Abstract
BACKGROUND:
Failure to rescue (FTR) is a new quality measure in the adult cardiac surgery database. The Society of Thoracic Surgeons (STS) defines FTR as mortality after permanent stroke, renal failure, reoperation, or prolonged ventilation. Our objective was to assess if cardiac arrest should be included in this definition.
METHODS:
Patients undergoing an STS index operation in a regional collaborative (2011–2021) were included. The performance of the STS definition of FTR was compared with a definition which included the STS complications plus cardiac arrest (“STS+”). Centers were grouped into FTR rate terciles using the STS and STS+ definitions of FTR and changes in their relative performance rating were assessed.
RESULTS:
A total of 43,641 patients were included across 17 centers. Cardiac arrest was the most lethal complication; 55.0% of patients who experienced cardiac arrest died. FTR after any complication (13 total) occurred among 884 patients. The STS definition of FTR accounted for 83% (735/884) of all FTR. The addition of cardiac arrest to the STS definition significantly increased the proportion of overall FTR accounted for (92.2%; 815/884; p < 0.001). Choice of FTR definition led to substantial differences in center-level relative performance rating by FTR rate.
CONCLUSIONS:
Mortality after cardiac arrest is not completely captured by the STS definition of FTR, and represents an important source of potentially preventable mortality after cardiac surgery. Future quality improvement efforts using the STS definition of FTR should account for this.
Classifications: Critical Care, Quality, Failure to Rescue
Graphical Abstract
Since its inception, the field of cardiac surgery has consistently spearheaded efforts to improve healthcare quality and outcomes via the analysis of clinical data. Central to these efforts in recent years have been efforts towards the inclusion of failure to rescue (FTR) within the Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database.1 Defined as mortality after a post-operative complication, failure to rescue is intended to identify center-level preventable factors associated with mortality after adult cardiac surgery.2–5
Although risk-adjusted FTR has been shown to be valuable in measuring institutional ability to effectively address post-operative complications after cardiac surgery, no studies have examined whether the STS’ definition of FTR includes the optimal outcomes to maximize its inferential power. Currently, STS-defined FTR includes four post-operative complications: prolonged ventilation, stroke, reoperation, and renal failure.1 In prior reports, low incidence complications have been excluded from the calculation of FTR to reduce statistical noise.6 While cardiac arrest after cardiac surgery is typically less common relative to other complications such as prolonged ventilation or reoperation, a large proportion of patients who experience cardiac arrest die.7 Thus, cardiac arrest is a high FTR complication and including it in FTR analyses may capture additional institutional variance in potentially preventable post-operative mortality which is otherwise missed by the STS definition of FTR.
The objective of this study was to determine if cardiac arrest should be included in the STS definition of failure to rescue. We hypothesized that cardiac arrest would represent an important source of potentially preventable mortality in cardiac surgery and should therefore be added to the post-operative complications included in the STS definition of failure to rescue.
MATERIAL AND METHODS
The Virginia Cardiac Services Quality Initiative (VCSQI) includes 17 hospitals and surgical groups in Virginia. VCSQI data include 99% of all adult cardiac surgeries in the region. Clinical data and cost methodology have been described previously.8–9 STS data from individual centers are compiled in a central registry. This study was exempt from review by the University of Virginia’s Institutional Review Board due to the de-identified nature of the quality database (Protocol #23305, deemed exempt July 14th, 2021).
All patients undergoing an STS index operation (CABG, CABG+AVR, CABG+MVR/r, AVR, MVR/r) between July 2011 and July 2021 were extracted from the VCSQI database. Patients were excluded if they were missing STS predicted risk of mortality (Figure 1).
Figure 1.
CONSORT Diagram
Standard STS definitions were used for all variables.10 Operative mortality is defined as in-hospital mortality or death within 30 days of surgery. Failure to rescue is defined in this manuscript in three ways: STS-defined FTR, “STS+” FTR, and FTR after any complication. STS-defined FTR is operative mortality after an STS-defined FTR complication (prolonged ventilation, post-operative renal failure, reoperation, and stroke). “STS+” FTR refers to operative mortality after the STS-defined FTR complications or cardiac arrest. FTR after any complication refers to operative mortality after the STS+ complications, as well as eight additional complications (Supplemental Table 1). These definitions are not exclusive. For example, if a patient experienced both an STS-defined FTR complication and cardiac arrest, they were included in both definitions of FTR.
