Developing a fast-track discharge protocol for patients with cerebral aneurysms treated via neuroendovascular techniques

Abstract

Introduction

As we emerge from the current pandemic, hospitals, staff, and resources will need to continue to adjust to meet ongoing healthcare demands. Lessons learned during past shortages can be used to optimize peri-procedural protocols to safely improve the utilization of hospital resources.

Methods

Retrospective review of patients who underwent elective endovascular intracranial aneurysm treatment was performed. Multivariable logistic regression was used to identify factors associated with patients who were able to be discharged within 24 h of elective procedures. Rates of complications (particularly readmission) were determined.

Results

330 patients underwent elective endovascular aneurysm treatment with 86 (26.1%) discharged within 24 h. Factors associated with earlier discharge included procedure years (2019–2021) and male sex. Patients were more likely to be discharged later (after 24 h) if they underwent stent-coil embolization or flow-diversion. There was no association between discharge timing and likelihood of readmission.

Discussion

Our review highlights the safety of earlier discharge and allowed us to prepare a fast-track protocol for same-day discharge in these patients. This protocol will be studied prospectively in the next phase of this study. As we gain more comfort with emerging, minimally invasive endovascular therapies, we hope to safely achieve same-day discharge on a protocolized and routine basis, reducing the demand of elective aneurysm treatments on our healthcare system.

Conclusion

We retrospectively demonstrate that early discharge following elective aneurysm treatment is safe in our cohort and provide a fast-track pathway based on these findings for other centers developing similar protocols.

Introduction

The COVID-19 pandemic challenged hospital systems to maintain performance benchmarks (e.g. patient care outcomes, financial productivity, etc.) in an era of added pressures and staffing shortages. ¹ While elective surgery waiting times increased, the need for these procedures has not wavered.^2–5 The progressive scarcity of available hospital beds for elective, outpatient admissions led to numerous cutbacks in the number of elective surgeries, which increased pressure to discharge patients earlier, amongst other restructuring and administrative efforts. ⁶

Similar to other subspecialties, neuroendovascular referrals and procedures demonstrated a steep decline during the pandemic.^7–9 The complexity of decision-making and treatment in addressing unruptured intracranial aneurysms was further burdened by the limitations in hospital resources, which required cerebrovascular specialists to consider alternative approaches for providing patient care.^8,10,11 Preparing for future waves and augmenting our healthcare strategies will help ensure success in addressing the backlog of elective cases. ¹²

Expedited recovery or fast-track discharge protocols (sometimes referred to as Enhanced Recovery After Surgery (ERAS)) have shifted the traditional algorithms of care for surgical patients in other subspecialties.^13–16 These multimodal perioperative pathways are often designed by collaborative teams (primarily led by the surgeons and anesthesiologists) with the aim for quicker recovery following invasive procedures, often leading to earlier discharges and improved outcomes. At the heart of these protocols is the optimized care of patients both pre- and post-operatively.

In this study, we conducted a retrospective review of all patients who underwent elective endovascular treatment for intracranial aneurysms at a comprehensive stroke center by four neuroendovascular neurosurgeons from 2015 to 2021. We aimed to evaluate trends in neuroendovascular practice patterns over this period to better identify and understand factors contributing to safe discharge within one post-operative day using a fast-track discharge protocol specific for elective aneurysm treatment. Identifying the patient group who may be most amendable to inclusion in such a protocol is the first step to achieving early discharge for elective aneurysm embolizations.

Methods

Retrospective, deidentified review from medical records collected data on all patients diagnosed with unruptured intracranial aneurysms who were treated electively with endovascular therapy at a comprehensive stroke center between 2015 and 2021. This study was approved by the Institutional Review Board.

All procedures were performed within a biplane angiography suite with endovascular embolization utilizing coil, stent-coil, flow-diverting stent, or Woven EndoBridge (WEB) under conscious sedation or general anesthesia. Our primary outcome was length of stay (LOS), calculated as the time from procedure check-in to the time of discharge. This was further classified into two separate groups (1: LOS <24 h; 2: LOS <24 h). Our secondary outcomes were adverse events. Any complications including those from procedure-related access (e.g. hematoma, pseudoaneurysm, dissection), stroke, neurologic deficit, and intracranial hemorrhage for the first 30 days post-procedure were captured. Readmissions during this period (including emergency department visits) were carefully noted.

Further procedural information was captured, including duration of each embolization as measured in total fluoroscopy time (in hours). Presence of anti-platelets/anti-coagulants before and after procedures were considered. Year of treatment was recorded as well. Patient factors including age, sex, medical comorbidities, aneurysm location, and previous aneurysm treatment were documented. Medical comorbidities were classified according to a modified version of the Charlson Comorbidity Index (mCCI), ¹⁷ which was stratified as mild (0-2), moderate (3-4), and severe (>5).

