- Email: [email protected]
Characterization of Planned and Unplanned 30-Day Readmissions Following Vascular Surgical Procedures
1540
Abstracts
Journal of Vascular Surgery May 2017
(0-6 months after randomization) mortality was lower in the EVAR groups (46 of 1393 vs 73 of 1390 deaths; P ¼ .010), primarily because the 30-day operative mortality was lower in the EVAR groups (16 deaths vs 40 for open repair). Later (#3 years), the survival curves converged and remained so to 8 years with a nonsignificant shift of the hazard ratio in favor of open repair. At 5 years, the estimated survival rate was 73.6% in both groups. Beyond 3 years, aneurysm-related mortality was significantly higher in the EVAR groups (19 deaths vs 3 for open repair; P ¼ .010). Patients with moderate renal dysfunction or previous coronary artery disease had no early survival advantage with EVAR. None of the morphologic aneurysm characteristics, smoking, diabetes, basal metabolic index, age, or gender was associated with mortality. Those with peripheral artery disease (ABI <0.9) had lower mortality under open repair (39 vs 62 deaths; P ¼ .022) in the period from 6 months to 4 years after randomization. The overall rates of reintervention were higher in the EVAR group (range, 5.1%8.5%) than the open group (range, 1.9%-4.6%). The commonest complications after EVAR was a type II endoleak (325 patients [11.7%]), with correction deemed necessary in 22.8%. The second most common was a type I endoleak (120 [4.3%]) and received early intervention in 65.8%. For those 37 patients with reported sac rupture, the median time to rupture was 3.5 years. All but 19 had a known graft endoleak or migration before rupture, of which approximately two-thirds had been addressed. The 30-day mortality with rupture was 62%. Comments: This meta-analysis confirms the fact that EVAR is not the definitive repair that an open operation provides. It may well benefit those with a limited life expectancy such as the elderly and others who are less able to weather the stresses of an open operation. In some patients (significant renal impairment and coronary artery disease), EVAR may not even provide an early mortality advantage. In addition, an open operation may well be the best therapy for the young and healthy. No matter, if you chose to treat a patient with an abdominal aortic aneurysm with EVAR, surveillance is critical to reduce aneurysmrelated deaths in the mid- and long-term. This may change with improved devices, but for now, this is the crucial clinical message.
Characterization of Planned and Unplanned 30-Day Readmissions Following Vascular Surgical Procedures Tahhan G, Farber A, Shah NK, Krafcik BM, Sachs TE, Kalish JA, et al. Vasc Endovascular Surg 2017;51:17-22. Conclusions: Causes for readmission of vascular surgery patients are multifactorial. Infections, both those related and unrelated to the surgical site, remain common reasons for readmission and represent an opportunity for improvement strategies. Improved understanding of readmissions after vascular surgery could help adjust policy benchmarks for targeted readmission rates and help reduce resource utilization. Summary: The goal of this single-center retrospective study was to evaluate 30-day readmissions after vascular surgery to identify patient factors that indicate a high risk of readmission. If these factors can be appropriately modified, it may be possible to prevent the need for readmission. All patients between October of 2012 and March of 2015 who were discharged from the vascular surgery service and readmitted in 30 days were included. Patient demographics, specifics of the operative intervention, the postoperative length of stay (LOS), and reasons for readmission were collected. A multivariate analysis of readmission LOS was conducted based on patient demographics, pre-existing comorbidities, index LOS, index procedure type, readmission service and location, and reason for readmission. A multivariate analysis of the reasons for short-term (0-7 days) readmission compared with long-term (8-30 days) readmission was also conducted for all readmissions and exclusively unplanned readmissions to determine whether there were hyperacute issues leading to very early readmission that could be targeted for modification. A P value of <.05 was considered statistically significant. During this interval, 649 vascular surgical discharges occurred, with 135 (21%) being readmitted. Most of the readmitted patients were overweight (35%) or obese (30%). These patients had common comorbidities associated with vascular disease: diabetes (56%), coronary artery disease (40%), congestive heart failure (24%), and chronic obstructive lung disease (19%). Index vascular operations included open lower extremity procedures (39%), diagnostic angiograms (35%), endovascular lower extremity procedures (16%), dialysis access procedures (7%), carotid/cerebrovascular procedures (7%), amputations (6%), and abdominal aortic procedures (5%). The average postoperative LOS of the index procedure was 7.48 6 6.73 days.
