD; NIV AD, MD
ABSTRACT Health care reform has helped streamline patient care and reimbursement by encouraging providers to provide the best outcome for the best value. Institutions with cardiac surgery programs need a methodology to monitor and improve outcomes linked to reimbursement. The Society of Thoracic Surgeons National Database (STSND) is a tool for monitoring outcomes and improving care. This article identifies the purpose, goals, and reporting system of the STSND and ways these data can be used for benchmarking, linking outcomes to the effectiveness of treatment, and identifying factors associated with mortality and complications. We explain the methodology used at Inova Heart and Vascular Institute, Falls Church, Virginia, to perform outcome management by using the STSND and address our performance-improvement cycle through discussion of data collection, analysis, and outcome reporting. We focus on the revision of clinical practice and offer examples of how patient outcomes have been improved using this methodology. AORN J 104 (September 2016) 198-205. ª AORN, Inc, 2016. http://dx.doi.org/10.1016/j.aorn.2016.06.017 Key words: Society of Thoracic Surgeons National Database, outcome management, performance improvement, pay for performance.
F
ollowing the work of Ernest Codman,1 Avedis Donabedian created what is known as modern-day outcome management. He coined the term outcome as part of his health care paradigm for quality assessment, which is composed of three components: structure (ie, the physical setting where health care is delivered, staffing ratios, patient volumes, equipment, information technology), process (ie, what is done for patients based on evidence), and outcome (ie, mortality, morbidity).2 Unfortunately, it is difficult to use Donabedian’s model in today’s complex health care environment, in which outcomes vary from institution to institution and from patient to patient, creating endless
treatment and outcome possibilities. Therefore, outcome data based on one patient, or a small group of patients, are insufficient to draw inferences related to the quality of health care.3 Current outcome management has evolved because of the move to consumer-driven health care, technological advances in medical science, and the subsequent rise in health care costs. Programs that attempted to control costs such as capitation (ie, a set fee paid by insurance companies to providers per patient regardless of the severity of the patient’s illness) in the 1980s and today’s pay-for-performance/value-based http://dx.doi.org/10.1016/j.aorn.2016.06.017 ª AORN, Inc, 2016
198 j AORN Journal
www.aornjournal.org
September 2016, Vol. 104, No. 3
purchasing initiative (ie, providers are rewarded for the quality of services provided rather than the quantity of services provided)4 as a result of health care reform have led to the need for comparative databases and data registries. Patient populationebased registries purchased by institutions provide the data to help health care personnel investigate the structure, process, and outcomes of a given patient population; recognize the effectiveness of treatment protocols; identify the risk factors associated with mortality and morbidity; and examine measures of overall riskeadjusted performance. The use of data registries will help health care personnel measure outcomes of care and continuously compare their data against set national benchmarks. If the results do not meet the set benchmark or target, personnel can implement a plan to improve performance before their data are publicly reported or used for reimbursement. Registries also allow the user to drill down into the data and further explore the reasons why performance is not meeting the set target. For example, if the postoperative length of stay is longer than the set benchmark, a registry allows the user to examine variables such as patient risk factors, complications, and ventilation time so that changes can be made in the process affecting the variable and, thus, improve length of stay. Having comparable data available to guide health care personnel in quality improvement (QI) initiatives will be of utmost importance in today’s health care environment as we move to a reimbursement system based on cost and outcome. The Society of Thoracic Surgeons National Database (STSND) is the premier clinical registry for cardiac surgery, and is one example of a registry that helps health care providers with QI and reimbursement efforts. Predetermined performance measures from the STSND are linked to the Medicare database of the Centers for Medicare & Medicaid Services5 so that in the near future, providers will need to meet their cost and outcome goals tied to these measures to receive their full reimbursement. In light of these challenges, it will be important to closely monitor the outcome measures that drive payment.
THE SOCIETY OF THORACIC SURGEONS NATIONAL DATABASE The STSND was established in 1989 to assist health care providers in their quest for QI. The database houses more than 4.7 million surgical records and gathers its information from more than 95% of the 1,100 groups and institutions that perform cardiac surgery in the United States.6 The STSND provides a way to look at the structure, process, and outcome of cardiac surgery patients by providing
www.aornjournal.org
Outcome Management in Cardiac Surgery
a standard format with defined data elements for examining the care of cardiac surgery patients (eg, demographics; risk factors; prior interventions; catheterization laboratory data; operative and other invasive procedure data, such as crossclamp time, internal mammary artery use, cardiopulmonary bypass/pump time, and complications); a tool that can be used to target specific areas of clinical practice improvement by comparison with data from similar hospitals and national benchmarks; an accurate look at practice patterns across facilities; the ability to access collective national data; and the opportunity to participate in national and statewide performance-improvement efforts in cardiac surgery.6 Our hospital also participates in a statewide QI effort through the Virginia Cardiac Surgery Initiative, which uses Society of Thoracic Surgeons (STS) data to improve quality for 18 hospitals in the state of Virginia.7 Involvement in this initiative allows hospitals to improve through use of comparative data on metrics such as incidence of postoperative atrial fibrillation, number of readmissions, blood product use, and ventilation time. Hospital personnel share the process they used to achieve their results, thus offering resources so that other hospitals may improve. Data collected by participating surgeons and institutions are stored locally at each institution and are submitted to the national database through the Duke Clinical Research Institute on a quarterly basis. The Duke Clinical Research Institute compiles, analyzes, and reports all STS data elements submitted for each institution by participant identification number. Data elements are arranged by patient population, including patients who have undergone the following procedures: isolated coronary artery bypass grafting (CABG), isolated aortic valve replacement, isolated mitral valve replacement, mitral valve repair, and each of these valve replacements/repairs in combination with CABG. Risk stratification models are applied for mortality, morbidity (eg, prolonged ventilation, pneumonia, mediastinitis, renal failure), and length of stay and then adjusted for differences in procedure mix. Participants receive their data reports with benchmarks approximately six months after the end of the current quarter. After compiling data for 20 years, the STSND committee determined that single-outcome measures were an insufficient basis for comparisons between cardiac surgery programs because of the low event rate of mortality and postoperative complications. In an effort to provide a more comprehensive measurement of the overall quality of a program, the STS developed the composite score. The composite score integrated 11 National Quality Forum measures of quality for CABG AORN Journal j 199