Patients were stratified by failure to rescue definitions (Table 1). Median (continuous) or mode (categorical) imputation was utilized for missing data(all variables missing less than 5% of data). Categorical variables are presented as counts (%) and continuous variables are presented as median (interquartile range) due to skewed distributions. Wilcoxon rank sum test was used for non-normal distributed continuous variables and the χ2 test was used for all categorical variables. All statistical analyses were carried out using SAS Version 9.4 (SAS Institutive, Cary, NC) with a p-value less than 0.05 determining significance.
Table 1.
Patient Characteristics by Definition of FTR, among Patients with Any Complication
| Characteristic | No Failure to Rescue (n = 13,958, 94.5 %) | STS-Defined FTR (n = 735, 4.98%) | STS + Cardiac Arrest FTR (n = 815, 5.50%) | p-value (Comparison is between No FTR and STS+ FTR groups) |
|---|---|---|---|---|
| Age, years | 68 (61 – 75) | 68 (61 –75) | 70 (63 – 77) | < 0.001 |
| Immunocompromised | 531 (3.80%) | 58 (7.89%) | 64 (7.85%) | < 0.001 |
| Peripheral Arterial Disease | 2002 (14.4%) | 192 (26.1%) | 214 (26.3%) | <0.001 |
| Hypertension | 12149 (87.0%) | 655 (89.1%) | 726 (89.1%) | 0.091 |
| Diabetes | 6067 (43.5%) | 359 (48.8%) | 404 (49.6%) | <0.001 |
| Prior Stroke | 1379 (9.88%) | 110 (15.0%) | 121 (14.9%) | <0.001 |
| Cerebrovascular Disease | 3232 (23.2%) | 229 (31.2%) | 248 (30.4%) | <0.001 |
| Race | 0.001 | |||
| White | 11517 (82.5%) | 564 (76.7%) | 630 (77.3%) | |
| Black | 1707 (12.2%) | 125 (17.0%) | 136 (16.7%) | |
| American Indian | 19 (0.14%) | 0 (0%) | 0 (0%) | |
| Asian | 355 (2.54%) | 26 (3.54%) | 27 (3.31%) | |
| Other | 360 (2.58%) | 20 (2.72%) | 22 (2.70%) | |
| Female Sex | 4037 (28.9%) | 309 (42.0%) | 338 (41.5%) | <0.001 |
| Liver Disease | 598 (4.28%) | 61 (8.30%) | 65 (7.98%) | <0.001 |
| MELD Score | 7.47 (6.40–9.38) | 9.22 (7.47–13.3) | 8.99 (7.47 – 13.1) | <0.001 |
| Pre-operative Serum Creatinine, mg/dL | 1.00 (0.80 – 1.20) | 1.12 (0.90 – 1.50) | 1.10 (0.90 – 1.50) | <0.001 |
| Pre-operative Serum Hemoglobin, mg/dL | 13.0 (11.7 – 14.2) | 11.9 (10.3 – 13.4) | 12.0 (10.4 – 13.4) | <0.001 |
| Pre-operative Serum WBC, × 109/L | 7.60 (6.20 – 9.30) | 8.40 (6.51 – 11.1) | 8.40 (6.59 – 10.9) | <0.001 |
| Pre-operative Serum Albumin, g/L | 3.80 (3.40 – 4.10) | 3.60 (3.00 –3.90) | 3.60 (3.00 –3.90) | <0.001 |
| Body Surface Area, m2 | 2.87 (2.75 – 3.00) | 2.83 (2.67 –2.97) | 2.83 (2.67 – 2.97) | <0.001 |
| End-Stage Renal Disease | 477 (3.42%) | 63 (8.57%) | 69 (8.47%) | <0.001 |
| Tobacco Use | 0.345 | |||
| None | 7137 (51.1%) | 362 (49.3%) | 396 (48.6%) | |
| Current | 2519 (18.1%) | 136 (18.5%) | 158 (19.4%) | |
| Former | 4302 (30.8%) | 237 (32.2%) | 261 (32.0%) | |
| Oxygen-Dependent Lung Disease | 77 (0.55%) | 11 (1.50%) | 12 (1.47%) | 0.001 |
| Chronic Lung Disease | 4536 (32.5%) | 302 (41.1%) | 337 (41.4%) | <0.001 |
| Sleep Apnea | 2604 (18.7%) | 115 (15.7%) | 130 (15.9%) | 0.053 |
| Pre-operative Arrhythmia, within 30 days | 1441 (10.3%) | 148 (20.1%) | 163 (20.0%) | <0.001 |