Descriptive statistics were considered based on LOS. Comparisons across LOS groups were made using chi-squared tests. Multivariable logistic regression analysis was performed to assess the relationship between predictor variables and LOS. All statistical analyses were conducted using Stata (version 17.0, Stata Corp LP, College Station, Texas, USA).

Results

A total of 330 patients were treated electively with endovascular embolization for unruptured intracranial aneurysms during years 2015 to 2021 (Table 1). Of these, there were 86 (26.1%) patients discharged within <24 h and 244 (73.9%) patients discharged >24 h. In the total cohort, most patients were female (75.2%) with a mean age of 57.9 ± 12.4 years. There was no significant difference in LOS based on aneurysm location (p = 0.15). Cases were more likely to be performed under general anesthesia (87.3%). There were 43 (13.0%) patients who underwent treatment for recurrence or residual aneurysm. Overall, there were 73 (22.1%) patients treated with coiling alone, 78 (23.6%) with stent-coiling, 159 (48.2%) with flow-diverting stent placement, and 20 (6.1%) cases treated with WEB embolization. A total of 12 (3.6%) cases were performed by radial access.

Table 1.

Sample characteristics. Significant p-values are noted with an asterisk (*). Percentages are calculated within columns, representing sub-group contributions based on LOS. DAPT = Dual antiplatelet therapy; LOS = Length of stay.

	Overall (n = 330)	<24 Hours (n = 86)	>24 Hours (n = 244)	p-value
YEAR
2015	54 (16.4%)	7 (8.1%)	47 (19.3%)	<0.001*
2016	39 (11.6%)	3 (3.5%)	36 (14.8%)
2017	35 (10.6%)	4 (4.7%)	31 (12.7%)
2018	42 (12.7%)	6 (7%)	36 (14.8%)
2019	60 (18.2%)	25 (29.1%)	35 (14.3%)
2020	50 (15.2%)	19 (22.1%)	31 (12.7%)
2021	50 (15.2%)	22 (25.6%)	28 (11.5%)
AGE
<40	31 (9.4%)	9 (10.5%)	22 (9%)	0.95
40–50	47 (14.2%)	12 (14%)	35 (14.3%)
50–60	114 (34.6%)	32 (37.2%)	82 (33.6%)
60–70	83 (25.2%)	20 (23.3%)	63 (25.8%)
>70	55 (16.7%)	13 (15.1%)	42 (17.2%)
SEX
Female	248 (75.2%)	59 (68.6%)	189 (77.5%)	0.10
mCCI SCORE
Mild (0–2)	206 (62.4%)	57 (66.3%)	149 (61%)	0.38
Moderate (3-4)	99 (30%)	21 (24.4%)	78 (32%)
Severe (>5)	25 (7.6%)	8 (9.3%)	17 (7%)
ANEURYSM LOCATION
Anterior Circulation	271 (82.1%)	70 (81.4%)	201 (82.4%)	0.84
Posterior Circulation	59 (17.6%)	16 (18.6%)	43 (17.6%)
SEDATION TYPE
Conscious Sedation	42 (12.7%)	6 (7%)	36 (14.7%)	0.06
General Anesthesia	288 (87.3%)	80 (93%)	208 (85.3%)
PROCEDURE DURATION
<0.58 Hours	84 (25.5%)	24 (27.9%)	60 (24.6%)	0.07
0.58–0.87 Hours	81 (24.6%)	28 (32.6%)	53 (21.7%)
0.87–1.28 Hours	83 (25.2%)	20 (23.3%)	63 (25.8%)
>1.28 Hours	82 (24.9%)	14 (16.3%)	68 (27.9%)
TREATMENT TYPE
Coil	73 (22.1%)	22 (25.6%)	51 (20.9%)	0.007*
Stent-Coil	78 (23.6%)	14 (16.3%)	64 (26.2%)
Flow Diversion	159 (48.2%)	39 (45.4%)	120 (49.2%)
WEB	20 (6.1%)	11 (12.8%)	9 (3.7%)
Prior treatment	43 (13.0%)	12 (14.0%)	31 (12.7%)	0.77
ACCESS SITE
Femoral access	318 (96.4%)	79 (91.9%)	239 (98%)	0.009*
Radial access	12 (3.6%)	7 (8.1%)	5 (2%)
PRE-TREATMENT ANTI-PLATELET REGIMEN
None	15 (4.5%)	2 (2.3%)	13 (5.3%)	0.75
Aspirin	37 (11.2%)	11 (12.8%)	26 (10.7%)
Aspirin/Ticagrelor	99 (30%)	26 (30.2%)	73 (29.9%)
Aspirin/Clopidogrel	155 (47%)	42 (48.8%)	113 (46.3%)
Other	24 (7.3%)	5 (5.8%)	19 (7.8%)
POST-TREATMENT MEDICATION ADJUSTMENT
No change	298 (90.3%)	77 (89.5%)	221 (90.6%)	0.83
Reduced to monotherapy	19 (5.8%)	6 (7%)	13 (5.3%)
Change to new medication	13 (3.9%)	3 (3.5%)	10 (4.1%)
COMPLICATIONS
Access site complication	9	1	8	0.3
30-Day Readmission	20	5	15	0.91