The average time to readmission was 13 days, but 35% came back in less than 8 days and 21% after 22 days. Planned readmission for same site (79%), different site (14%), or foot operation (7%) was the reason for the 30-day readmission in 21.5% of cases. Most of the unplanned readmissions were medical in nature (43%), specifically related to failure to thrive (28%), infection unrelated to the site of operation (24%), and hypoxia/ congestive heart failure (21%). Surgical complications accounted for 35.5% of readmissions and were generally related to surgical site infections (69%), graft failure (19%), or bleeding (8%). Readmission LOS averaged 7.43 6 7.22 days. Multivariate analysis did not find a statistical difference between those readmitted early vs after 7 days. The multivariate analysis of readmission LOS found no factor of significance. Comments: Reimbursement for 30-day readmissions will likely be severely curtailed in the future based on the thought that the surgeon has control to prevent or modify the reasons for readmission. Unfortunately, the authors did not analyze from a statistical standpoint what factors drove readmissions, but that may not have been productive considering the many reasons for readmissions. Vascular surgeons will likely have to modify their practice to limit planned readmissions (or prove to insurers that it is required); this could cut the number of readmissions by one-fourth. Increased attention to those factors that prevent surgical wound infection (surgical preps, perioperative antibiotics, blood use, etc.) could decrease readmission by another quarter at least.
Management of Immediate Post-Endovascular Aortic Aneurysm Repair Type 1A Endoleaks and Late Outcomes AbuRahma AF, Hass SM, AbuRahma ZT, Yacoub M, Mousa AY, Abu-Halimah S, et al. J Am Coll Surg 2017;xx:1-9. http://dx.doi.org/ 10.1016/j.jamcollsurg.2016.12.014 Conclusions: Those patients with a type Ia endoleak on completion angiogram immediately after an endovascular aortic aneurysm repair (EVAR) have higher rates of early intervention, late endoleaks, and reinterventions, which highlights the need for a strict post-EVAR surveillance, especially in this patient cohort. Summary: This 10-year retrospective study of a prospectively collected EVAR database was used to compare the early (30 days) and late outcomes of those with (group A) and without (group B) a type Ia endoleak on completion angiogram. The aim is to analyze the incidence, management, and late outcome of these endoleaks. Six different EVAR graft designs were used during the conduct of this review. The standard surveillance protocol consisted of computed tomography angiography (CTA) or color duplex ultrasound imaging #30 days, and if normal, a color duplex study at 6, 12, and every 12 months thereafter. A CTA was obtained only if there was evidence of sac enlargement or endoleak, or both, on the duplex study. Definitions primary end points closely followed the Society for Vascular Surgery recommendations on stent graft standardized reporting. An “early endoleak” was that seen on completion angiography or #30 days, and a “late endoleak” occurred >30 days. Sac expansion was a >5 mm increase compared with preoperative sac size, and significant shrinkage was the reverse. Migration was displacement from the lowest renal to the most cephalad part of the stent graft by >10 mm compared with the previous study. The primary end point included early outcomes: rate of early endoleak and rate of early intervention. Late clinical outcomes included late type Ia endoleaks, aortic sac expansion, late reintervention to treat endoleak or other complications, stent migration, conversion to open repair, and late mortality (aneurysm-related deaths). All deaths were verified using the Social Security Death Index. Individual comparison used contingency table analysis with c2 or Fisher exact test and t-tests. The Kaplan-Meier method was used to compare rates of freedom from late type Ia endoleak, late intervention, sac expansion, and survival for both groups. Comparisons were based on the log-rank test. An immediate post-EVAR type Ia endoleak occurred in 71 of 565 patients (12.6%). The only demographic difference was age, in that the group A patients were older (mean, 75.7 [range, 48-96] vs 72.9 [range, 45-101] years; P ¼ .01). Early intervention (proximal aortic cuffs or stenting, or both) was used in 56 of 71 (79%) in group A vs 31 of 494 (6%) in group B (P < .0001) to correct the endoleak. The 15 remaining patients in group A, who were designated to have no early intervention, were treated with percutaneous transluminal angioplasty of the proximal aortic attachment site alone or were believed to have minimal proximal type Ia endoleak, which did not justify early intervention. Thirty-one patients in group B had early intervention (proximal aortic cuffs and/or aortic stents) for type Ia endoleaks that
Abstracts