| CHF | 5089 (36.5%) | 409 (55.6%) | 446 (54.7%) | <0.001 |
| Previous PCI | 3543 (25.4%) | 215 (29.3%) | 238 (29.2%) | 0.015 |
| Previous CABG | 393 (2.82%) | 45 (6.12%) | 51 (6.26%) | <0.001 |
| Previous Valve Surgery | 400 (2.87%) | 39 (5.31%) | 42 (5.15%) | <0.001 |
| Prior MI | 6455 (46.3%) | 441 (60.0%) | 495 (60.7%) | <0.001 |
| Reoperative Surgery | 726 (5.2%) | 64 (8.7%) | 70 (8.6%) | <0.001 |
| Aortic Stenosis | 2908 (20.8%) | 159 (21.6%) | 176 (21.6%) | 0.603 |
| Aortic Regurgitation | 0.617 | |||
| None | 9056 (64.9%) | 472 (64.2%) | 525 (64.4%) | |
| Trivial/Trace | 1924 (13.8%) | 110 (15.0%) | 119 (14.6%) | |
| Mild | 1903 (13.6%) | 105 (14.3%) | 116 (14.2%) | |
| Moderate | 687 (4.92%) | 27 (3.67%) | 31 (3.80%) | |
| Severe | 388 (2.78%) | 21 (2.86%) | 24 (2.94%) | |
| Mitral Stenosis | 424 (3.04%) | 54 (7.35%) | 54 (6.63%) | <0.001 |
| Mitral Regurgitation | <0.001 | |||
| None | 4140 (29.7%) | 159 (21.6%) | 178 (21.8%) | |
| Trivial/Trace | 3193 (22.9%) | 129 (17.6%) | 138 (16.9%) | |
| Mild | 3676 (26.3%) | 194 (26.4%) | 227 (27.9%) | |
| Moderate | 1340 (9.60%) | 124 (16.9%) | 136 (16.7%) | |
| Severe | 1609 (11.5%) | 129 (17.6%) | 136 (16.7%) | |
| Tricuspid Stenosis | 29 (0.21%) | 1 (0.14%) | 1 (0.12%) | 0.600 |
| Tricuspid Regurgitation | <0.001 | |||
| None | 5364 (39.4%) | 232 (31.5%) | 265 (32.5%) | |
| Trivial/Trace | 4193 (30.0%) | 171 (23.3%) | 186 (22.8%) | |
| Mild | 3348 (24.0%) | 212 (28.8%) | 232 (28.5%) | |
| Moderate | 909 (6.51%) | 89 (12.1%) | 100 (12.3%) | |
| Severe | 144 (1.03%) | 31 (4.22%) | 32 (3.93%) | |
| Pre-operative Ejection Fraction, % | 55.0 (45.0 – 60.0) | 53.0 (35.0 – 60.0) | 53.0 (35.0 – 60.0) | <0.001 |
| Status | <0.001 | |||
| Elective | 6375 (45.7%) | 220 (29.9%) | 248 (30.4%) | |
| Urgent | 6994 (50.1%) | 407 (55.4%) | 450 (55.2%) | |
| Emergent | 554 (3.97%) | 92 (12.5%) | 100 (12.3%) | |
| Emergent Salvage | 35 (0.25%) | 16 (2.18%) | 17 (2.09%) | |
| Intra-Aortic Balloon Pump | 1518 (10.9%) | 256 (34.8%) | 282 (34.6%) | <0.001 |
| Procedure Type | <0.001 | |||
| AV Replacement | 1557 (11.2%) | 57 (7.76%) | 64 (7.85%) | |
| AV Replacement + CABG | 1294 (9.27%) | 78 (10.6%) | 86 (10.6%) | |
| Isolated CABG | 9060 (64.9%) | 422 (57.4%) | 476 (58.4%) | |
| MV Repair | 664 (4.76%) | 33 (4.49%) | 35 (4.29%) | |
| MV Repair + CABG | 412 (2.95%) | 38 (5.17%) | 40 (4.91%) | |
| MV Replacement + CABG | 267 (1.91%) | 37 (5.03%) | 40 (4.91%) | |
| MV Replacement Only | 704 (5.04%) | 70 (9.52%) | 74 (9.08%) | |
| Cross Clamp Time, minutes | 76.0 (60.0 –101) | 84.0 (66.0 –119) | 82.0 (65.0 – 117) | <0.001 |
| Cardiopulmonary Bypass Time, minutes | 105 (84.0 – 138) | 126 (95.0 –178) | 124 (92.0 –176) | <0.001 |
| Intra-operative Blood Products, Units | 0 (0–0) | 1 (0 – 3.00) | 1 (0–3.00) | <0.001 |
| STS Predicted Risk of Mortality, % | 1.72 (0.88 – 3.57) | 5.07 (2.21 – 13.3) | 4.72 (2.07 – 11.4) | <0.001 |
For Table 1, statistical comparisons are made between the No FTR group and the STS + Cardiac Arrest group.
