Abstract
Objective:
Postoperative pulmonary embolism (PE) is a rare but potentially life-threatening complication, which can be treated with extracorporeal membrane oxygenation (ECMO) therapy, a novel therapy option for acute cardiorespiratory failure. We postulate that hospitals with ECMO availability have more experienced staff, technical capabilities, and expertise in treating cardiorespiratory failure.
Design:
A retrospective analysis of surgical procedures in Germany between 2012 and 2019 was performed using hospital billing data. High-risk surgical procedures for postoperative PE were analyzed according to the availability of and expertise in ECMO therapy and its effect on outcome, regardless of whether ECMO was used in patients with PE.
Methods:
Descriptive, univariate, and multivariate analyses were applied to identify possible associations and correct for confounding factors (complications, complication management, and mortality).
Results:
A total of 13,976,606 surgical procedures were analyzed, of which 2,407,805 were defined as high-risk surgeries. The overall failure to rescue (FtR) rate was 24.4% and increased significantly with patient age, as well as type of surgery. The availability of and experience in ECMO therapy (defined as at least 20 ECMO applications per year; ECMO centers) are associated with a significantly reduced FtR in patients with PE after high-risk surgical procedures. In a multivariate analysis, the odds ratio (OR) for FtR after postoperative PE was significantly lower in ECMO centers (OR, 0.75 [0.70–0.81], P < 0.001).
Conclusions:
The availability of and expertise in ECMO therapy lead to a significantly reduced FtR rate of postoperative PE. This improved outcome is independent of the use of ECMO in these patients.
Keywords: ECMO, FtR, pulmonary embolism, surgery
Mini abstract:
Postoperative pulmonary embolism (PE) is a potentially life-threatening complication. We analyzed the impact of availability and experience for ECMO, independent of its use, in 2.4 million high-risk surgical procedures for PE and found a significantly reduced failure to rescue after postoperative PE (odds ratio, 0.75 [0.70–0.81], P < 0.001) for hospitals with ECMO.
INTRODUCTION
Approximately 266 million surgical procedures are performed worldwide each year,1 each procedure having a varying morbidity and mortality rate. Postoperative morbidity is influenced by procedure-specific complications, such as anastomotic insufficiency or postoperative bleeding, and nonspecific complications, such as inflammation or cardiopulmonary events. The expertise in perioperative care greatly influences the failure to rescue (FtR), defined as a death following a severe complication. It has been shown that a higher caseload per hospital does not necessarily decrease the rate of postoperative complications, however, it correlates with a decreased FtR rate. Similarly, a higher caseload decreases the postoperative mortality rate.2–7
In addition to the pure number of cases, the hospital’s infrastructure is also important in reducing the FtR rate.8 For example, the availability of an interventional radiology department reduces the FtR after pancreatic surgery, and resuscitation teams reduce the FtR and by this mortality rate after cardiopulmonary events.9
Pulmonary embolism (PE) is a relatively rare but potentially fatal postoperative complication.10 Perioperative thrombosis prophylaxis with low-molecular-weight heparins has significantly reduced the incidence and is recommended by international guidelines with a high level of evidence.11,12 However, despite advances in the prophylaxis of PE, it is still associated with a high mortality rate.13 PE can lead to a fulminant cardiopulmonary event requiring immediate treatment. To date, one gold standard for the treatment of acute respiratory distress syndrome beside others is extracorporeal membrane oxygenation (ECMO) therapy. Over the last 10 years, ECMO has been developed for the treatment of acute pulmonary and cardiocirculatory events. ECMO therapy withdraws blood from the venous system, which then is oxygenated and decarboxylated and, depending on the indication, returned to the body either venously or arterially. In cases of refractory pulmonary failure, the blood is returned venously (venovenous ECMO [VV-ECMO]), whereas in cardiac failure, blood is returned arterially to maintain circulation (venoarterial ECMO [VA-ECMO], also extracorporeal life support).
