Introduction There are few published data evaluating the incidence of mechanical thrombectomy among stroke centers or the times at which they occur.
Methods A multicenter retrospective study was performed to identify all patients undergoing emergent thrombectomy for acute ischemic stroke during a 3-month period (June through August 2016). Consultations that did not undergo thrombectomy were not included.
Results Ten institutions participated in the study. During the 92-day study period, a total of 189 patients underwent mechanical thrombectomy. The average number of procedures per hospital over the study period was 18.9 (average of 0.2 cases per day per or 75.6 cases per year). This ranged from 0.09 cases per day at the lowest volume center to 0.49 cases per day at the highest volume center. Procedures were more common on weekdays (p<0.001) and during non-work hours (p<0.001). The most common period for thrombectomy procedures was between 20:00 and 21:00 hours. The median time from notification to groin puncture was 84 min (IQR 56–145 min) and from puncture to closure was 57 min (IQR 33–80 min). The median time from imaging completion to procedural start was 52 min longer for non-work hours than during work hours (p<0.001). There were no differences in procedural length based on day of the week or time of day.
Conclusions These findings indicate that the majority of mechanical thrombectomy cases occur during non-work hours, with associated off-hours delays, which has important operational implications for hospitals implementing stroke call coverage.
- Mechanical thrombectomy
- Acute ischemic stroke
Statistics from Altmetric.com
Recently, seven randomised controlled trials were published showing benefit to mechanical thrombectomy in the setting of anterior circulation emergent large vessel occlusion (ELVO).1–7 The positive results from these trials led the American Heart Association/American Stroke Association to update societal guidelines and provide a Class 1, level of evidence A recommendation that thrombectomy be pursued in patients with ELVO who meet the appropriate criteria.8 Subsequently, a recent national survey of physicians who perform thrombectomy reported increasing procedural volumes with many reporting increased aggressiveness in patient selection since the publication of these trials.9 The proven benefit of mechanical thrombectomy for patients with ELVO has led to renewed efforts to optimize logistics to improve patient outcomes, but also debate regarding who should be performing these procedures and in what hospital settings. Debating this requires not just an understanding of current limitations to stroke care but also a clear grasp of operational factors. These factors include, among other considerations, the time commitment expected of call teams, available procedural suites, and the percentage of procedures occurring during work hours versus non-work hours.
While the literature on emergent thrombectomy procedural volumes is relatively scant, evidence suggests that the number of cases being treated with mechanical thrombectomy has been increasing rapidly over the past several years.9–12 There are currently no published data evaluating the incidence of mechanical thrombectomy among stroke centers or the times at which they occur. This study was designed to quantify these factors by assessing thrombectomy cases performed at 10 geographically diverse stroke centers, ranging in programme activity, from around the USA.
Institutional review board approval was obtained at all participating centers. Each institution queried its respective stroke database in order to retrospectively identify all patients undergoing emergent angiography with intent to perform thrombectomy for ELVO during a 3-month time period (1 June 2016 to 31 August 2016). All patients who underwent either mechanical thrombectomy or emergent cerebral angiography without revascularization but with the intent to perform thrombectomy were included in the study. De-identified data on each patient were collected and transmitted to the primary center for analysis. Data collected included: origin of patient (inpatient, hospital–hospital transfer, or emergency room arrival), time of initial consultation or completion of angiographic imaging (representing time of initial neuroendovascular team notification), time of groin puncture, time of groin closure (representing end time of procedure), and day of the week (Monday to Sunday). For the purposes of analysis, ‘day time’ was defined as 06:00 hours to 21:59 hours and ‘night time’ as 22:00 hours to 05:59 hours, based on typical sleep and wake cycles. ‘Work hours’ were defined as 07:00 hours to 16:59 hours Monday through Friday while ‘non-work hours’ were defined as 17:00 hours to 06:59 hours Monday through Friday and all day Saturday and Sunday.
Data analysis was performed using SPSS Version 24.0 for Mac (IBM Corp, Armonk, New York, USA). The Kolomogorov–Smirnov test was used to evaluate for normality of distribution of data. The χ2 test for goodness of fit, Mann–Whitney U test, and Kruskal–Wallis test were used when appropriate. Univariate general linear model analysis was used to identify predictors of outcomes. R Studio (R Foundation for Statistical Computing, 2016, Vienna, Austria) with ggplot2 package was used to generate graphical representation of data.
Ten institutions participated in the study (table 1). During the 92-day study period a total of 189 patients underwent emergent angiography with intent to pursue mechanical thrombectomy for ELVO at the 10 centers.
The average number of procedures per hospital over the study period was 18.9 (average of 0.2 cases per day per center). This ranged from 0.09 cases per day at the lowest volume center to 0.49 cases per day at the highest volume center (75.6 cases per year on average, ranging from 32 to 180 cases per year). Figures 1 and 2 and table 2 demonstrate the most common time periods when mechanical thrombectomy occurred.
Procedures were more commonly initiated during daytime hours (p<0.001) and were more common during weekdays (p<0.001). However, procedures were most commonly started during non-work hours (58.7% between 17:00 hours and 07:00 hours; p<0.001). The most common periods for thrombectomy procedures were between 20:00 and 21:00 hours with a second peak between 13:00 and 14:00 hours.