RESULTS
Patient characteristics by definition of FTR
A total of 43,641 patients were identified during the study period. Operative mortality occurred among 1,017 (2.33%) patients, and 14,773 (33.8%) patients experienced one or more of the thirteen complications reported. FTR from any complication occurred among 884 (5.98%, 884 / 14,773) patients. A total of 5,315 (12.2%) patients experienced one or more of the four STS-defined FTR complications. STS-defined FTR occurred among 735 patients(13.8%, 735 / 5315). A total of 5,475 (12.5%) patients experienced one or more of the STS+ FTR complications, and STS+ FTR occurred among 815 patients (14.9%, 815 / 5475). Patients who experienced STS+ defined FTR were significantly older (70 vs. 68 years, p < 0.001), and presented with a significantly higher burden of comorbidities including peripheral arterial disease (26.3% vs. 14.4%, p < 0.001), end-stage renal disease (8.47% vs. 3.42%, p < 0.001), and congestive heart failure (54.7% vs. 36.5%, p < 0.001) relative to those who did not experience FTR (Table 1). Patients who experienced STS+ FTR were more often of non-white race (22.7% vs. 17.5%, p < 0.001) and female sex (41.5% vs. 28.9%, p < 0.001), relative to those who did not experience FTR. STS-PROM was significantly elevated among those who experienced STS+ FTR (4.72% vs. 1.72%, p < 0.001), relative to those who did not experience FTR.
Complication-specific incidence and FTR rates
Complication-specific incidence rates as well as complication-specific FTR rates can be found in Table 2. The most common individual complication was atrial fibrillation (10,360, 23.7% of all patients), followed by prolonged ventilation (3,998, 9.16%) and reoperation (1,175, 2.69%). The most lethal individual complication was cardiac arrest (444/807 cardiac arrest cases resulting in operative mortality, 55.0%), followed by sepsis (151/420 cases resulting in operative mortality, 35.9%) and post-operative renal failure (356/1165 cases resulting in operative mortality, 30.6%). The composite of complications included in the STS-definition of FTR occurred among 5,315 (12.2%) patients, whereas the STS+ definition of FTR occurred among 5475 (12.5%) patients.
Table 2.
Complication-specific incidence and Failure to Rescue Rates
| Complication | Complication-Specific Incidence; Overall: 33.9% (n = 14,773 / 43,641) | Complication-specific Failure to Rescue Rate; Overall: 6.02% (n = 884 / 14,773 |
|---|---|---|
| Post-operative Renal Failure | 2.67% (1165) | 30.6% (356) |
| Reoperation | 2.69% (1175) | 17.6% (207) |
| Permanent Stroke | 1.31% (570) | 16.8% (96) |
| Prolonged Ventilation | 9.16% (3998) | 16.7% (667) |
| STS-Defined FTR Composite (i.e., the above 4 complications) | 12.2% (5315) | 13.8% (735) |
| Cardiac Arrest | 1.85% (807) | 55.0% (444) |
| STS Complications + Cardiac Arrest | 12.5% (5475) | 14.9% (815) |
| Sepsis | 0.96% (420) | 35.9% (151) |
| Anticoagulation event | 0.60% (263) | 25.9% (68) |
| GI Event | 2.53% (1105) | 17.6% (194) |
| Pneumonia | 2.31% (1009) | 17.1% (173) |
| Venothromboembolism | 1.23% (538) | 8.36% (45) |
| Atrial Fibrillation | 23.7% (10360) | 3.41% (353) |
| Surgical Site Infection | 1.3% (586) | 3.92% (23) |
| Deep Sternal Wound Infection | 0.29% (126) | 4.76% (6) |
Complication-specific incidence and FTR rates are not exclusive (i.e., a patient with post-operative renal failure and cardiac arrest will be counted in both rows). Anticoagulation Event and GI Event are defined as per the STS ACSD (Supplemental Table 3).10
Cardiac arrest occurred among 807 patients (1.85%). Among patients with cardiac arrest, 160 (19.8%) did not experience one or more of the STS-defined FTR complications. Of the 160 patients who experienced cardiac arrest alone, the rate of mortality was 50% (n = 80). This FTR rate is not significantly different from the rate among patients who experience cardiac arrest plus one or more of the STS-defined complications (56.3%, p = 0.154).