More than half of all ECMO applications in Germany were performed in large hospitals with more than 1000 beds and more than 75% of all ECMO applications in Germany took place in hospitals with more than 20 annual ECMO applications. Hospitals with ECMO availability often have a high level of expertise in the treatment of acute cardiopulmonary events, such as PE, independent of the use of ECMO.14
The aim of this study was to investigate whether the availability of and expertise in ECMO as a structural parameter of a hospital are associated with a lower probability for FtR after postoperative PE regardless of whether the PE was treated with ECMO. Expertise was defined as at least 20 annual ECMO therapies in the year of admission according to the recent German guideline.
METHODS
This is a register-based, retrospective, nationwide cohort study of anonymized Diagnosis-related Groups billing data provided by the “Statistisches Bundesamt” (Federal Statistical Office in Germany) for all mentioned surgical procedures performed in adults between 2012 and 2019 in Germany.
Data collection was carried out with the Research Centre of the Federal Statistical Office via remote analysis (Forschungsdatenzentrum of the Federal Statistical Offices and the federal states; data source: Diagnosis-related Group Statistics [2012–2019]) and in accordance with their privacy guidelines. Each patient record was assigned a sex and age category (18–54, 55–74, and over 75 years), an anonymized institute identifier, procedural codes (Operationen- und Prozedurenschlüssel [OPS]), main and secondary diagnoses (International Statistical Classification of Diseases and Related Health Problems; up to 100 secondary diagnoses), and reason for discharge. When duplicate case ID was identified, one data set was randomly selected to minimize bias. Only complete records were analyzed.
Due to the hospital billing system in Germany, the nature of the data in this analysis provided the opportunity to include all surgeries performed in all (private and public) hospitals nationwide. The data were analyzed according to the reason for discharge with in-hospital death as the primary endpoint. Due to the anonymization protocol of the “Statistische Bundesamt” and the system of Diagnosis-related Group statistics, it was not possible to analyze the outcome of patients after discharge or transfer to another hospital. Therefore, the data only contain information on the index admission, that is, the admission with a surgical procedure and no conclusions can be drawn about long-term readmissions, complication rates, or mortality beyond this index hospitalization.
Patients were identified using OPS codes (Supplemental Tables 1 and 2, see http://links.lww.com/AOSO/A315). The OPS codes (Supplemental Table 3, see http://links.lww.com/AOSO/A315) were grouped into procedures with a risk profile.
The next step was to identify procedure groups with a high risk for PE. The cutoff was set at 1 PE in 200 cases.
Hospitals were divided into 2 groups, according to the volume of ECMO applications performed in the same year as the surgical procedure. Hospitals with at least 20 annual ECMO applications were defined as “ECMO centers” according to national guidelines. Hospitals were allocated to one of these 2 groups for each year individually. For the purpose of this study, as a dichotomous factor influencing survival or FtR after PE, the availability of ECMO therapy at a hospital is used regardless of the exact assignment (VV-ECMO vs VA-ECMO). The German guideline for invasive ventilation and use of extracorporeal procedures in acute respiratory insufficiency, which is the basis for setting the limit at 20 applications per year, does not distinguish between VV-ECMO and VA-ECMO but provides for the sum of applications of both technologies. ICD codes for complications were identified in the secondary diagnoses of the patients.
In-house mortality was determined using the coded reason for hospital discharge as a highly validated and reliable variable. In this article, “mortality” refers to in-house mortality.
Odds ratios (ORs) were calculated as a risk assessment between the primary dependent variable, hospital mortality, and the primary independent variable, ECMO center, as well as secondary independent variables. A multivariable logistic regression model was used to analyze the association between ECMO provision and hospital mortality while taking into account possible confounding variables. To account for different comorbidity structures, we used the comorbidity score first introduced by Stausberg and Hagn,15 whose validity has been confirmed in the German variant of the ICD system. In this score, patterns of ICD secondary diagnoses reflect a comorbidity structure of a patient cohort. Potential confounding variables including comorbidity, age category, and sex were controlled for. Trends were assessed using nonparametric trend analysis.16 The presence of significant multicollinearity between confounding variables was excluded. Effect modification was tested using the Mantel–Haenszel method.
For the multivariable logistic regression model, the association between ECMO availability and in-house mortality was determined while accounting for potential confounders and the clustered data structure, treating the constant hospital identifier as a random effect. Likelihood tests were used to assess the accuracy of the regression model. Refitting the models for different quadrature points and comparing the values of the estimators helped check the accuracy of the random effects estimators. A resulting maximum difference of ≤10-4 between the distinct quadrature points was accepted.
