Ninety-Day Readmissions of Bundled Valve Patients: Implications for Healthcare Policy.

Medicare's Bundle Payment for Care Improvement (BPCI) Model 2 groups reimbursement for valve surgery into 90-day episodes of care, which include operative costs, inpatient stay, physician fees, postacute care, and readmissions up to 90 days postprocedure. We analyzed our BPCI patients' 90-day outcomes to understand the late financial risks and implications of the bundle payment system for valve patients. All BPCI valve patients from October 2013 (start of risk-sharing phase) to December 2015 were included. Readmissions were categorized as early (≤30 days) or late (31-90 days). Data were collected from institutional databases as well as Medicare claims. Analysis included 376 BPCI valve patients: 202 open and 174 transcatheter aortic valves (TAVR). TAVR patients were older (83.6 vs 73.8 years; P = 0.001) and had higher Society of Thoracic Surgery predicted risk (7.1% vs 2.8%; P = 0.001). Overall, 18.6% of patients (70/376) had one-or-more 90-day readmission, and total claim was on average 51% greater for these patients. Overall readmissions were more common among TAVR patients (22.4% (39/174) vs 15.3% (31/202), P = 0.052) as was late readmission. TAVR patients had significantly higher late readmission claims, and early readmission was predictive of late readmission for TAVR patients only (P = 0.04). Bundled claims for a 90-day episode of care are significantly increased in patients with readmissions. TAVR patients represent a high-risk group for late readmission, possibly a reflection of their chronic disease processes. Being able to identify patients at highest risk for 90-day readmission and the associated claims will be valuable as we enter into risk-bearing episodes of care agreements with Medicare.

Del Nido cardioplegia for minimally invasive aortic valve replacement.

BACKGROUND: We analyzed the impact and safety of del Nido Cardioplegia (DNC) in patients undergoing minimally invasive aortic valve replacement (MIAVR). METHODS: We analyzed all isolated MIAVR replacements from 5/2013-6/2015 excluding re-operative patients. The approach was a hemi-median sternotomy in all patients. Patients were divided into two cohorts, those who received 4:1 crystalloid:blood DNC solution and those in whom standard 1:4 Buckberg-based cardioplegia (WBC) was used. One-to-one propensity case matching of DNC to WBC was performed based on standard risk factors and differences between groups were analyzed using chi-square and non-parametric methods. RESULTS: MIAVR was performed in 181 patients; DNC was used in 59 and WBC in 122. Case matching resulted in 59 patients per cohort. DNC was associated with reduced re-dosing (5/59 (8.5%) versus 39/59 (61.0%), P < 0.001) and less total cardioplegia volume (1290 ± 347 mL vs 2284 ± 828 mL, P < 0.001). Antegrade cardioplegia alone was used in 89.8% (53/59) of DNC patients versus 33.9% (20/59) of WBC patients (P < 0.001). Median bypass and aortic cross-clamp times were similar. Clinical outcomes were similar with respect to post-operative hematocrit, transfusion requirements, need for inotropic/pressor support, duration of intensive care unit stay, re-intubation, length of stay, new onset atrial fibrillation, and mortality. CONCLUSIONS: Del Nido cardioplegia usage during MIAVR minimized re-dosing and the need for retrograde delivery. Patient safety was not compromised with this technique in this group of low-risk patients undergoing MIAVR.

Use of electromagnetic navigational bronchoscopy in robotic pulmonary resection.

Robotic resection of pulmonary lesions has become a more common approach in the field of thoracic surgery. The greatest drawback of robotic resection is the lack of tactile feedback as compared to open approaches, making identification of intrapulmonary lesion difficult. Electromagnetic navigational bronchoscopy (navibronch) enables pre-incisional marking of pulmonary lesions for intraoperative identification. We sought to determine how effective navibronch was in our institution's robotic cases. Thirty-one patients underwent robotically assisted resection of 35 lesions with the assistance of navibronch from 7/2014 to 9/2015. Retrospective demographic and operative data were collected on these patients, and statistical analysis was conducted using ANOVA means testing, Chi-square, and non-parametric tests. The average age in this patient population was 63.7 ± 13.5 years. Eight patients (25.8%) were male. Twenty-five (80.6%) of the patients had pathology involving one lobe, with six (19.4%) in two lobes. 34 of the resections (97.1%) resulted in dye being localized to the first specimen; 34 (97.1%) were found to have the target pathology in the initial specimen. Further resection was carried out in 22 (62.9%) cases, with the final resection resulting in a segment in 2 (5.7%) and a lobe in 14 (40.0%). The mean number of lung specimens collected was 1.94 ± 0.13. The mean number of tumors in each target resection was 1.46 ± 0.66 in final pathology. Malignancy was found in 19 (54.3%) of final specimens. There were no complications related to navibronch. Navibronch is an effective technique in the identification and localization of pulmonary lesions in robotically assisted lung resections.

A Contemporary Approach to Reoperative Aortic Valve Surgery: When is Less, More?

OBJECTIVE: Although the benefits of minimally invasive valvular surgery are well established, the applicability of extending these techniques to reoperative aortic valve surgery is unknown. We evaluated our experience with a minimally invasive approach to this patient population. METHODS: From January 2010 to September 2015, 21 patients underwent reoperative isolated aortic valve replacement via a minimally invasive approach by a single surgeon. All patients had preoperative evaluation with computerized tomography and coronary catheterization. Surgical approaches were right anterior thoracotomy (6/21) or upper hemisternotomy (15/21). Central aortic cannulation was preferred with femoral artery cannulation used in four patients (19%). In patients with left internal mammary artery (LIMA) grafts, no attempt to dissect or occlude the graft was made. Cold blood cardioplegia was administered antegrade (12/21) or retrograde (9/21); systemic cooling with a mean low temperature of 27.5 °C was employed. RESULTS: Mean age was 75.1 years with a range from 33 to 92 years, and 67% (14/21) were male. All procedures were completed with a minimally invasive approach. Mean ± SD cross-clamp time was 51.5 ± 9.2 minutes. Fourteen patients had patent LIMA grafts. No aortic, LIMA, or cardiac injuries occurred. There were no hospital deaths nor occurrences of perioperative myocardial infarction, stroke, wound infection, renal failure, or endocarditis/sepsis. One patient required a reoperation for bleeding. Sixty-two percent of patients were discharged to home; mean ± SD length of stay was 6 ± 3 days. CONCLUSIONS: With appropriate preoperative evaluation and careful surgical planning, a minimally invasive approach to reoperative aortic valve surgery can be performed in a safe and effective manner.