Time requirement: notification to puncture
The median time from notification to groin puncture for all cases was 84 min (IQR 56–145 min; table 3). Time from notification to puncture was longer on weekends, during night-time hours, and during non-work hours. The median time from imaging completion to procedural start was 52 min longer for non-work hours than during work hours (p<0.001).
Time requirement: procedural length
The median intra-procedural length from puncture to closure for all cases was 57 min (IQR 33–80 min; table 3). There were no statistically significant associations between day of the week or time of day and procedural length. There were 11 (5.8%) procedures (from groin puncture to groin closure) that were longer than 2 hours and one (0.01%) procedure that was longer than 3 hours.
Time requirement: total
The median overall neuroendovascular call team time required for each individual patient from time of notification until groin closure was just over 2.5 hours (154 min, IQR 114–207 min; table 3). The majority of stroke events had a total time requirement of less than 3 hours (126 cases; 66.7% total). There were 36 (19%) stroke events that required 3–4 hours, 20 (10.6%) that required 4–6 hours, and 7 (3.7%) that required >6 hours. The total time requirement was longer during weekends and non-work hours than during weekdays and work hours, respectively. The median time from imaging completion or physician notification to procedural completion was 47 min longer for non-work hours than during work hours (p=0.01).
This is the first study to quantify the frequency and time requirement of mechanical thrombectomy procedures at stroke centers. A small number of the busiest stroke centers in the USA are performing mechanical thrombectomy in the range of 200 patients per year. Only two of the sites included in this study would project to treating over 100 patients per year. By design, the group of stroke centers included in this study ranged in their respective activity level, providing a real-world perspective. Over a 3-month period at these 10 stroke centers across the USA, mechanical thrombectomy procedures occurred once every 5 days on average. Almost 60% of procedures occurred outside regular work hours, with the highest frequency occurring between the hours of 20:00 hours and 21:00 hours. The median time from notification to procedural end was just over 2.5 hours, with most actual procedure times lasting less than 1 hour.
This study provides detailed information about the burden presented to the hospital by acute endovascular stroke cases. Importantly, this study intentionally does not capture the total workload of neuroendovascular physicians during stroke call coverage as it focuses specifically on the time spent performing procedures. Additional peri-procedural time required for documentation, transfer of care communications, family discussions, or commute is not included in this analysis because it is not captured in most centers. Furthermore, false positives, where neuroendovascular physicians are contacted regarding potential thrombectomy candidates but the decision is made for medical management, occur much more frequently than consultations that ultimately undergo thrombectomy procedures. False positives are not systematically recorded at most centers and were therefore not included in this retrospective analysis. Thus, the time requirement reported in this study significantly underestimates the actual time required of physicians during thrombectomy call. Prospective studies are currently planned to more completely quantify the total burden of thrombectomy call on neuroendovascular physicians and stroke centers.
The findings of this study have important operational implications regarding stroke call staff commitment during coverage. Endovascular stroke procedures are resource intensive, requiring on-call stroke and endovascular physicians and also nurses and technologists, and often anaesthesia teams. Centers performing these procedures should account for the need for frequent evening, night-time, or weekend coverage by staff. This is particularly important in light of paradigm shifts in payment policy that will certainly impact facility and possibly physician fees for treatment of ischemic stroke.13–15 Mechanisms should be in place for appropriate compensation of staff, but an additional important consideration is the risk of fatigue or provider ‘burnout’ should night-time call be too burdensome or intrusive for employees. This is particularly true if they are expected to work an additional daytime shift the day following a call night.
This study found no difference in procedural length based on day of the week or time of day. This is consistent with a previous study that suggested the absence of a weekend effect on stroke procedural times.16 However, the present study did suggest a delay of approximately 50 min from notification to puncture during non-work hours compared with work hours. The longer time from notification to groin puncture during non-work hours is probably largely explained by the necessity for staff and physicians to commute to the hospital from home when the call team is activated. Attempts at further reductions in door-to-puncture times will require consideration of these night-time delays and methods to mitigate this, such as parallel activation paradigms.
There are other important limitations to this study. All centers included are primary or comprehensive academic stroke centers with robust stroke triage systems of care in place. These data may therefore not reflect volumes or procedural times at private hospitals or community hospitals. Despite being high efficiency centers, the process per center is non-uniform and this variance may affect some of the average times. However, this circumstance is likely to reflect real-life variations in practice logistics. Additionally, the data extended over only a 3-month time period, an intentional decision to capture only the most recent data available given the current pace of change in stroke care. Finally, the busiest stroke centers in the USA were not included in this pilot study as the intent was to provide information that was broadly applicable and in particular useful for institutions that are considering expansion of their embolectomy practice.
This is the first study to quantify the frequency and time of day of mechanical thrombectomy procedures at stroke centers. Over a 3-month period at 10 stroke centers across the USA, mechanical thrombectomy procedures occurred once every 5 days on average, although the majority occurred during non-work hours, peaking between 20:00 hours and 21:00 hours. Our findings indicate that the majority of cases occur off-hours, with important operational implications for hospitals implementing stroke call coverage. Future prospective studies will evaluate the total call burden more completely, including the staff burden incurred by the high rate of cases evaluated that do not undergo treatment.
Contributor Conception and design: KF. Data collection and interpretation: All authors. Statistical analysis: TW. Drafting the article: TW, TL, JH, KF. Critical revision of article: All authors. Final approval of article: All authors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.