Stata (Version 16; StataCorp LP, USA) was used for all statistical analyses. P values of ≤0.05 were considered significant.
Ethical approval was not required due to the anonymous data character. The project was, however, formally assessed and approved by the Forschungsdatenzentrum. Due to the completely anonymous nature of the data, approval by the Ethics Committee of the University of Würzburg was not required.
RESULTS
First, we defined 17 OPS groups covering the majority of noncardiac surgical interventions, identifying a total of 13,695,533 cases in Germany between 2012 and 2019 (Supplemental Table 1, see http://links.lww.com/AOSO/A315). The annual caseload was evenly distributed over the time period. The largest OPS group was trauma and orthopedic surgery with 5,932,395 cases, and the smallest group was esophageal resection with 27,888 cases. The age and sex distribution varied depending on the type of surgery (Fig. 1; Supplemental Table 1, see http://links.lww.com/AOSO/A315).
FIGURE 1.
Inclusion criteria as a flow chart (Supplemental Table 1, see http://links.lww.com/AOSO/A315 for OPS codes for primary identification and final inclusion of patient records, respectively). PE was identified by ICD code I26. ECMO application was defined by cases with OPS code 8-852.0 or 8-852.3. All patient records were complete.
A total of 44,985 postoperative pulmonary embolisms were documented in the selected patient cohort (overall incidence of 0.33%) with a slight increase over time and an overall mortality of 24.42%. The incidence of PE increased with patient age (age: 18–54 years, 0.14%; 55.74 years, 0.31%; >75 years, 0.53%) and was comparable in men and women. The incidence of postoperative PE differed significantly depending on surgical procedure ranging from 0.06% (357 of 552,808) for thyroid resections to 2.98% (831 of 27,888) for esophageal resection (Supplemental Table 1, see http://links.lww.com/AOSO/A315). Thus, we then defined procedures with more than 1 PE in 200 surgeries as “high-risk surgeries” to better tease out a potential difference in the FtR in different centers. A total of 2,407,805 patients underwent a “high-risk surgery.” These were colorectal, gastric, and urological surgeries, procedures on esophagus, pancreas, liver, and lung, and surgery on the abdominal aorta and other blood vessels. In this cohort, 23,467 PE were documented (0.97%) with a higher incidence in females (1.05% vs 0.92%; P < 0.001), in adipose patients (0.93% vs 1.5%; P < 0.001), and in patients with a cancer diagnosis (0.78% vs 1.69%; P < 0.001) and an increased incidence with patient age (18–5 years, 4:0.92%; 55–74 years, 0.94%; >75 years, 1.06%; Table 1). All further analyses were done using the patient cohort with “high-risk surgeries.”
TABLE 1.
Description of Study Cohort (Surgery With High Risk of PE) and Incidence of PE
| Total | PE, n (%) | |
|---|---|---|
| Total | 2,407,805 | 23,467 (0.97%) |
| Gender | ||
| Male | 1,449,787 (60.19%) | 13,403 (0.92%) |
| Female | 958,433 (39.81%) | 10,064 (1.05%) |
| Adipositas | ||
| BMI <30 | 2,223,153 (92.3%) | 20.384 (0.93%) |
| BMI ≥ 30 | 208,231 (7.7%) | 3084 (1.5%) |
| Cancer diagnosis | ||
| No | 1,901,007 (78.95%) | 14,893 (0.78%) |
| Yes | 506,909 (21.05%) | 8575 (1.69%) |
| Age | ||
| 18–54 | 414,505 (17.21%) | 3816 (0.92%) |
| 55–74 | 1,261,094 (52.37%) | 11,886 (0.94%) |
| >75 | 732,317 (30.41%) | 7766 (1.06%) |
| Surgery | ||
| Colorectal | 914,143 | 9387 (1.03%) |
| Urological | 368,651 | 3380 (0.92%) |
| Pancreas | 85,379 | 1685 (1.97%) |
| Gastric | 65,186 | 1318 (2.02%) |
| Esophagus | 27,888 | 831 (2.98%) |
| Liver | 37,326 | 746 (2.00%) |
| Lung | 250,484 | 2649 (1.06%) |
| Aorta | 40,201 | 296 (0.74%) |
| Vascular | 691,830 | 5288 (0.76%) |
Next, we analyzed the frequency of ECMO treatment in Germany between 2012 and 2019. In total, 39,429 patients were treated with an ECMO therapy. While the number of VV-ECMO remained relatively stable at around 2100 cases per year the number of VA-ECMO (2012: 557 to 2019: 4310) and the total number (2012: 2737; 2019: 6366) increased steadily. Also, the number of hospitals performing more than 20 ECMO therapies per year (defined as ECMO centers), as well as the percentage of ECMO therapies performed in these hospitals, increased over time (2012: 39 centers performing 74.61%, n = 2042; 2019: 80 centers performing 86.84%, n = 5528).