Halpin et al
September 2016, Vol. 104, No. 3
master of science in nursing degree, serves as the critical link between the database and the clinical care of the patient. The COS ensures the reliability and validity of the STS data by reviewing each record to make sure every data element adheres to the STS definition and is a correct reflection of the documentation found in the patient’s electronic medical record. Clinicians enter data into the local STS-approved vendor database, beginning with the preoperative history through discharge for every patient encounter. The COS extracts missing data from the medical record and enters these data directly into the local database soon after the patient is discharged so that the data are as close to “real time” as possible. The database administrator submits STS data to the Duke Clinical Research Institute on a quarterly basis during the required submission period. Figure 1. An example of the Society of Thoracic Surgeons (STS) coronary artery bypass grafting (CABG) composite quality rating. Adapted with permission from the STS National Database, Chicago, IL; the Duke Clinical Research Institute, Durham, NC; and the Inova Heart and Vascular Institute, Falls Church, VA. (eg, reoperations, renal failure, sternal incision site infections, ventilation for more than 24 hours) grouped into four domains: avoidance of mortality, avoidance of morbidity, use of the internal mammary artery, and medications.8 Composite score ratings or “star ratings” (Figure 1) are reported to database participants yearly for isolated CABG, isolated aortic valve replacement, and combined CABG and aortic valve replacement. Institutions that have voluntarily consented will have their star rating publicly reported by the STS on their web site, in Consumer Reports magazine, and in the Consumer Reports online rankings of hospitals and physicians.9
OUTCOME MANAGEMENT AT THE INOVA HEART AND VASCULAR INSTITUTE Since 1993, the Inova Heart and Vascular Institute (IHVI) in Falls Church, Virginia, has been a participant in the STSND. Our local database includes thousands of records of cardiac surgery patients to access for outcome reporting, QI, and research. The database serves as the basis for our cardiac surgery outcome management program.
Data Collection The IHVI clinical team collects clinical data elements concurrently on every cardiac surgery patient’s episode of care. The clinical outcome specialist (COS), who is an RN with a 200 j AORN Journal
Data Reporting The COS shares outcome and quality data from the STS quarterly reports with surgeons and staff members; these data include mortality and morbidity data that compose the STS composite scores, our own clinical dashboard data, and data to be used for focused QI projects (eg, improving postoperative extubation time). Clinical dashboard data include both process and administrative measures, such as patient volumes, length of stay, infections, intensive care unit (ICU) hours, extubation time, and readmissions. The COS uses the dashboard to keep track of how we are performing on each of these measures in comparison with either STS benchmarks or our own internal targets. For reporting these data, the COS uses a variety of forums, such as nursing staff meetings, nursing “lunch and learn” activities, and a yearly update provided to cardiovascular OR (CVOR) staff members. These data reports are always reviewed with the IHVI multidisciplinary QI team for cardiac surgery. The purpose of this QI team is to revise clinical practice by comparing our data with the STS benchmarks to identify areas in which we can improve and then to implement those improvements using evidence-based guidelines and clinical research findings. Representatives from each cardiac patient care area attend monthly quality meetings and are as follows: Administration: vice president for cardiac services, nursing senior director for cardiac services, medical director of quality and safety, pharmacist, director of quality (ad hoc), and infection control practitioner (ad hoc); Preoperative services: nurse practitioners from the surgeon’s office and nurse practitioners for in-hospital consultation; Cardiac surgery OR: cardiac surgeons, anesthesia professionals; nurse manager, chief of perfusion, and physician assistants;
www.aornjournal.org
September 2016, Vol. 104, No. 3
Outcome Management in Cardiac Surgery
Cardiac surgery ICU: physician intensivist, nurse manager, charge nurse, physician assistants, respiratory therapist (ad hoc), and physical therapist (ad hoc); and Cardiac surgery stepdown unit: nurse manager, nurse navigator, physician assistants, social workers, and case managers. Our QI initiatives are focused on six postoperative complications (ie, reoperations for bleeding, sternal dehiscence, mediastinal/deep sternal infection, superficial incision site infection, stroke, and mortality) and are geared toward both surgeon and caregiver performance. During our QI meetings, the COS provides a status report on these measures and any additional QI indicators chosen for the current year. For example, during the past several years, our quality initiatives have focused on the number of patients fast-tracked to the telemetry unit, the number of patients extubated postoperatively within six hours, the length of patient stay in both the ICU and hospital, the incidence of postoperative atrial fibrillation, glucose control, and the number of patient readmissions to the hospital after discharge. To present these data to the QI committee, the IHVI STSND administrator queries the database for the selected indicators and converts the data into the following appropriate reporting formats: Preformatted STS slides: these come with each data report and compare our hospital with all STS hospitals, providing useful data for benchmarking; Clinical dashboard: trended graphs with targets for each QI indicator; Quality grid: all quality indicators with targets and action plans for current and previous years; Physician profile: reports patient volumes and outcomes for each individual surgeon and allows for performance comparison and peer review; and Outcome brochure: an IHVI outcomes report that is mailed to local/referring cardiology physicians yearly.
Revision of Clinical Practice At the monthly QI committee meeting, the COS presents indicators that fall above or below the STS benchmark. As a result of this discussion and the degree to which improvement is needed, the QI team may choose a new indicator for improvement, such as reducing the incidence of sternal incision infections, renal failure, or readmissions for CABG. After a new indicator is chosen, the COS reviews the research literature to determine best practice and consults with clinical experts. A practice group of clinician volunteers from the larger QI group works with the COS to perform an in-depth study
www.aornjournal.org
Figure 2. The performance-improvement cycle used as a guide to structure our process at Inova Heart and Vascular Institute. QI ¼ quality improvement; COS ¼ clinical outcome specialist. Adapted with permission from the Inova Heart and Vascular Institute, Falls Church, VA. of the problem and develops an action plan that can include additional retrospective data review and statistical analysis. After completion, the practice group makes recommendations to the QI committee, whose primary goal is to improve patient care through changes in clinical practice. These changes can include the development of a new practice guideline, revision of standing order sets, or a clinical pathway/algorithm. To track the effect of a practice change, the QI committee determines a target or benchmark, which may match the STS benchmark, and the new indicator is added to our dashboard and quality grid. Progress toward the target is reported at subsequent QI meetings and refined if necessary. Our complete performance-improvement cycle is depicted in Figure 2.
