Patient Safety Indicators are an insufficient performance metric to track and grade outcomes of open aortic repair

Midterm outcomes of isolated thoracic aortic replacement in congenital versus degenerative aortopathy in a 15-year institutional cohort

BACKGROUND

Open aortic replacement represents the only approved option to address thoracic aortopathy among patients with connective tissue disorders (CTD). The aim of our study was to investigate contemporary midterm outcomes of isolated thoracic aortic replacement in patients with CTD versus degenerative pathology in a large institutional cohort.

METHODS

All patients undergoing isolated open thoracic aortic replacement at a single academic center from 2005 to 2020 were included. Patients were classified as having CTD or not having CTD based on documented genetic mutations associated with congenital aortopathy. In-hospital and midterm outcomes, including mortality, paraplegia, development of new arterial pathology on surveillance imaging, and the need for future operations, were compared between groups using descriptive statistics and Kaplan-Meier survival analysis.

RESULTS

Overall, 62 patients were included with a median follow-up of 58 months (range, 19-81 months) (59 months for those with CTD vs 51.5 months for those without CTD). CTD was present in 18 patients (29%), with 16 having Marfan syndrome (77.8%). Patients with CTD were younger than patients without CTD (45.8 years vs 60.9 years) and had lower rates of smoking (5.6% vs 56.8%) and hypertension (97.7% vs 72.2%; all P < .01). Patients with CTD were more likely to have a dissection component at the time of repair compared with patients without CTD (100% vs 59.1%) and underwent repair at smaller aortic diameters (5.9 cm vs 6.6 cm; both P < .05). There were no differences in in-hospital outcomes between the two groups, including mortality (4.5% vs 5.6%) and paraplegia (2.3% vs 0%; both P > .05). At 5 years, patients with CTD were more likely to have developed aneurysmal changes distal to their thoracic repair (88.9% vs 47.7%) and extra-aortic arterial aneurysms (41.2% vs 2.3%; both P < .05). However, on survival analysis, there were no differences in freedom from additional vascular procedures (hazard ratio,1.76; P = .333) or, specifically, additional aortic procedures (hazard ratio, 1.81; P = .380) between the two groups. There was only one anastomotic complication identified on longitudinal follow-up, which occurred in a patient without CTD 8 years after the index operation.

CONCLUSIONS

Although carrying significant operative risks and the potential for morbidity, open thoracic aortic replacement represents a well-tolerated, durable treatment option for patients with congenitally mediated thoracic aortic disease. Because both patients with and without CTD who required thoracic aortic replacement often need future aortic intervention, vigilant surveillance is warranted. Equivalent intervention rates between the two groups suggest remodeling of the CTD aorta is almost universally characterized by initial postrepair dilation, but the majority of these changes successfully stabilize and do not progress to higher rates of intervention.

A retrospective study of the effects of minimally invasive colorectal surgery on Patient Safety Indicators across a five-hospital system

BACKGROUND

The Agency for Healthcare Research and Quality uses Patient Safety Indicators (PSI) to gauge quality of care and patient safety in hospitals. PSI 90 is a weighted combination of several PSIs that primarily comprises perioperative events. This score can affect reimbursement through Medicare and hospital quality ratings. Minimally invasive surgery (MIS) has been shown to decrease adverse events and outcomes. We sought to evaluate individual PSI and PSI 90 outcomes of minimally invasive versus open colorectal surgeries using a large medical database from 5 hospitals.

METHODS

A health system administrative database including all inpatients from 5 acute care hospitals was queried based on ICD 10 PC codes for colon and rectal surgery procedures performed between January 2, 2018 and December 31, 2019. Surgeries were labeled as MIS (laparoscopic) or open colorectal resection surgery. Patient demographics, health information, and case characteristics were analyzed with respect to surgical approach and PSI events. Statistical relationships between surgical approach and PSI were investigated using univariate methods and multivariate logarithmic regression analysis. PSIs of interest were PSI 8, PSI 9 PSI 11, PSI 12, and PSI 13.

RESULTS

There were 1382 operations identified, with 861 (62%) being open and 521 (38%) being minimally invasive. Logistic modeling showed no significant difference between the 2 groups for PSI 3, 6, or 8 through 15.