Looking at the “ECMO centers,” a total of 434,966 (18.06%) “high-risk surgeries” were performed with a steady increase over time (2012: 11.50%; 2019: 22.87%) (Supplemental Table 2, see http://links.lww.com/AOSO/A315).
Comparing ECMO and non-ECMO centers, patients undergoing high-risk surgery in ECMO centers were younger (64.2 years vs 67.0 years) and more frequently male (Table 2). The distribution of the types of high-risk surgeries differed with 44.3% of all liver surgeries being performed in ECMO centers compared with 11.47% of colorectal surgeries (Table 2). The percentage of patients with documented postoperative PE in high-risk surgeries was higher in ECMO than non-ECMO centers (0.85% vs 1.56%; P < 0.001). This also applied when looking at each individual surgical procedure group (Table 2).
TABLE 2.
Description of Study Cohort According to ECMO Availability
| Non-ECMO Center | ECMO Center | |
|---|---|---|
| Total | 1,972,950 (81.94%) | 434,966 (18.06%) |
| Gender | ||
| Male | 1,167,585 (59.18%) | 281,787 (64.79%) |
| Female | 805,266 (40.82%) | 153,167 (35.21%) |
| Age category | ||
| 18–54 | 323,636 (16.40%) | 90,869 (20.89%) |
| 55–74 | 1,020,647 (51.73%) | 240,447 (55.28%) |
| >75 | 628,667 (31.86%) | 103,650 (23.83%) |
| Mean age | 66.95 (66.93–66.97) | 64.19 (64.15–64.23) |
| Number | ||
| Colorectal | 809,309 (88.53%) | 104,834 (11.47%) |
| Urological | 271,183 (73.56%) | 97,468 (26.44%) |
| Pancreas | 61,365 (71.87%) | 24,014 (28.13%) |
| Gastric | 53,629 (82.27%) | 11,557 (17.73%) |
| Esophagus | 17,189 (61.64%) | 10,699 (38.36%) |
| Liver | 20,790 (55.70%) | 16,536 (44.30%) |
| Lung | 186,745 (74.45%) | 64,103 (25.55%) |
| Aorta | 30,664 (76.28%) | 9537 (23.72%) |
| Vascular | 573,169 (82.85%) | 118,661 (17.15%) |
| Incidence of PE | ||
| Total | 16,704 (0.85%) | 6766 (1.56%) |
| Colorectal | 7402 (0.91%) | 1985 (1.89%) |
| Urological | 2300 (0.85%) | 1080 (1.11%) |
| Pancreas | 1068 (1.74%) | 617 (2.57%) |
| Gastric | 952 (1.78%) | 366 (3.17%) |
| Esophagus | 434 (2.52%) | 397 (3.71%) |
| Liver | 352 (1.69%) | 394 (2.38%) |
| Lung | 1793 (0.96%) | 856 (1.34%) |
| Aorta | 195 (0.64%) | 101 (1.06%) |
| Vascular | 3.346 (0.58%) | 1942 (1.64%) |
The overall mortality rate from PE (FtR) in the high-risk surgery study population was 24.42% (5729 of 23,467). It was higher in men (25.31%) than in female (23.22%) and increased with age (18–54 years: 15.25%; 55–74 years: 22.25%; >75 years: 32.23%). The mortality/FtR rate after PE varied in the different surgery types, with the highest rate in aortic (29.05%) and vascular surgery (29.44%) (Table 3).