Examples of Performance Improvement There are several examples of how IHVI has successfully used data from the STSND to change and improve clinical practice and patient outcomes. One example of using ICU data on length of stay, postoperative extubation time, and postoperative length of stay occurred when a nurse manager in the cardiovascular ICU at IHVI developed the Rapid After Bypass Back Into Telemetry (RABBIT) fast-track program. AORN Journal j 201
Halpin et al
September 2016, Vol. 104, No. 3
Figure 3. An example of a dashboard graph of data collected for the Rapid After Bypass Back Into Telemetry (RABBIT) program, representing the percentage of patients who met program criteria over time. Adapted with permission from the Inova Heart and Vascular Institute, Falls Church, VA.
The program was designed to increase patient satisfaction, decrease the time that patients remained intubated after surgery, decrease ICU stay, decrease postoperative length of stay, and lower costs, all while maintaining high-quality care.10 Patients in the RABBIT program were transferred out of the ICU to the telemetry unit after they were hemodynamically stable, on the same day as their surgery rather than the morning after surgery. To accomplish this goal, patients were allowed to awaken early after arrival to the ICU, which facilitated earlier extubation from the ventilator and allowed them to progress to dangling their legs off the side of the bed. Transferring patients out of the ICU to the telemetry unit on the same day as surgery not only improved patient satisfaction but also decreased costs because of shortened ICU hours (ie, an average of 9 hours rather than the usual 24 hours) and postoperative length of stay in the hospital; most RABBIT patients were discharged from the hospital on the afternoon of the third postoperative day instead of the fourth or fifth postoperative day. To accomplish this extensive change in practice, morbidity, mortality, and patient satisfaction were tracked very closely. After the QI team determined that RABBIT patients did as well or better than patients before implementation of the RABBIT program, the practice change was made. Today, the QI committee continues to monitor RABBIT data by coding patients as such in the STSND (Figure 3). Time spent in the 202 j AORN Journal
ICU and time before extubation are monitored quarterly in an effort to increase and understand any fluctuations in the number of patients who qualify for RABBIT. The STSND is also used to make comparisons of performance indicators between RABBIT patients and non-RABBIT patients and understand the differences between the two groups. In another example, our STSND report in 2008 showed us to be above the STS benchmark for gastrointestinal (GI) complications in CABG patients (ie, 5.5% versus the STS benchmark of 1.9%). The COS used our hospital STS database to identify patients with GI events and reviewed the patient medical records. Findings indicated that postoperative ileus and ischemic bowel events had increased, which led to an increase in GI events. The COS presented these data at a QI meeting, and the QI team put together a GI practice group. The group consisted of a surgeon, an intensivist, an ICU physician assistant, an ICU pharmacist, and the COS. Accordingly, the QI team reviewed current best practice literature and formed an action plan. The practice group developed and implemented an algorithm for prevention of GI events in CABG patients (Figure 4). In subsequent years, the incidence of GI events has been below the current STSND benchmark. The STSND continues to play an integral part in QI in our practice of cardiac surgery. Several other performanceimprovement initiatives were based on data from the STSND
www.aornjournal.org
September 2016, Vol. 104, No. 3
Outcome Management in Cardiac Surgery
Figure 4. Inova’s algorithm for the prevention of adverse gastrointestinal events after cardiac surgery. Adapted with permission from the Inova Heart and Vascular Institute, Falls Church, VA.
and changed clinical practice and improved patient outcomes at our facility. These projects resulted in decreased average number of hours that a patient spent in the ICU (from 84 hours to 75 hours), increased number of patients extubated within six hours after surgery, decreased incidence of postoperative atrial fibrillation in CABG patients, and decreased blood product use during cardiac procedures.11
www.aornjournal.org
These improvements led to substantial cost savings for the hospital because they allowed for decreased use of expensive hospital resources and improved processes that previously kept patients in the hospital for longer periods of time.
Application to Perioperative Practice Even though the QI projects previously described occurred in the postoperative setting, many of these changes would not have been possible if the improvement process had not begun with the nursing staff members in the CVOR. Collaboration
AORN Journal j 203
Halpin et al
September 2016, Vol. 104, No. 3
Table 1. Use of the Society of Thoracic Surgeons National Database in Perioperative Nursing STS Data Elements for Isolated CABG1,2 (Selected)
Perioperative Nursing Considerations
Underweight (BMI < 18.5) Albumin < 3
Frailty Increased risk for pressure injury Increased likelihood of poor outcome
Diabetes (diet, oral, insulin, or no control), HgA1C > 6.5
Blood glucose harder to control in OR Higher risk of SSI postoperatively
Obesity (BMI ¼ 30.0þ)
Peripheral arterial disease
Potential difficulty inserting IABP catheter or aortic blood pressure line through atherosclerotic aortic-iliac vessels
Immunosuppressive treatment
Increased potential for SSI
Previous CABG
Potential for increased bleeding Adhesions Potential injury to pre-existing bypass conduits
Cross-clamp time
Importance of having instruments and supplies ready Knowledge of proposed surgery Possible contingencies during surgery to minimize/prevent myocardial injury from anoxia/hypoxia
Preoperative hematocrit (anemia)
Increased likelihood of intraoperative transfusions Higher likelihood of perioperative morbidity
Postoperative deep/mediastinal infection or veinharvest or cannulation site infection
Any infection occurring in the early postoperative period could indicate problems with sterility in the OR
Difficulty positioning Risk of patient fall Increased risk for pressure injury Heparin rebound with prolonged bleeding Additional stress on chest incision site closure
STS ¼ Society of Thoracic Surgeons; CABG ¼ coronary artery bypass grafting; BMI ¼ body mass index; Hg ¼ hemoglobin; SSI ¼ surgical site infection; IABP ¼ intra-aortic balloon pump. References 1. Isolated CAB procedures. The Society of Thoracic Surgeons. http://www.sts.org/sites/default/files/documents/pdf/ndb2010/On_Pump_110-112.pdf. Accessed May 10, 2016. 2. Adult Cardiac Surgery Database. The Society of Thoracic Surgeons. http://www.sts.org/sts-national-database/database-managers/adult-cardiac -surgery-database. Accessed May 10, 2016.