Journal of Vascular Surgery May 2017
(0-6 months after randomization) mortality was lower in the EVAR groups (46 of 1393 vs 73 of 1390 deaths; P ¼ .010), primarily because the 30-day operative mortality was lower in the EVAR groups (16 deaths vs 40 for open repair). Later (#3 years), the survival curves converged and remained so to 8 years with a nonsignificant shift of the hazard ratio in favor of open repair. At 5 years, the estimated survival rate was 73.6% in both groups. Beyond 3 years, aneurysm-related mortality was significantly higher in the EVAR groups (19 deaths vs 3 for open repair; P ¼ .010). Patients with moderate renal dysfunction or previous coronary artery disease had no early survival advantage with EVAR. None of the morphologic aneurysm characteristics, smoking, diabetes, basal metabolic index, age, or gender was associated with mortality. Those with peripheral artery disease (ABI <0.9) had lower mortality under open repair (39 vs 62 deaths; P ¼ .022) in the period from 6 months to 4 years after randomization. The overall rates of reintervention were higher in the EVAR group (range, 5.1%8.5%) than the open group (range, 1.9%-4.6%). The commonest complications after EVAR was a type II endoleak (325 patients [11.7%]), with correction deemed necessary in 22.8%. The second most common was a type I endoleak (120 [4.3%]) and received early intervention in 65.8%. For those 37 patients with reported sac rupture, the median time to rupture was 3.5 years. All but 19 had a known graft endoleak or migration before rupture, of which approximately two-thirds had been addressed. The 30-day mortality with rupture was 62%. Comments: This meta-analysis confirms the fact that EVAR is not the definitive repair that an open operation provides. It may well benefit those with a limited life expectancy such as the elderly and others who are less able to weather the stresses of an open operation. In some patients (significant renal impairment and coronary artery disease), EVAR may not even provide an early mortality advantage. In addition, an open operation may well be the best therapy for the young and healthy. No matter, if you chose to treat a patient with an abdominal aortic aneurysm with EVAR, surveillance is critical to reduce aneurysmrelated deaths in the mid- and long-term. This may change with improved devices, but for now, this is the crucial clinical message.
Characterization of Planned and Unplanned 30-Day Readmissions Following Vascular Surgical Procedures Tahhan G, Farber A, Shah NK, Krafcik BM, Sachs TE, Kalish JA, et al. Vasc Endovascular Surg 2017;51:17-22. Conclusions: Causes for readmission of vascular surgery patients are multifactorial. Infections, both those related and unrelated to the surgical site, remain common reasons for readmission and represent an opportunity for improvement strategies. Improved understanding of readmissions after vascular surgery could help adjust policy benchmarks for targeted readmission rates and help reduce resource utilization. Summary: The goal of this single-center retrospective study was to evaluate 30-day readmissions after vascular surgery to identify patient factors that indicate a high risk of readmission. If these factors can be appropriately modified, it may be possible to prevent the need for readmission. All patients between October of 2012 and March of 2015 who were discharged from the vascular surgery service and readmitted in 30 days were included. Patient demographics, specifics of the operative intervention, the postoperative length of stay (LOS), and reasons for readmission were collected. A multivariate analysis of readmission LOS was conducted based on patient demographics, pre-existing comorbidities, index LOS, index procedure type, readmission service and location, and reason for readmission. A multivariate analysis of the reasons for short-term (0-7 days) readmission compared with long-term (8-30 days) readmission was also conducted for all readmissions and exclusively unplanned readmissions to determine whether there were hyperacute issues leading to very early readmission that could be targeted for modification. A P value of <.05 was considered statistically significant. During this interval, 649 vascular surgical discharges occurred, with 135 (21%) being readmitted. Most of the readmitted patients were overweight (35%) or obese (30%). These patients had common comorbidities associated with vascular disease: diabetes (56%), coronary artery disease (40%), congestive heart failure (24%), and chronic obstructive lung disease (19%). Index vascular operations included open lower extremity procedures (39%), diagnostic angiograms (35%), endovascular lower extremity procedures (16%), dialysis access procedures (7%), carotid/cerebrovascular procedures (7%), amputations (6%), and abdominal aortic procedures (5%). The average postoperative LOS of the index procedure was 7.48 6 6.73 days.