TABLE 3.
FtR in Patients With Postoperative PE
| Total | Non-ECMO Center | ECMO Center | P | |
|---|---|---|---|---|
| Overall | 5729 (24.41%) | 4169 (24.96%) | 1560 (23.06%) | 0.002 |
| Gender | ||||
| Male | 3392 (25.31%) | 2457 (26.35%) | 935 (22.92%) | n/a |
| Female | 2337 (23.22%) | 1712 (23.20%) | 625 (23.27%) | n/a |
| Age category | ||||
| 18–54 | 582 (15.25%) | 322 (13.85%) | 260 (17.44%) | n/a |
| 55–74 | 2645 (22.25%) | 1867 (22.58%) | 778 (21.50%) | n/a |
| >75 | 2503 (32.23%) | 1981 (32.43%) | 522 (31.50%) | n/a |
| Type of surgery | ||||
| Colorectal | 2505 (26.69%) | 1985 (26.82%) | 520 (26.20%) | 0.579 |
| Urological | 590 (17.46%) | 418 (18.17%) | 172 (15.93%) | 0.108 |
| Pancreas | 361 (21.42%) | 249 (23.31%) | 112 (18.15%) | 0.013 |
| Gastric | 315 (23.90%) | 239 (25.11%) | 76 (20.77%) | 0.098 |
| Esophagus | 166 (19.98%) | 101 (23.27%) | 65 (16.37%) | 0.013 |
| Liver | 184 (24.66%) | 81 (23.01%) | 103 (26.14%) | 0.322 |
| Lung | 645 (24.35%) | 437 (24.37%) | 208 (24.30%) | 0.967 |
| Aorta | 86 (29.05%) | 59 (30.26%) | 27 (26.73%) | 0.527 |
| Vascular | 1.557 (29.44%) | 987 (29.50%) | 570 (29.35%) | 0.910 |
n/a indicates not available.
Interestingly, the mortality/FtR rate in patients with postoperative PE was significantly lower in ECMO centers compared with non-ECMO centers (24.96% vs 23.06%, P < 0.001). While the FtR rate did not differ in more frequent procedures, such as colorectal surgery, it was significantly lower in ECMO centers in patients undergoing more complex and less frequent gastrointestinal surgeries, such as pancreas or esophageal resections (Table 2). When analyzing the FtR for postoperative PE during the analyzed time period, a significant decrease was observed over time (Fig. 2).
FIGRUE 2.
Risk-adjusted ORs with 95% CI for in-hospital mortality. (A) According to patients’ age, (B) according to patients’ gender, and (C) according to hospitals’ ECMO experience. *P = 0.002, **P < 0.001 (logistic regression analysis).
In the univariate analysis, the in-house mortality after postoperative PE (FtR) significantly increased with increasing age, male sex, comorbidity, and surgery performed in a non-ECMO center. In a multivariate model adjusting for identified confounders, there was a significantly lower OR for in-house mortality after postoperative PE in ECMO centers compared with non-ECMO centers (OR, 0.75 [0.7–0.81]; P < 0.001) (Table 4). Additionally, the subgroup analysis for high-risk abdominal surgery showed a significantly lower FtR rate in ECMO centers than in non-ECMO centers (OR, 0.69 [0.63–0.75]; P < 0.001).
TABLE 4.