Univariate outcomes by definition of FTR are detailed in Table 3. Total cost was significantly higher among those who experienced either definition of FTR (STS-defined: $87,824 [$53,450 – 147,872], STS+: $79,335 [$48,373–138,729], No FTR: $42,473 [$32,463 – 60,893], p<0.001), as was ICU and post-operative length of stay.
Table 3.
Univariate Outcomes by Definition of FTR
| Characteristic | No Failure to Rescue (n = 13,958, 94.5 %) | STS-Defined FTR (n = 735, 4.98%) | STS + Cardiac Arrest FTR (n = 815, 5.50%) | p-value (Comparison is between No FTR and STS+ FTR groups) |
|---|---|---|---|---|
| Operative Mortality | 69 (0.49%) | 735 (100%) | 815 (100%) | <0.001 |
| Major Morbidity | 4387 (31.4%) | 727 (98.9%) | 727 (89.2%) | <0.001 |
| Prolonged Ventilation | 3331 (23.9%) | 667 (90.8%) | 667 (81.8%) | <0.001 |
| Post-operative Renal Failure | 809 (5.76%) | 356 (48.4%) | 356 (43.7%) | <0.001 |
| Reoperation, any cause | 968 (6.94%) | 207 (28.2%) | 207 (25.4%) | <0.001 |
| Permanent Stroke | 474 (3.40%) | 96 (13.1%) | 96 (11.8%) | <0.001 |
| Post-operative Dialysis | 537 (3.85%) | 275 (37.4%) | 275 (33.7%) | <0.001 |
| Readmission | 1788 (12.8%) | 10 (1.36%) | 11 (1.35%) | <0.001 |
| Atrial Fibrillation | 10065 (72.1%) | 281 (38.2%) | 295 (36.2%) | <0.001 |
| Cardiac Arrest | 363 (2.60%) | 364 (49.5%) | 444 (54.5%) | <0.001 |
| Deep Sternal Wound Infection | 121 (0.87%) | 5 (0.68%) | 5 (0.61%) | 0.602 |
| Surgical Site Infection (Non-DSWI) | 566 (4.06%) | 20 (2.72%) | 20 (2.45%) | 0.076 |
| Deep Vein Thrombus / Pulmonary Embolism | 498 (3.57%) | 38 (5.17%) | 40 (4.91%) | 0.047 |
| Pneumonia | 843 (6.04%) | 166 (22.6%) | 166 (20.4%) | <0.001 |
| Non-home Discharge | 4652 (33.3%) | 723 (98.3%) | 803 (98.5%) | <0.001 |
| Total ICU Stay, Hours | 76.1 (45.0 – 144) | 221 (96.0 – 461) | 309 (83.0 – 423) | <0.001 |
| Length of Stay (Surgery to Discharge), Days | 8.00 (6.00–12.0) | 10 (5.00 – 21.0) | 9.00 (4.00 – 20.0) | <0.001 |
| Total Cost, USD | 42473 (32463 60893) | 87824 (53450 – 147872) | 79335 (48373 – 138729) | <0.001 |
For Table 3, statistical comparisons are made between the No FTR group and the STS + Cardiac Arrest group.