The average time to readmission was 13 days, but 35% came back in less than 8 days and 21% after 22 days. Planned readmission for same site (79%), different site (14%), or foot operation (7%) was the reason for the 30-day readmission in 21.5% of cases. Most of the unplanned readmissions were medical in nature (43%), specifically related to failure to thrive (28%), infection unrelated to the site of operation (24%), and hypoxia/ congestive heart failure (21%). Surgical complications accounted for 35.5% of readmissions and were generally related to surgical site infections (69%), graft failure (19%), or bleeding (8%). Readmission LOS averaged 7.43 6 7.22 days. Multivariate analysis did not find a statistical difference between those readmitted early vs after 7 days. The multivariate analysis of readmission LOS found no factor of significance. Comments: Reimbursement for 30-day readmissions will likely be severely curtailed in the future based on the thought that the surgeon has control to prevent or modify the reasons for readmission. Unfortunately, the authors did not analyze from a statistical standpoint what factors drove readmissions, but that may not have been productive considering the many reasons for readmissions. Vascular surgeons will likely have to modify their practice to limit planned readmissions (or prove to insurers that it is required); this could cut the number of readmissions by one-fourth. Increased attention to those factors that prevent surgical wound infection (surgical preps, perioperative antibiotics, blood use, etc.) could decrease readmission by another quarter at least.
Management of Immediate Post-Endovascular Aortic Aneurysm Repair Type 1A Endoleaks and Late Outcomes AbuRahma AF, Hass SM, AbuRahma ZT, Yacoub M, Mousa AY, Abu-Halimah S, et al. J Am Coll Surg 2017;xx:1-9. http://dx.doi.org/ 10.1016/j.jamcollsurg.2016.12.014 Conclusions: Those patients with a type Ia endoleak on completion angiogram immediately after an endovascular aortic aneurysm repair (EVAR) have higher rates of early intervention, late endoleaks, and reinterventions, which highlights the need for a strict post-EVAR surveillance, especially in this patient cohort. Summary: This 10-year retrospective study of a prospectively collected EVAR database was used to compare the early (30 days) and late outcomes of those with (group A) and without (group B) a type Ia endoleak on completion angiogram. The aim is to analyze the incidence, management, and late outcome of these endoleaks. Six different EVAR graft designs were used during the conduct of this review. The standard surveillance protocol consisted of computed tomography angiography (CTA) or color duplex ultrasound imaging #30 days, and if normal, a color duplex study at 6, 12, and every 12 months thereafter. A CTA was obtained only if there was evidence of sac enlargement or endoleak, or both, on the duplex study. Definitions primary end points closely followed the Society for Vascular Surgery recommendations on stent graft standardized reporting. An “early endoleak” was that seen on completion angiography or #30 days, and a “late endoleak” occurred >30 days. Sac expansion was a >5 mm increase compared with preoperative sac size, and significant shrinkage was the reverse. Migration was displacement from the lowest renal to the most cephalad part of the stent graft by >10 mm compared with the previous study. The primary end point included early outcomes: rate of early endoleak and rate of early intervention. Late clinical outcomes included late type Ia endoleaks, aortic sac expansion, late reintervention to treat endoleak or other complications, stent migration, conversion to open repair, and late mortality (aneurysm-related deaths). All deaths were verified using the Social Security Death Index. Individual comparison used contingency table analysis with c2 or Fisher exact test and t-tests. The Kaplan-Meier method was used to compare rates of freedom from late type Ia endoleak, late intervention, sac expansion, and survival for both groups. Comparisons were based on the log-rank test. An immediate post-EVAR type Ia endoleak occurred in 71 of 565 patients (12.6%). The only demographic difference was age, in that the group A patients were older (mean, 75.7 [range, 48-96] vs 72.9 [range, 45-101] years; P ¼ .01). Early intervention (proximal aortic cuffs or stenting, or both) was used in 56 of 71 (79%) in group A vs 31 of 494 (6%) in group B (P < .0001) to correct the endoleak. The 15 remaining patients in group A, who were designated to have no early intervention, were treated with percutaneous transluminal angioplasty of the proximal aortic attachment site alone or were believed to have minimal proximal type Ia endoleak, which did not justify early intervention. Thirty-one patients in group B had early intervention (proximal aortic cuffs and/or aortic stents) for type Ia endoleaks that