and cooperation are enhanced between the COS and CVOR staff members through a yearly presentation of the clinical outcome data. Discussion during the presentation between the COS and CVOR staff members often generates questions and concerns that lead to process improvement. For example, in a previous year, STS data showed an increase in leg veineharvest site infections above the benchmark. Discussion in the QI committee meeting led us to believe that we had problems with our newer staff members, who were learning to perform endoscopic vein-harvest procedures. To help properly educate our staff members on these procedures, the vendor for our equipment watched and guided the RN first assistants and physician assistants during the procedure. Improvements related to endoscopic vein harvest were made by revising the 204 j AORN Journal
process to keep the vein-harvest tract dry. The nursing and physician assistant staff members involved in the improvement effort wrote a guideline for performing the procedure. After implementation of these improvements, the incidence of vein-harvest site infections disappeared. Similar collaborative efforts have helped our surgical team members reduce the incidence of sternal incision site infections by changing the prepping and draping practices for cardiac procedures. In addition, we have reduced intraoperative use of blood products through implementation of conservation efforts like cell saving and autotransfusion.4 After our data showed that the incidence of pressure ulcers in the ICU was greater in patients with an open chest, CVOR staff
www.aornjournal.org
September 2016, Vol. 104, No. 3
Outcome Management in Cardiac Surgery
members implemented a process change in which open-chest patients were placed on special pressure-reducing beds at the end of the procedure before they were moved to the cardiac ICU. Although the STSND data elements are not specific to CVOR nursing practice, many of the data elements collected before surgery could be used to enhance and improve care in the CVOR. Table 1 lists some of these elements and describes how the information might be used for patient care in the OR. When participating in the STSND, a hospital must provide access to these data elements for staff members because this keeps them engaged in the QI process and helps them provide better patient care. At our facility, access to these STSND elements is available on the computers in our preoperative area in the history and physical information section, which is part of our electronic medical record documentation.
6.
7.
8.
9.
10.
11.
CONCLUSION Using data to improve patient care is a significant step as we continue to transition our health care system to one that focuses on the quantity of services provided to one that streamlines care to achieve the best outcome for the best value.4,11 At our facility, close communication between the COS, staff members, and leadership of the CVOR enables potential or actual problems to be identified, tracked, and corrected as needed. Such cooperation enables CVOR team members to see themselves as active participants in the care of their patients and to identify specific areas of patient care in which they can make a significant difference.
Cardiac Surgery Database. Ann Thorac Surg. 2016;101(1):33-41; discussion 41. Health policy compendium: healthcare associated infections. The Society of Thoracic Surgeons. https://www.sts.org/sites/default/files/ documents/pdf/advocacy/HAI.pdf. Accessed June 6, 2016. Speir AM, Rich JB, Crosby I, Fonner E Jr. Regional collaboration as a model for fostering accountability and transforming health care. Semin Thorac Cardiovasc Surg. 2009;21(1):12-19. Report overview: STS composite quality rating and NQF measures. The Society of Thoracic Surgeons. January 27, 2013. https:// www.sts.org/sites/default/files/documents/pdf/ndb2010/Report_ OV_NQF_44-73.pdf. Accessed June 6, 2016. Doctors & hospitals. Consumer Reports. www.consumerreports .org/cro/health/doctors-and-hospitals/index.htm. Accessed June 6, 2016. Sakallaris BR, Halpin LS, Knapp M, Sheridan MJ. Same-day transfer of patients to the cardiac telemetry unit after surgery: the Rapid After Bypass Back into Telemetry (RABBIT) program. Crit Care Nurse. 2000;20(2):50-55, 59-63, 65-68. LaPar DJ, Crosby IK, Ailawadi G, et al. Blood product conservation is associated with improved outcomes and reduced costs after cardiac surgery. J Thorac Cardiovasc Surg. 2013;145(3): 796-803; discussion 803-804.
Linda S. Halpin, MSN, RN, is a clinical outcome specialist in adult cardiac surgery at Inova Heart and Vascular Institute, Falls Church, VA. Ms Halpin has no declared affiliation that could be perceived as posing a potential conflict of interest in the publication of this article.
Bret E. Gallardo, MIS, CIS, is a data registry admin-
Acknowledgment: The authors thank Patricia C. Seifert, MSN, RN, CNOR, CRNFA(E), FAAN, independent cardiac consultant, for her suggestions and assistance in creating Table 1.
istrator at Inova Heart and Vascular Institute, Falls Church, VA. Mr Gallardo has no declared affiliation that could be perceived as posing a potential conflict of interest in the publication of this article.
References
Alan M. Speir, MD, is the medical director of cardiac surgery at Inova Heart and Vascular Institute, Falls Church, VA. Dr Speir has no declared affiliation that could be perceived as posing a potential conflict of interest in the publication of this article.
1. Codman EA. The product of a hospital. Surg Gynecol Obstet. 1914;18:491-496. 2. Donabedian A. Evaluating the quality of medical care. Milbank Mem Fund Q. 1966;44(3):166-206. 3. Halpin LS, Barnett SD, Beachy J. Cardiac surgery outcomes. Outcomes Manag. 2003;7(4):144-147. 4. Sherman RO. The drive for cost transparency in health care. Am Nurse Today. 2014;9(5):32-33. https://americannursetoday.com/ the-drive-for-cost-transparency-in-health-care [subscription required]. Accessed May 19, 2016. 5. Jacobs JP, Shahian DM, He X, et al. Penetration, completeness, and representativeness of the Society of Thoracic Surgeons Adult
www.aornjournal.org
Niv Ad, MD, is the section chief of adult cardiac surgery, and the director of cardiac surgery research at Inova Heart and Vascular Institute, Falls Church, VA. Dr Ad has no declared affiliation that could be perceived as posing a potential conflict of interest in the publication of this article.