Unadjusted and Adjusted Odds Ratio for In-House Mortality
| Unadjusted OR for In-Hospital Mortality After Postoperative PE | Multivariable Logistic Regression Model for In-House Mortality After Postoperative PE | |||
|---|---|---|---|---|
| Odds Ratio (95% CI) | P | Odds Ratio (95% CI) | P | |
| Age (yr) | ||||
| ≤54 | 1 | 1 | ||
| 55–74 | 1.59 (1.44–1.75) | <0.001 | 1.46 (1.32–1.62) | <0.001 |
| ≥75 | 2.64 (2.39–2.92) | <0.001 | 2.36 (2.12–2.63) | <0.001 |
| Gender | ||||
| Female | 1 | 1 | ||
| Male | 1.12 (1.05–1.19) | <0.001 | 1.11 (1.04–1.19) | 0.002 |
| Adipositas | ||||
| BMI < 30 | 1 | 1 | ||
| BMI ≥ 30 | 1.16 (1.03–1.23) | <0.001 | 1.18 (1.06–1.30) | <0.001 |
| Cancer diagnosis | ||||
| No | 1 | 1 | ||
| Yes | 0.86 (0.81–0.91) | <0.001 | 0.52 (0.48–0.55) | <0.001 |
| ECMO center | ||||
| No | 1 | 1 | ||
| Yes | 0.90 (0.84–0.96) | 0.002 | 0.75 (0.70–0.81) | <0.001 |
| Comorbidity | ||||
| No | 1 | 1 | ||
| Yes | 1.13 (1.12–1.13) | <0.001 | 1.13 (1.12–1.13) | <0.001 |
DISCUSSION
Overall, pulmonary embolism is a rare but potentially life-threatening postoperative complication with an overall perioperative mortality rate of about 25%. To date, there has been a significant improvement in the prophylaxis and reducing the risk of PE through the use of heparins; however, the high mortality rate is still unchanged. Thus, defining structural markers associated with improved treatment outcome are highly relevant and necessary to improve the outcome.
In this nationwide analysis of over 12 million patients, we identified risk factors associated with an increased likelihood of pulmonary embolism. The rate of PE correlates with the complexity of the surgical procedure and increases with patient age and comorbidities. For example, perioperative pulmonary embolism occurred in 0.10% of all appendectomies, in about 1% of colorectal procedures, and in almost 3% of esophagectomies. Similarly, previous studies have shown a higher risk of PE in women, older patients, and following surgery with a longer duration, reflecting the greater complexity of the procedure.17,18 Interestingly, trauma and orthopedic surgery had a PE rate of 0.3%, which was lower than expected. However, we included a wide range of procedures and the known high-risk procedures, such as hip surgery, only make up for a relatively small part of the selected cohort.
In recent years, knowledge about and use of cardiovascular and pulmonary replacement therapies (ECMO) has increased significantly. In order to ensure sufficient user experience, the German national guideline recommends a minimum of 20 annual ECMO applications.
Recent publications have yielded conflicting results with regard to VA-ECMO alone or in conjunction with embolectomy or thrombolysis for the resuscitation of patients with massive PE with or without cardiac arrest.19 Patients with VA-ECMO who have had an embolectomy appear to have a higher survival rate and the use of VA-ECMO also seems to reduce the incidence of organ failure.20–22 Recently, a post hoc analysis of patients treated with VA-ECMO due to refractory out-of-hospital cardiac arrest caused by PE found an inferior outcome compared to other cardiac arrest patients.23 A German study of patients with PE and VA-ECMO therapy indicated, however, that VA-ECMO may be part of a multimodal approach when treating massive PE and is associated with positive outcomes.24 In our study, we used the indicator variable, ECMO center, as a proxy for high expertise in the therapy of PE associated with major surgery. The higher expertise might include better early recognition of PE, the availability of embolectomy, and higher expertise in the resuscitation of patients with major cardiovascular events. Our study suggests that the survival after postoperative PE is improved in ECMO centers, independent of the actual use of ECMO as a treatment for PE. Thus, this improved outcome after PE might be due to the higher overall expertise in the field. Similarly, the CESAR (conventional ventilatory support vs extracorporeal membrane oxygenation for severe adult respiratory failure) trial showed that the transfer of patients with acute respiratory distress syndrome to an ECMO center was associated with lower mortality, even though not all patients received an ECMO treatment after transfer.25
Previous studies have shown that the perioperative care and expertise of the hospital are important to reduce the FtR rate. For example, the availability of an interventional radiologist significantly reduces the FtR rate after pancreatic surgery.26 In this study, we show that the availability ECMO therapy decreases the FtR rate after postoperative PE. Interestingly, this association is independent of the use of ECMO in the affected patient. This is in line with the observation already found several times in literature showing a positive correlation between the frequency of a therapy performed in a hospital and its successful outcome. In this study, ECMO maintenance represents a surrogate parameter for experience and expertise in the management of cardiorespiratory failure.