Effect of FTR definition choice on relative center performance
STS-defined FTR accounted for 83.0% (735/884) of all FTR, whereas STS+ FTR accounted for 92.2% (815/884), a significantly greater proportion of all FTR (p < 0.001). The remaining 69 cases of FTR not accounted for by the STS+ definition are attributable to the other eight complications detailed in Supplemental Table 1. As seen in Figure 2, the choice of FTR definition materially impacts centers’ relative FTR performance ratings (Supplemental Table 2). For example, when the STS+ definition is used relative to STS-defined FTR, two lowest-performing FTR tercile centers move into the mid-performing tercile and one top-performing FTR tercile center moves into the low-performing tercile. In total, 5 centers (29.4%) had discordant performance tercile ratings when the STS-definition and STS+ definition of FTR were used, respectively. Figure 3 demonstrates that when cardiac arrest is included in center-level calculations of FTR, STS-defined FTR underestimates centers’ FTR rates.
Figure 2.
Change in center-level FTR performance rating based on inclusion of cardiac arrest. Legend: Each pair of bars reflects one center. Solid color bars reflect center-level STS-defined FTR rate. Cross hatched bars reflect center-level STS definition + cardiac arrest FTR rate. When the STS+ definition is used relative to STS-defined FTR, two lowest-performing FTR tercile centers move into the mid-performing tercile and one top-performing FTR tercile center moves into the low-performing tercile. In total, 5 centers (29.4%) had discordant performance tercile rating when the STS-definition and STS+ definition of FTR were used, respectively.
Figure 3.
Center-level STS-Defined FTR rate (X-Axis) versus STS + Cardiac Arrest FTR Rate (Y-Axis). Each data point on the plot represents a different center, and compares how their performance varies with the use of the two FTR definitions (STS-defined vs. STS+). The farther a center is from the solid equivalency line, the more their performance varied based on which definition was selected.
COMMENT
In this retrospective, multi-center cohort study, we demonstrated that mortality after cardiac arrest is a source of potentially preventable mortality after cardiac surgery that is currently not captured by the STS definition of FTR. Of all the post-operative complications assessed, cardiac arrest was the most likely to result in failure to rescue. When cardiac arrest was explicitly added to the STS definition of FTR, the proportion of overall FTR accounted for by the STS definition increased substantially from 83% to 92.2% (P < 0.001). Additionally, the inclusion of cardiac arrest drove appreciable differences in relative center-level FTR performance. This suggests that expansion of the STS definition of FTR to include cardiac arrest would allow for more accurate stratification of hospitals’ FTR performance, serving as a basis for improvement at all centers.
Cardiac arrest following cardiac surgery is a relatively uncommon complication with a reported incidence of 0.7% to 8% and approximately half of these patients survive to discharge.11 Relatively speaking, these outcomes are improved when compared to cardiac arrest in other settings as the etiology tends to be reversible in the post-operative period. There have been efforts by hospital systems to develop rapid response algorithms to manage this complication in the post-operative sternotomy patient. The European Resuscitation Council (ERC) incorporated previous work by Dunning et alto develop guidelines on the management of cardiac arrest in these patients.12–13 These guidelines have since been incorporated into the Cardiac Surgery Advanced Life Support (CALS) course and now serve as the standard of care in both Europe and many centers in the United States. Implementation of this program has led to significant quality improvements in rescue from this lethal complication.14 Maccaroni et al demonstrated improvement in survival from 36.3% to 63.8% when moving away from conventional resuscitative practices toward CALS.15During this same time frame, the authors also saw a decrease in the number of arrests and rates of re-sternotomy, likely a result of staff better equipped to recognize clinical deterioration and pre-arrest situations. Thus, centers with CALS training may be superior in rescuing patients from this devastating complication. These efforts underline an important reality: rescuing a patient from cardiac arrest after cardiac surgery requires an immediate, coordinated response from the surgical team, sometimes including emergent re-sternotomy and initiation of mechanical support. A highly lethal complication which also requires a near instantaneous and extremely complex systems-level response for rescue is well-suited as a candidate complication for the STS’ FTR definition.
Our study builds upon previous research illustrating that interhospital variation in mortality is driven in part by differences in failure to rescue. Hospitals with higher observed to expected (O:E) mortality ratios have significantly higher FTR rates.4,16 In support of this assertion, Ghaferi and colleagues analyzed post-operative mortality in CABG patients; they reported 1.1-fold complication rate variability (low: 21.1% vs. high quintile: 24.2%), but 3.1-fold FTR variability (6.2% vs. 18.9%), across hospitals in different mortality quintiles.17 Likosky and colleagues noted that hospital failure to rescue rates were positively correlated with major and overall complications.4 The authors concluded that hospitals in the lower observed to expected mortality tercile group were likely more adept at recognizing and managing post-operative complications.