AORN Journal j 205
ABSTRACT Health care reform has helped streamline patient care and reimbursement by encouraging providers to provide the best outcome for the best value. Institutions with cardiac surgery programs need a methodology to monitor and improve outcomes linked to reimbursement. The Society of Thoracic Surgeons National Database (STSND) is a tool for monitoring outcomes and improving care. This article identifies the purpose, goals, and reporting system of the STSND and ways these data can be used for benchmarking, linking outcomes to the effectiveness of treatment, and identifying factors associated with mortality and complications. We explain the methodology used at Inova Heart and Vascular Institute, Falls Church, Virginia, to perform outcome management by using the STSND and address our performance-improvement cycle through discussion of data collection, analysis, and outcome reporting. We focus on the revision of clinical practice and offer examples of how patient outcomes have been improved using this methodology. AORN J 104 (September 2016) 198-205. ª AORN, Inc, 2016. http://dx.doi.org/10.1016/j.aorn.2016.06.017 Key words: Society of Thoracic Surgeons National Database, outcome management, performance improvement, pay for performance.
F
ollowing the work of Ernest Codman,1 Avedis Donabedian created what is known as modern-day outcome management. He coined the term outcome as part of his health care paradigm for quality assessment, which is composed of three components: structure (ie, the physical setting where health care is delivered, staffing ratios, patient volumes, equipment, information technology), process (ie, what is done for patients based on evidence), and outcome (ie, mortality, morbidity).2 Unfortunately, it is difficult to use Donabedian’s model in today’s complex health care environment, in which outcomes vary from institution to institution and from patient to patient, creating endless
treatment and outcome possibilities. Therefore, outcome data based on one patient, or a small group of patients, are insufficient to draw inferences related to the quality of health care.3 Current outcome management has evolved because of the move to consumer-driven health care, technological advances in medical science, and the subsequent rise in health care costs. Programs that attempted to control costs such as capitation (ie, a set fee paid by insurance companies to providers per patient regardless of the severity of the patient’s illness) in the 1980s and today’s pay-for-performance/value-based http://dx.doi.org/10.1016/j.aorn.2016.06.017 ª AORN, Inc, 2016
198 j AORN Journal
www.aornjournal.org
September 2016, Vol. 104, No. 3
purchasing initiative (ie, providers are rewarded for the quality of services provided rather than the quantity of services provided)4 as a result of health care reform have led to the need for comparative databases and data registries. Patient populationebased registries purchased by institutions provide the data to help health care personnel investigate the structure, process, and outcomes of a given patient population; recognize the effectiveness of treatment protocols; identify the risk factors associated with mortality and morbidity; and examine measures of overall riskeadjusted performance. The use of data registries will help health care personnel measure outcomes of care and continuously compare their data against set national benchmarks. If the results do not meet the set benchmark or target, personnel can implement a plan to improve performance before their data are publicly reported or used for reimbursement. Registries also allow the user to drill down into the data and further explore the reasons why performance is not meeting the set target. For example, if the postoperative length of stay is longer than the set benchmark, a registry allows the user to examine variables such as patient risk factors, complications, and ventilation time so that changes can be made in the process affecting the variable and, thus, improve length of stay. Having comparable data available to guide health care personnel in quality improvement (QI) initiatives will be of utmost importance in today’s health care environment as we move to a reimbursement system based on cost and outcome. The Society of Thoracic Surgeons National Database (STSND) is the premier clinical registry for cardiac surgery, and is one example of a registry that helps health care providers with QI and reimbursement efforts. Predetermined performance measures from the STSND are linked to the Medicare database of the Centers for Medicare & Medicaid Services5 so that in the near future, providers will need to meet their cost and outcome goals tied to these measures to receive their full reimbursement. In light of these challenges, it will be important to closely monitor the outcome measures that drive payment.
THE SOCIETY OF THORACIC SURGEONS NATIONAL DATABASE The STSND was established in 1989 to assist health care providers in their quest for QI. The database houses more than 4.7 million surgical records and gathers its information from more than 95% of the 1,100 groups and institutions that perform cardiac surgery in the United States.6 The STSND provides a way to look at the structure, process, and outcome of cardiac surgery patients by providing
www.aornjournal.org
Outcome Management in Cardiac Surgery
a standard format with defined data elements for examining the care of cardiac surgery patients (eg, demographics; risk factors; prior interventions; catheterization laboratory data; operative and other invasive procedure data, such as crossclamp time, internal mammary artery use, cardiopulmonary bypass/pump time, and complications); a tool that can be used to target specific areas of clinical practice improvement by comparison with data from similar hospitals and national benchmarks; an accurate look at practice patterns across facilities; the ability to access collective national data; and the opportunity to participate in national and statewide performance-improvement efforts in cardiac surgery.6 Our hospital also participates in a statewide QI effort through the Virginia Cardiac Surgery Initiative, which uses Society of Thoracic Surgeons (STS) data to improve quality for 18 hospitals in the state of Virginia.7 Involvement in this initiative allows hospitals to improve through use of comparative data on metrics such as incidence of postoperative atrial fibrillation, number of readmissions, blood product use, and ventilation time. Hospital personnel share the process they used to achieve their results, thus offering resources so that other hospitals may improve. Data collected by participating surgeons and institutions are stored locally at each institution and are submitted to the national database through the Duke Clinical Research Institute on a quarterly basis. The Duke Clinical Research Institute compiles, analyzes, and reports all STS data elements submitted for each institution by participant identification number. Data elements are arranged by patient population, including patients who have undergone the following procedures: isolated coronary artery bypass grafting (CABG), isolated aortic valve replacement, isolated mitral valve replacement, mitral valve repair, and each of these valve replacements/repairs in combination with CABG. Risk stratification models are applied for mortality, morbidity (eg, prolonged ventilation, pneumonia, mediastinitis, renal failure), and length of stay and then adjusted for differences in procedure mix. Participants receive their data reports with benchmarks approximately six months after the end of the current quarter. After compiling data for 20 years, the STSND committee determined that single-outcome measures were an insufficient basis for comparisons between cardiac surgery programs because of the low event rate of mortality and postoperative complications. In an effort to provide a more comprehensive measurement of the overall quality of a program, the STS developed the composite score. The composite score integrated 11 National Quality Forum measures of quality for CABG AORN Journal j 199