Patients treated in ECMO centers are generally younger and undergo more complex surgical procedures compared to those treated in non-ECMO centers. Surprisingly, the rate of postoperative PE is higher at ECMO centers, potentially due to a higher risk profile of the patients and procedure for PE. An indication that there is indeed a higher incidence of postoperative PE and not just a better recognition of it is supported by the fact that we were also able to see a higher incidence of postoperative resuscitations in ECMO centers. The detection of circulatory arrest is probably much less examiner dependent than the detection of PE.
Known risk factors such as obesity or the presence of a tumor disease are associated with an increased risk of postoperative PE. It is interesting to note, however, that despite an increased incidence of PE in tumor patients, mortality is significantly lower in this population in univariate and multivariate analyses. This observation can possibly be explained by the fact that these patients are more likely to be screened for PE or that they are detected during staging computed tomography scans.
The main strength of this study is the large, nationwide sample size and the completeness of the data. This allowed an adjustment for mortality and patient comorbidities.
A limitation of this analysis is the lack of information on the physicians` expertise and standard procedures in the individual hospitals. In addition, more detailed information on medical conditions, such as tumor stage, general performance score, or long-term survival of patients, was not available. Another limitation is the lack of data on readmission, as the statistics only include individual cases per hospital and do not take readmission into account. Furthermore, we could not track the course of patients after transfer to another hospital. A patient who is transferred in cardiogenic shock and dies shortly afterward is not recorded as a death in our data. However, this probably leads to an underestimation of the mortality rate in non-ECMO centers, as ECMO centers presumably transfer unstable patients with pulmonary embolisms less frequently.
In summary, we are able to show that the availability of ECMO therapy and the associated expertise in the management of cardiorespiratory failure are associated with a significant reduction of the FtR rate in postoperative PE. Given the strong correlation found between the availability of ECMO therapy and FtR after postoperative PE, we recommend high-risk surgical procedures to be performed in specialized and adequately equipped centers to improve perioperative patient outcome.
Supplementary Material
Footnotes
Published online 2 April 2024
Disclosure: The authors declare that they have nothing to disclose. This work was supported by the Open Access Publication Fund of the University of Wuerzburg.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.annalsofsurgery.com).
REFERENCES
- 1.Platz EA, Yegnasubramanian S, Liu JO, et al. A novel two-stage, transdisciplinary study identifies digoxin as a possible drug for prostate cancer treatment. Cancer Discov. 2011;1:68–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ghaferi AA, Birkmeyer JD, Dimick JB. Complications, failure to rescue, and mortality with major inpatient surgery in Medicare patients. Ann Surg. 2009;250:1029–1034. [DOI] [PubMed] [Google Scholar]
- 3.Diers J, Baum P, Wagner JC, et al. Hospital volume following major surgery for gastric cancer determines in-hospital mortality rate and failure to rescue: a nation-wide study based on German billing data (2009-2017). Gastric Cancer. 2021;24:959–969. [DOI] [PubMed] [Google Scholar]
- 4.Diers J, Baum P, Matthes H, et al. Mortality and complication management after surgery for colorectal cancer depending on the DKG minimum amounts for hospital volume. Eur J Surg Oncol. 2020;47:850–857. [DOI] [PubMed] [Google Scholar]
- 5.Uttinger KL, Diers J, Baum P, et al. Mortality, complications and failure to rescue after surgery for esophageal, gastric, pancreatic and liver cancer patients based on minimum caseloads set by the German Cancer Society. Eur J Surg Oncol. 2021;48:924–932. [DOI] [PubMed] [Google Scholar]