Many different factors potentially associated with FTR after cardiac arrest have been studied to determine which are most strongly correlated with FTR, and how to identify opportunities for improvement. LaPar and colleagues found that individual hospital, individual surgeon, and overall hospital volume were more strongly associated with mortality after cardiac arrest than the patient’s pre-operative risk profile, procedure type, or the urgency of the case.16 LaPar’s finding further demonstrates that interhospital variability in management of cardiac surgery patients is a significant determinant of their outcomes. In their review, Taenzer et al. suggest that high-performing centers have higher surgical volumes, increased surgeon experience, better physician and nursing staffing ratios in the ICU, and protocols in place for early recognition of clinical deterioration.18
We view our analysis as complimentary to that of Kurlansky et al. [1] In their study, Kurlansky and co-authors developed a complex and highly discriminatory multivariable model to predict patient-level risk of FTR. This model was derived from a national cohort of patients undergoing adult cardiac surgery who developed one or more of the following complications: prolonged ventilation, post-operative renal failure, reoperation, and stroke. This model included patients’ STS PROM, intra-operative characteristics, as well as interaction terms between the four complications themselves, and between the complications and procedure type. The model’s performance was excellent (c-statistic: 0.806). However, it is critical to note that this model was derived from a cohort of patients which excluded those who developed cardiac arrest without any of the four STS-defined FTR complications. This implies that the STS’ FTR model may not provide accurate estimated risk of FTR for patients who experience cardiac arrest in isolation, and that a center’s O:E FTR rate may not completely reflect their ability to rescue patients from cardiac arrest. In the light of our finding that the STS-definition of FTR fails to capture nearly 20% of all cases of cardiac arrest – 50% of which result in operative mortality – we believe this may suggest the need for a national analysis of the ACSD which explores how center-level O:E FTR changes when the current STS-defined FTR definition and model is used, versus when the STS+ definition and a modified risk model which includes cardiac arrest is utilized.
This study has several limitations. As in all retrospective studies, our results may be affected by unmeasured confounding, though this has been minimized by risk adjustment. Our results also may not be generalizable outside of our regional collaborative. However, this risk is reduced by the demographically and socioeconomically diverse composition of the population studied. Additionally, the cause of cardiac arrest leading to failure to rescue and methods of resuscitation were not available for analysis. Future studies may be able to stratify the underlying etiology of failure to rescue complications and variation in resuscitation techniques to further elucidate specific areas of improvement. Lastly, the STS adult cardiac surgery database does not capture the timing of post-operative complications (i.e., date/time when complication occurred), so we are unable to comment on the relative timing of patients’ complications and mortality.
Conclusion
The inclusion of cardiac arrest to the STS definition of failure to rescue improved its ability to capture failure to rescue after cardiac surgery. Future quality improvement efforts using the STS definition of FTR should consider the addition of cardiac arrest as an FTR-eligible complication.
Supplementary Material
Acknowledgements and Disclosures
Sources of Funding: This work was funded in part by a grant from the Cardiothoracic Surgery Trials Network of the National Institutes of Health Clinical Research and Implementation Skills Program under Award Number 2UM HL088925, as well as by the National Heart, Lung, and Blood Institute (grant T32 HL007849).
Funding Statement:
This work was funded in part by a grant under Award Number 2UM HL088925, as well as by the National Heart, Lung, and Blood Institute (grant T32 HL007849).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Read at the Fifty-ninth Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, January 21st – 23rd, 2023.
Conflict of Interest Statement: The authors have no conflicts of interest to disclose.