Halpin et al
September 2016, Vol. 104, No. 3
master of science in nursing degree, serves as the critical link between the database and the clinical care of the patient. The COS ensures the reliability and validity of the STS data by reviewing each record to make sure every data element adheres to the STS definition and is a correct reflection of the documentation found in the patient’s electronic medical record. Clinicians enter data into the local STS-approved vendor database, beginning with the preoperative history through discharge for every patient encounter. The COS extracts missing data from the medical record and enters these data directly into the local database soon after the patient is discharged so that the data are as close to “real time” as possible. The database administrator submits STS data to the Duke Clinical Research Institute on a quarterly basis during the required submission period. Figure 1. An example of the Society of Thoracic Surgeons (STS) coronary artery bypass grafting (CABG) composite quality rating. Adapted with permission from the STS National Database, Chicago, IL; the Duke Clinical Research Institute, Durham, NC; and the Inova Heart and Vascular Institute, Falls Church, VA. (eg, reoperations, renal failure, sternal incision site infections, ventilation for more than 24 hours) grouped into four domains: avoidance of mortality, avoidance of morbidity, use of the internal mammary artery, and medications.8 Composite score ratings or “star ratings” (Figure 1) are reported to database participants yearly for isolated CABG, isolated aortic valve replacement, and combined CABG and aortic valve replacement. Institutions that have voluntarily consented will have their star rating publicly reported by the STS on their web site, in Consumer Reports magazine, and in the Consumer Reports online rankings of hospitals and physicians.9
OUTCOME MANAGEMENT AT THE INOVA HEART AND VASCULAR INSTITUTE Since 1993, the Inova Heart and Vascular Institute (IHVI) in Falls Church, Virginia, has been a participant in the STSND. Our local database includes thousands of records of cardiac surgery patients to access for outcome reporting, QI, and research. The database serves as the basis for our cardiac surgery outcome management program.
Data Collection The IHVI clinical team collects clinical data elements concurrently on every cardiac surgery patient’s episode of care. The clinical outcome specialist (COS), who is an RN with a 200 j AORN Journal
Data Reporting The COS shares outcome and quality data from the STS quarterly reports with surgeons and staff members; these data include mortality and morbidity data that compose the STS composite scores, our own clinical dashboard data, and data to be used for focused QI projects (eg, improving postoperative extubation time). Clinical dashboard data include both process and administrative measures, such as patient volumes, length of stay, infections, intensive care unit (ICU) hours, extubation time, and readmissions. The COS uses the dashboard to keep track of how we are performing on each of these measures in comparison with either STS benchmarks or our own internal targets. For reporting these data, the COS uses a variety of forums, such as nursing staff meetings, nursing “lunch and learn” activities, and a yearly update provided to cardiovascular OR (CVOR) staff members. These data reports are always reviewed with the IHVI multidisciplinary QI team for cardiac surgery. The purpose of this QI team is to revise clinical practice by comparing our data with the STS benchmarks to identify areas in which we can improve and then to implement those improvements using evidence-based guidelines and clinical research findings. Representatives from each cardiac patient care area attend monthly quality meetings and are as follows: Administration: vice president for cardiac services, nursing senior director for cardiac services, medical director of quality and safety, pharmacist, director of quality (ad hoc), and infection control practitioner (ad hoc); Preoperative services: nurse practitioners from the surgeon’s office and nurse practitioners for in-hospital consultation; Cardiac surgery OR: cardiac surgeons, anesthesia professionals; nurse manager, chief of perfusion, and physician assistants;
www.aornjournal.org
September 2016, Vol. 104, No. 3
Outcome Management in Cardiac Surgery
Cardiac surgery ICU: physician intensivist, nurse manager, charge nurse, physician assistants, respiratory therapist (ad hoc), and physical therapist (ad hoc); and Cardiac surgery stepdown unit: nurse manager, nurse navigator, physician assistants, social workers, and case managers. Our QI initiatives are focused on six postoperative complications (ie, reoperations for bleeding, sternal dehiscence, mediastinal/deep sternal infection, superficial incision site infection, stroke, and mortality) and are geared toward both surgeon and caregiver performance. During our QI meetings, the COS provides a status report on these measures and any additional QI indicators chosen for the current year. For example, during the past several years, our quality initiatives have focused on the number of patients fast-tracked to the telemetry unit, the number of patients extubated postoperatively within six hours, the length of patient stay in both the ICU and hospital, the incidence of postoperative atrial fibrillation, glucose control, and the number of patient readmissions to the hospital after discharge. To present these data to the QI committee, the IHVI STSND administrator queries the database for the selected indicators and converts the data into the following appropriate reporting formats: Preformatted STS slides: these come with each data report and compare our hospital with all STS hospitals, providing useful data for benchmarking; Clinical dashboard: trended graphs with targets for each QI indicator; Quality grid: all quality indicators with targets and action plans for current and previous years; Physician profile: reports patient volumes and outcomes for each individual surgeon and allows for performance comparison and peer review; and Outcome brochure: an IHVI outcomes report that is mailed to local/referring cardiology physicians yearly.
Revision of Clinical Practice At the monthly QI committee meeting, the COS presents indicators that fall above or below the STS benchmark. As a result of this discussion and the degree to which improvement is needed, the QI team may choose a new indicator for improvement, such as reducing the incidence of sternal incision infections, renal failure, or readmissions for CABG. After a new indicator is chosen, the COS reviews the research literature to determine best practice and consults with clinical experts. A practice group of clinician volunteers from the larger QI group works with the COS to perform an in-depth study
www.aornjournal.org
Figure 2. The performance-improvement cycle used as a guide to structure our process at Inova Heart and Vascular Institute. QI ¼ quality improvement; COS ¼ clinical outcome specialist. Adapted with permission from the Inova Heart and Vascular Institute, Falls Church, VA. of the problem and develops an action plan that can include additional retrospective data review and statistical analysis. After completion, the practice group makes recommendations to the QI committee, whose primary goal is to improve patient care through changes in clinical practice. These changes can include the development of a new practice guideline, revision of standing order sets, or a clinical pathway/algorithm. To track the effect of a practice change, the QI committee determines a target or benchmark, which may match the STS benchmark, and the new indicator is added to our dashboard and quality grid. Progress toward the target is reported at subsequent QI meetings and refined if necessary. Our complete performance-improvement cycle is depicted in Figure 2.