- 6.Diers J, Wagner J, Baum P, et al. Nationwide in-hospital mortality following colonic cancer resection according to hospital volume in Germany. BJS Open. 2019;3:672–677. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Diers J, Wagner J, Baum P, et al. Nationwide in-hospital mortality rate following rectal resection for rectal cancer according to annual hospital volume in Germany. BJS Open. 2020;4:310–319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Huijts DD, Dekker JWT, van Bodegom-Vos L, et al. Differences in organization of care are associated with mortality, severe complication and failure to rescue in emergency colon cancer surgery. Int J Qual Health Care. 2021;33:mzab038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Krautz C, Nimptsch U, Weber GF, et al. Effect of hospital volume on in-hospital morbidity and mortality following pancreatic surgery in Germany. Ann Surg. 2018;267:411–417. [DOI] [PubMed] [Google Scholar]
- 10.Porres-Aguilar M, Rivera-Lebron BN, Anaya-Ayala JE, et al. Perioperative acute pulmonary embolism: a concise review with emphasis on multidisciplinary approach. Int J Angiol. 2020;29:183–188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Geerts WH, Heit JA, Clagett GP, et al. Prevention of venous thromboembolism. Chest. 2001;119(1 Suppl):132S–175S. [DOI] [PubMed] [Google Scholar]
- 12.Hunt BJ. Preventing hospital associated venous thromboembolism. BMJ. 2019;365:l4239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ryan L, Maharjan J, Mataraso S, et al. Predicting pulmonary embolism among hospitalized patients with machine learning algorithms. Pulm Circ. 2022;12:e12013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Bercker S, Petroff D, Polze N, et al. ECMO use in Germany: an analysis of 29,929 ECMO runs. PLoS One. 2021;16:e0260324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Stausberg J, Hagn S. New morbidity and comorbidity scores based on the structure of the ICD-10. PLoS One. 2015;10:e0143365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cuzick J. A Wilcoxon-type test for trend statistics in medicine. Stat Med. 1985;4:87–90. [DOI] [PubMed] [Google Scholar]
- 17.Theochari NA, Theochari CA, Kokkinidis DG, et al. Venous thromboembolism after esophagectomy for cancer: a systematic review of the literature to evaluate incidence, risk factors, and prophylaxis. Surg Today. 2022;52:171–181. [DOI] [PubMed] [Google Scholar]
- 18.Inderbitzin DT, Opitz I, Giger U, et al. Incidence of clinical pulmonary embolism after laparoscopic surgery. Br J Surg. 2007;94:599–603. [DOI] [PubMed] [Google Scholar]
- 19.Steinhorn R, Dalia AA, Bittner EA, et al. Surgical pulmonary embolectomy on VA-ECMO. Respir Med Case Rep. 2021;34:101551. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Karami M, Mandigers L, Miranda DDR, et al. ; DUTCH ECLS Study Group. Survival of patients with acute pulmonary embolism treated with venoarterial extracorporeal membrane oxygenation: a systematic review and meta-analysis. J Crit Care. 2021;64:245–254. [DOI] [PubMed] [Google Scholar]
- 21.Guliani S, Das Gupta J, Osofsky R, et al. Venoarterial extracorporeal membrane oxygenation is an effective management strategy for massive pulmonary embolism patients. J Vasc Surg Venous Lymphat Disord. 2021;9:307–314. [DOI] [PubMed] [Google Scholar]
- 22.Kmiec L, Philipp A, Floerchinger B, et al. Extracorporeal membrane oxygenation for massive pulmonary embolism as bridge to therapy. ASAIO J. 2020;66:146–152. [DOI] [PubMed] [Google Scholar]
- 23.Pudil J, Rob D, Smalcova J, et al. Pulmonary embolism-related refractory out-of-hospital cardiac arrest and extracorporeal cardiopulmonary resuscitation: Prague OHCA study post hoc analysis. Eur Heart J Acute Cardiovasc Care. 2023;12:507–512. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hobohm L, Sagoschen I, Habertheuer A, et al. Clinical use and outcome of extracorporeal membrane oxygenation in patients with pulmonary embolism. Resuscitation. 2022;170:285–292. [DOI] [PubMed] [Google Scholar]
- 25.Peek GJ, Mugford M, Tiruvoipati R, et al. ; CESAR trial collaboration. Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial. Lancet. 2009;374:1351–1363. [DOI] [PubMed] [Google Scholar]
- 26.Sakai N, Yoshitomi H, Furukawa K, et al. Outcome of interventional radiology for delayed postoperative hemorrhage in hepatobiliary and pancreatic surgery. J Gastroenterol Hepatol. 2020;35:2264–2272. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.