No abstract

WHAT THIS PAPER ADDSThe peri-operative outcome of carotid endarterectomy is still mostly reported as a composite end point of combined ipsilateral stroke and death rate, both at individual patient level and at hospital level. This paper shows that textbook outcome, a composite measure achieved for an individual patient when all undesirable outcomes are absent, could be added to individual outcome measures to better evaluate hospital performance, especially in surgical interventions with low baseline risk such as carotid interventions.Objective: Composite measures may better objectify hospital performance than individual outcome measures (IOM). Textbook outcome (TO) is an outcome measure achieved for an individual patient when all undesirable outcomes are absent. The aim of this study was to assess TO as an additional outcome measure to evaluate quality of care in symptomatic patients treated by carotid endarterectomy (CEA). Methods: All symptomatic patients treated by CEA in 2018, registered in the Dutch Audit for Carotid Interventions, were included. TO was defined as a composite of the absence of 30 day mortality, neurological events (any stroke or transient ischaemic attack [TIA]), cranial nerve deficit, haemorrhage, 30 day readmission, prolonged length of stay (LOS; > 5 days) and any other surgical complication. Multivariable logistic regression was used to identify covariables associated with achieving TO, which were used for casemix adjustment for hospital comparison. For each hospital, an observed vs. expected number of events ratio (O/E ratio) was calculated and plotted in a funnel plot with 95% control limits. Results: In total, 70.7% of patients had a desired outcome within 30 days after CEA and therefore achieved TO. Prolonged LOS was the most common parameter (85%) and mortality the least common (1.1%) for not achieving TO. Covariates associated with achieving TO were younger age, the absence of pulmonary comorbidity, higher haemoglobin levels, and TIA as index event. In the case mix adjusted funnel plot, the O/E ratios between hospitals ranged between 0.63 and 1.27, with two hospitals revealing a statistically significantly lower rate of TO (with O/E ratios of 0.63 and 0.66). Conclusion:In the Netherlands, most patients treated by CEA achieve TO. Variation between hospitals in achieving TO might imply differences in performance. TO may be used as an additive to the pre-existing IOM, especially in surgical care with low baseline risk such as CEA.

This study demonstrates that ACD use in patients with peripheral artery disease can lead to serious adverse events resulting in increased morbidity. Therefore, the potential benefits of an ACD over MC should be carefully weighed.

Response:We thank the authors for their careful and positive feedback. Especially, we are pleased by their full support concerning the main conclusion of our article stating that irradiated patients need not to be considered a "high-risk" group for surgical revascularization.To our best knowledge, we included all available literature data on carotid revascularization after irradiation. However, the authors make an appropriate remark on our omission of not including the article by White et al. 1 When running our search queries a second time, the article by White et al again was not identified neither within applied databases nor by repeat crossreference checking despite the data from the Vascular Registry as reported by White et al 1 indeed fit our meta-analysis inclusion criteria.White et al reported 5 perioperative strokes (4.2%) in 119 carotid artery stenting procedures in previously radiated patients making carotid artery stenting clearly less advantageous. Including these outcome data in our meta-analysis would cause a shift toward better overall short-term outcome in the carotid endarterectomy group as compared with the carotid artery stenting group, although it approaches our estimated pooled risk of stroke with carotid artery stenting (3.9%) and therefore does not change final conclusions on perioperative outcome. Although the Vascular Registry also holds endarterectomy data, the analysis of patients with carotid stenosis and previous cervical radiotherapy by White et al was restricted to outcome on carotid artery stenting. Furthermore, the study only revealed periprocedural data, whereas long-term clinical outcome and restenosis, which are relevant to estimate the long-term prevention of stroke, were unreported.With respect to the other 3 comments brought up, all were actually addressed in the limitation section of our article, but perhaps not as clear as we thought it to be. As a consequence of small individual sample sizes and lack of reporting details, we were not able to distinguish between results of symptomatic versus asymptomatic status as the initial indication for revascularization. Likewise, we were not able to assess for stroke and death rates separately, and in most studies, surgical technique could not be further specified. In the descriptive result section, we summarized all adverse events individually in detail (fatal, disabling, nondisabling, or stated as "not further defined") to provide all available information in detail.The main difficulty in our analysis was the lack of individual patient data. A meta-analysis pooling the individual patient data would actually be a major step forward because such an approach allows for valid subgroup analyses and adequate control for confounding factors and not having to rely on what is reported on an aggregated level by the authors. Specified data analysis was especially difficult in articles reporting outcome on "high"-risk patients in which patients undergoing irradiation were not well stratified from the other high-risk groups.Along with Abbott ...

No abstract