Examples of Performance Improvement There are several examples of how IHVI has successfully used data from the STSND to change and improve clinical practice and patient outcomes. One example of using ICU data on length of stay, postoperative extubation time, and postoperative length of stay occurred when a nurse manager in the cardiovascular ICU at IHVI developed the Rapid After Bypass Back Into Telemetry (RABBIT) fast-track program. AORN Journal j 201
Halpin et al
September 2016, Vol. 104, No. 3
Figure 3. An example of a dashboard graph of data collected for the Rapid After Bypass Back Into Telemetry (RABBIT) program, representing the percentage of patients who met program criteria over time. Adapted with permission from the Inova Heart and Vascular Institute, Falls Church, VA.
The program was designed to increase patient satisfaction, decrease the time that patients remained intubated after surgery, decrease ICU stay, decrease postoperative length of stay, and lower costs, all while maintaining high-quality care.10 Patients in the RABBIT program were transferred out of the ICU to the telemetry unit after they were hemodynamically stable, on the same day as their surgery rather than the morning after surgery. To accomplish this goal, patients were allowed to awaken early after arrival to the ICU, which facilitated earlier extubation from the ventilator and allowed them to progress to dangling their legs off the side of the bed. Transferring patients out of the ICU to the telemetry unit on the same day as surgery not only improved patient satisfaction but also decreased costs because of shortened ICU hours (ie, an average of 9 hours rather than the usual 24 hours) and postoperative length of stay in the hospital; most RABBIT patients were discharged from the hospital on the afternoon of the third postoperative day instead of the fourth or fifth postoperative day. To accomplish this extensive change in practice, morbidity, mortality, and patient satisfaction were tracked very closely. After the QI team determined that RABBIT patients did as well or better than patients before implementation of the RABBIT program, the practice change was made. Today, the QI committee continues to monitor RABBIT data by coding patients as such in the STSND (Figure 3). Time spent in the 202 j AORN Journal
ICU and time before extubation are monitored quarterly in an effort to increase and understand any fluctuations in the number of patients who qualify for RABBIT. The STSND is also used to make comparisons of performance indicators between RABBIT patients and non-RABBIT patients and understand the differences between the two groups. In another example, our STSND report in 2008 showed us to be above the STS benchmark for gastrointestinal (GI) complications in CABG patients (ie, 5.5% versus the STS benchmark of 1.9%). The COS used our hospital STS database to identify patients with GI events and reviewed the patient medical records. Findings indicated that postoperative ileus and ischemic bowel events had increased, which led to an increase in GI events. The COS presented these data at a QI meeting, and the QI team put together a GI practice group. The group consisted of a surgeon, an intensivist, an ICU physician assistant, an ICU pharmacist, and the COS. Accordingly, the QI team reviewed current best practice literature and formed an action plan. The practice group developed and implemented an algorithm for prevention of GI events in CABG patients (Figure 4). In subsequent years, the incidence of GI events has been below the current STSND benchmark. The STSND continues to play an integral part in QI in our practice of cardiac surgery. Several other performanceimprovement initiatives were based on data from the STSND
www.aornjournal.org
September 2016, Vol. 104, No. 3
Outcome Management in Cardiac Surgery
Figure 4. Inova’s algorithm for the prevention of adverse gastrointestinal events after cardiac surgery. Adapted with permission from the Inova Heart and Vascular Institute, Falls Church, VA.
and changed clinical practice and improved patient outcomes at our facility. These projects resulted in decreased average number of hours that a patient spent in the ICU (from 84 hours to 75 hours), increased number of patients extubated within six hours after surgery, decreased incidence of postoperative atrial fibrillation in CABG patients, and decreased blood product use during cardiac procedures.11
www.aornjournal.org
These improvements led to substantial cost savings for the hospital because they allowed for decreased use of expensive hospital resources and improved processes that previously kept patients in the hospital for longer periods of time.
Application to Perioperative Practice Even though the QI projects previously described occurred in the postoperative setting, many of these changes would not have been possible if the improvement process had not begun with the nursing staff members in the CVOR. Collaboration
AORN Journal j 203
Halpin et al
September 2016, Vol. 104, No. 3
Table 1. Use of the Society of Thoracic Surgeons National Database in Perioperative Nursing STS Data Elements for Isolated CABG1,2 (Selected)
Perioperative Nursing Considerations
Underweight (BMI < 18.5) Albumin < 3
Frailty Increased risk for pressure injury Increased likelihood of poor outcome
Diabetes (diet, oral, insulin, or no control), HgA1C > 6.5
Blood glucose harder to control in OR Higher risk of SSI postoperatively
Obesity (BMI ¼ 30.0þ)
Peripheral arterial disease
Potential difficulty inserting IABP catheter or aortic blood pressure line through atherosclerotic aortic-iliac vessels
Immunosuppressive treatment
Increased potential for SSI
Previous CABG
Potential for increased bleeding Adhesions Potential injury to pre-existing bypass conduits
Cross-clamp time
Importance of having instruments and supplies ready Knowledge of proposed surgery Possible contingencies during surgery to minimize/prevent myocardial injury from anoxia/hypoxia
Preoperative hematocrit (anemia)
Increased likelihood of intraoperative transfusions Higher likelihood of perioperative morbidity
Postoperative deep/mediastinal infection or veinharvest or cannulation site infection
Any infection occurring in the early postoperative period could indicate problems with sterility in the OR
Difficulty positioning Risk of patient fall Increased risk for pressure injury Heparin rebound with prolonged bleeding Additional stress on chest incision site closure
STS ¼ Society of Thoracic Surgeons; CABG ¼ coronary artery bypass grafting; BMI ¼ body mass index; Hg ¼ hemoglobin; SSI ¼ surgical site infection; IABP ¼ intra-aortic balloon pump. References 1. Isolated CAB procedures. The Society of Thoracic Surgeons. http://www.sts.org/sites/default/files/documents/pdf/ndb2010/On_Pump_110-112.pdf. Accessed May 10, 2016. 2. Adult Cardiac Surgery Database. The Society of Thoracic Surgeons. http://www.sts.org/sts-national-database/database-managers/adult-cardiac -surgery-database. Accessed May 10, 2016.