CONCLUSION

Understanding PSI 90 and its components is important to enhance perioperative patient care and optimize reimbursement rates. We showed that MIS, despite providing known clinical benefits, may not affect scores in the PSI 90. Surgical approach may have little effect on PSIs, and other patient and system components that are more important to these outcome measures should be pursued.

Connective tissue disease type mediates branch patency of grafts in open thoracoabdominal aortic reconstruction

OBJECTIVE

Despite a shared degenerative vascular phenotype, Marfan syndrome (MFS), Loeys-Dietz syndrome (LDS), and other genetically distinct connective tissue diseases (CTDs) have unique extravascular pathologies that impact the outcomes of aortic replacement. The aim of our study was to investigate the association of CTD genotype with postoperative outcomes and branch patency following open thoracoabdominal aortic replacement in a large institutional cohort.

METHODS

All patients undergoing open branched thoracoabdominal aortic replacement at a single academic center from 2006 to 2020 were included and classified as CTD or non-CTD based on the presence of genotypic documentation. Outcomes were compared using analysis of variance and χ² testing for continuous and discrete variables, respectively. Kaplan-Meier curves were utilized to examine patency of graft branches over time.

RESULTS

Overall, 172 patients were included, with a mean follow-up of 30.5 ± 34.9 months. CTD was present in 45 patients (26%); specifically, 32 had MFS, five had LDS, and eight had another CTD. Patients with CTDs had more extent II thoracoabdominal aneurysms (40% vs 15%), more reconstructed branches (3.5 vs 1.8), more frequently reconstructed visceral branches (86.7% vs 22.7%), and higher intraoperative blood loss (13.3 vs 6.8 L; all P < .05) compared with non-CTD patients. Patients with MFS were more frequently systemically anticoagulated preoperatively (50% vs 5%) and demonstrated higher rates of postoperative deep vein thrombosis/pulmonary embolism compared with non-CTD patients (9% vs 2%; both P < .05). Five-year renal branch patency was decreased among all patients compared with visceral branches (87.3% vs 95.6%; P = .05), but there were no individual branch patency differences between patients with and without CTDs (P = .086). Overall branch patency at 1 and 5 years was significantly higher in patients with MFS than in non-CTD patients (98.9% vs 89.1% at 5 years); there were no significant patency differences between non-CTD patients and any other CTD subgroup, mostly due to early patency loss.

CONCLUSIONS

Open thoracoabdominal reconstruction in patients with CTD is technically challenging and associated with increased transfusion and postoperative thromboembolic events when compared with non-CTD patients. Technical outcomes of the procedure are excellent and are differentially associated with genotype, with patients with MFS experiencing significantly improved branch patency over both non-CTD patients and patients with other CTDs, a finding which has multifactorial drivers.

Comparative analysis of open abdominal aortic aneurysm repair outcomes across national registries

OBJECTIVE

In a recent analysis, we discovered lower mortality after open abdominal aortic aneurysm repair (OAAA) in the Society for Vascular Surgery Vascular Quality Initiative (VQI) database when compared with previously published reports of other national registries. Understanding differentials in these registries is essential for their utility because such datasets increasingly inform clinical guidelines and health policy.

METHODS

The VQI, American College of Surgeons National Surgical Quality Improvement Program (NSQIP), and National Inpatient Sample (NIS) databases were queried to identify patients who had undergone elective OAAA between 2013 and 2016. χ² tests were used for frequencies and analysis of variance for continuous variables.

RESULTS

In total, data from 8775 patients were analyzed. Significant differences were seen across the baseline characteristics included. Additionally, the availability of patient and procedural data varied across datasets, with VQI including a number of procedure-specific variables and NIS with the most limited clinical data. Length of stay, primary insurer, and discharge destination differed significantly. Unadjusted in-hospital mortality also varied significantly between datasets: NIS, 5.5%; NSQIP, 5.2%; and VQI, 3.3%; P < .001. Similarly, 30-day mortality was found to be 3.5% in VQI and 5.9% in NSQIP (P < .001).

CONCLUSIONS

There are fundamental important differences in patient demographic/comorbidity profiles, payer mix, and outcomes after OAAA across widely used national registries. This may represent differences in outcomes between institutions that elect to participate in the VQI and NSQIP compared with patient sampling in the NIS. In addition to avoiding direct comparison of information derived from these databases, it is critical these differences are considered when making policy decisions and guidelines based on these "real-world" data repositories.