REFERENCES
- 1.Kurlansky PA, O’Brien SM, Vassileva CM, et al. Failure to Rescue: A New Society of Thoracic Surgeons Quality Metric for Cardiac Surgery. Ann Thorac Surg. 2022. Jun;113(6):1935–1942. [DOI] [PubMed] [Google Scholar]
- 2.Strobel RJ, Kaplan EF, Young AM, Rotar EP, Mehaffey JH, Hawkins RB, et al. Socioeconomic distress is associated with failure to rescue in cardiac surgery. J Thorac Cardiovasc Surg 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Strobel RJ, Charles EJ, Mehaffey JH, Hawkins RB, Quader MA, Rich JB, et al. Effect of Socioeconomic Distress on Risk-Adjusted Mortality After Valve Surgery for Infective Endocarditis. Semin Thorac Cardiovasc Surg 2022. [DOI] [PubMed] [Google Scholar]
- 4.Likosky DS, Strobel RJ, Wu X, Kramer RS, Hamman BL, Brevig JK, et al. Interhospital failure to rescue after coronary artery bypass grafting. J Thorac Cardiovasc Surg 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Reddy HG, Shih T, Englesbe MJ, Shannon FL, Theurer PF, Herbert MA, et al. Analyzing “failure to rescue”: is this an opportunity for outcome improvement in cardiac surgery? Ann Thorac Surg 2013;95:1976–81; discussion 1981. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Edwards FH, Ferraris VA, Kurlansky PA, et al. Failure to Rescue Rates After Coronary Artery Bypass Grafting: An Analysis From The Society of Thoracic Surgeons Adult Cardiac Surgery Database. Ann Thorac Surg. 2016. Aug;102(2):458–64. [DOI] [PubMed] [Google Scholar]
- 7.Society of Thoracic Surgeons Task Force on Resuscitation After Cardiac Surgery. The Society of Thoracic Surgeons Expert Consensus for the Resuscitation of Patients Who Arrest After Cardiac Surgery. Ann Thorac Surg 2017;103:1005–20. [DOI] [PubMed] [Google Scholar]
- 8.Osnabrugge RLJ, Speir AM, Head SJ, Fonner CE, Fonner E Jr, Ailawadi G, et al. Costs for surgical aortic valve replacement according to preoperative risk categories. Ann Thorac Surg 2013;96:500–6. [DOI] [PubMed] [Google Scholar]
- 9.Hawkins RB, Downs EA, Johnston LE, Mehaffey JH, Fonner CE, Ghanta RK, et al. Impact of Transcatheter Technology on Surgical Aortic Valve Replacement Volume, Outcomes, and Cost. Ann Thorac Surg 2017;103:1815–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.https://www.sts.org/registries-research-center/sts-national-database/adult-cardiac-surgery-database/data-collection. Accessed 03/01/2022.
- 11.Dunning J, Levine A, Ley J, Strang T, Lizotte DE, Lamarche Y, et al. The Society of Thoracic Surgeons Expert Consensus for the Resuscitation of Patients Who Arrest After Cardiac Surgery. Ann Thorac Surg 2017;103:1005–20. 10.1016/j.athoracsur.2016.10.033. [DOI] [PubMed] [Google Scholar]
- 12.Truhlář A, Deakin CD, Soar J, Khalifa GEA, Alfonzo A, Bierens JJLM, et al. European Resuscitation Council Guidelines for Resuscitation 2015: Section 4. Cardiac arrest in special circumstances. Resuscitation 2015;95:148–201. 10.1016/j.resuscitation.2015.07.017. [DOI] [PubMed] [Google Scholar]
- 13.Dunning J, Fabbri A, Kolh PH, Levine A, Lockowandt U, Mackay J, et al. Guideline for resuscitation in cardiac arrest after cardiac surgery. Eur J Cardiothorac Surg 2009;36:3–28. [DOI] [PubMed] [Google Scholar]
- 14.Michaelis P, Leone RJ. Cardiac Arrest After Cardiac Surgery: An Evidence-Based Resuscitation Protocol. Crit Care Nurse 2019;39:15–25. [DOI] [PubMed] [Google Scholar]
- 15.Maccaroni MR, Watson ND, Mukherjee S, Ngaage DL. Managing cardiac arrest after cardiac surgery: the impact of a five year evolving re-sternotomy policy and a review of the literature. Analg Resusc. 2013;S1. [Google Scholar]
- 16.LaPar DJ, Ghanta RK, Kern JA, Crosby IK, Rich JB, Speir AM, et al. Hospital variation in mortality from cardiac arrest after cardiac surgery: an opportunity for improvement? Ann Thorac Surg 2014;98:534–9; discussion 539–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ghaferi AA, Birkmeyer JD, Dimick JB. Complications, failure to rescue, and mortality with major inpatient surgery in Medicare patients. Ann Surg. 2009;250(6):1029–1034. [DOI] [PubMed] [Google Scholar]
- 18.Taenzer AH, Pyke JB, McGrath SP, & Warner DS (2011). A review of current and emerging approaches to address failure-to-rescue. Anesthesiology, 115(2), 421–431. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.