and cooperation are enhanced between the COS and CVOR staff members through a yearly presentation of the clinical outcome data. Discussion during the presentation between the COS and CVOR staff members often generates questions and concerns that lead to process improvement. For example, in a previous year, STS data showed an increase in leg veineharvest site infections above the benchmark. Discussion in the QI committee meeting led us to believe that we had problems with our newer staff members, who were learning to perform endoscopic vein-harvest procedures. To help properly educate our staff members on these procedures, the vendor for our equipment watched and guided the RN first assistants and physician assistants during the procedure. Improvements related to endoscopic vein harvest were made by revising the 204 j AORN Journal
process to keep the vein-harvest tract dry. The nursing and physician assistant staff members involved in the improvement effort wrote a guideline for performing the procedure. After implementation of these improvements, the incidence of vein-harvest site infections disappeared. Similar collaborative efforts have helped our surgical team members reduce the incidence of sternal incision site infections by changing the prepping and draping practices for cardiac procedures. In addition, we have reduced intraoperative use of blood products through implementation of conservation efforts like cell saving and autotransfusion.4 After our data showed that the incidence of pressure ulcers in the ICU was greater in patients with an open chest, CVOR staff
www.aornjournal.org
September 2016, Vol. 104, No. 3
Outcome Management in Cardiac Surgery
members implemented a process change in which open-chest patients were placed on special pressure-reducing beds at the end of the procedure before they were moved to the cardiac ICU. Although the STSND data elements are not specific to CVOR nursing practice, many of the data elements collected before surgery could be used to enhance and improve care in the CVOR. Table 1 lists some of these elements and describes how the information might be used for patient care in the OR. When participating in the STSND, a hospital must provide access to these data elements for staff members because this keeps them engaged in the QI process and helps them provide better patient care. At our facility, access to these STSND elements is available on the computers in our preoperative area in the history and physical information section, which is part of our electronic medical record documentation.
6.
7.
8.
9.
10.
11.
CONCLUSION Using data to improve patient care is a significant step as we continue to transition our health care system to one that focuses on the quantity of services provided to one that streamlines care to achieve the best outcome for the best value.4,11 At our facility, close communication between the COS, staff members, and leadership of the CVOR enables potential or actual problems to be identified, tracked, and corrected as needed. Such cooperation enables CVOR team members to see themselves as active participants in the care of their patients and to identify specific areas of patient care in which they can make a significant difference.
Cardiac Surgery Database. Ann Thorac Surg. 2016;101(1):33-41; discussion 41. Health policy compendium: healthcare associated infections. The Society of Thoracic Surgeons. https://www.sts.org/sites/default/files/ documents/pdf/advocacy/HAI.pdf. Accessed June 6, 2016. Speir AM, Rich JB, Crosby I, Fonner E Jr. Regional collaboration as a model for fostering accountability and transforming health care. Semin Thorac Cardiovasc Surg. 2009;21(1):12-19. Report overview: STS composite quality rating and NQF measures. The Society of Thoracic Surgeons. January 27, 2013. https:// www.sts.org/sites/default/files/documents/pdf/ndb2010/Report_ OV_NQF_44-73.pdf. Accessed June 6, 2016. Doctors & hospitals. Consumer Reports. www.consumerreports .org/cro/health/doctors-and-hospitals/index.htm. Accessed June 6, 2016. Sakallaris BR, Halpin LS, Knapp M, Sheridan MJ. Same-day transfer of patients to the cardiac telemetry unit after surgery: the Rapid After Bypass Back into Telemetry (RABBIT) program. Crit Care Nurse. 2000;20(2):50-55, 59-63, 65-68. LaPar DJ, Crosby IK, Ailawadi G, et al. Blood product conservation is associated with improved outcomes and reduced costs after cardiac surgery. J Thorac Cardiovasc Surg. 2013;145(3): 796-803; discussion 803-804.
Linda S. Halpin, MSN, RN, is a clinical outcome specialist in adult cardiac surgery at Inova Heart and Vascular Institute, Falls Church, VA. Ms Halpin has no declared affiliation that could be perceived as posing a potential conflict of interest in the publication of this article.
Bret E. Gallardo, MIS, CIS, is a data registry admin-
Acknowledgment: The authors thank Patricia C. Seifert, MSN, RN, CNOR, CRNFA(E), FAAN, independent cardiac consultant, for her suggestions and assistance in creating Table 1.
istrator at Inova Heart and Vascular Institute, Falls Church, VA. Mr Gallardo has no declared affiliation that could be perceived as posing a potential conflict of interest in the publication of this article.
References
Alan M. Speir, MD, is the medical director of cardiac surgery at Inova Heart and Vascular Institute, Falls Church, VA. Dr Speir has no declared affiliation that could be perceived as posing a potential conflict of interest in the publication of this article.
1. Codman EA. The product of a hospital. Surg Gynecol Obstet. 1914;18:491-496. 2. Donabedian A. Evaluating the quality of medical care. Milbank Mem Fund Q. 1966;44(3):166-206. 3. Halpin LS, Barnett SD, Beachy J. Cardiac surgery outcomes. Outcomes Manag. 2003;7(4):144-147. 4. Sherman RO. The drive for cost transparency in health care. Am Nurse Today. 2014;9(5):32-33. https://americannursetoday.com/ the-drive-for-cost-transparency-in-health-care [subscription required]. Accessed May 19, 2016. 5. Jacobs JP, Shahian DM, He X, et al. Penetration, completeness, and representativeness of the Society of Thoracic Surgeons Adult
www.aornjournal.org
Niv Ad, MD, is the section chief of adult cardiac surgery, and the director of cardiac surgery research at Inova Heart and Vascular Institute, Falls Church, VA. Dr Ad has no declared affiliation that could be perceived as posing a potential conflict of interest in the publication of this article.
AORN Journal j 205