There were significant differences in the LOS based on treatment type. Patients with aneurysms treated with stent-coil embolization (OR: 3.55, 95% CI: 1.16-10.84, p = 0.03) and flow-diversion (OR: 3.61, 95% CI: 1.31-9.35, p = 0.01) had a higher likelihood of being discharged >24 h following the procedure when compared to those treated with coil embolization. Treatment with WEB embolization did not predict longer LOS compared to coil embolization alone (OR: 1.10, 95% CI: 0.30-3.95, p = 0.98).

Results from multivariable logistic regression analysis are presented in Table 2. The year of treatment significantly predicted shorter LOS. Patients were more likely to be discharged within <24 h for procedures performed during years: 2019 (OR: 0.08, 95% CI: 0.03-0.28, p < 0.001), 2020 (OR: 0.13, 95% CI: 0.04-0.42, p = 0.001), and 2021 (OR: 0.06, 95% CI: 0.02-0.24, p < 0.001) when compared to year 2015.

Table 2.

Multivariate analysis. Significant p-values are noted with an asterisk (*). CI = Confidence interval; LOS = Length of stay; mCCI = modified Charlson comorbidity index; OR = Odds radio; WEB = Woven EndoBridge intrasaccular device.

	OR (95% CI)	p-value
YEAR
2015	1.00 (Reference)	-
2016	1.30 (0.28–5.98)	0.73
2017	0.63 (0.14–2.78)	0.54
2018	0.38 (0.10–1.48)	0.16
2019	0.08 (0.03–0.28)	<0.001
2020	0.13 (0.04–0.42)	0.001
2021	0.06 (0.02–0.24)	<0.001
AGE
<40	1.00 (Reference)	-
40–50	1.11 (0.33–3.69)	0.87
50–60	0.95 (0.33–2.79)	0.93
60–70	1.57 (0.47–5.24)	0.46
>70	1.49 (0.34–6.56)	0.6
SEX
Male	0.50 (0.25–0.97)	0.04
mCCI SCORE
Mild (0–2)	1.00 (Reference)	-
Moderate (3-4)	0.97 (0.42–2.27)	0.68
Severe (>5)	0.87 (0.23–3.24)	0.84
ANEURYSM LOCATION
Anterior Circulation	1.00 (Reference)	-
Posterior Circulation	0.84 (0.37–1.90)	0.68
SEDATION TYPE
Conscious Sedation	1.00 (Reference)	-
General Anesthesia	0.79 (0.27–2.37)	0.68
PROCEDURE DURATION
<0.58 Hours	1.00 (Reference)	-
0.58–0.87 Hours	0.62 (0.28–1.36)	0.24
0.87–1.28 Hours	0.89 (0.37–2.11)	0.79
>1.28 Hours	1.18 (0.47–2.96)	0.73
TREATMENT TYPE
Coil	1.00 (Reference)	-
Stent-Coil	3.55 (1.16–10.84)	0.03
Flow Diversion	3.61 (1.34–9.69)	0.01
WEB	1.10 (0.30–3.95)	0.89
Prior treatment	0.50 (0.20–1.22)	0.13
ACCESS SITE
Femoral access	1.00 (Reference)	-
Radial access	0.46 (0.11–1.84)	0.27
PRE-TREATMENT ANTI-PLATELET REGIMEN
None	1.00 (Reference)	-
Aspirin	0.96 (0.14–6.75)	0.97
Aspirin/Ticagrelor	0.52 (0.07–3.68)	0.51
Aspirin/Clopidogrel	0.28 (0.04–1.91)	0.19
Other	0.80 (0.09–7.16)	0.84
POST-TREATMENT MEDICATION ADJUSTMENT
No change	1.00 (Reference)	-
Reduced to monotherapy	0.83 (0.20–3.49)	0.8
Change to new medication	0.42 (0.08–2.25)	0.31
COMPLICATIONS
Access site complication	2.96 (0.27–32.68)	0.38
Other complication	6.12 (1.38–27.07)	0.02

Male sex had 50% lower odds (OR: 0.50, 95% CI: 0.25-0.97, p = 0.04) of being discharged >24 h when compared to females. Age, mCCI, and aneurysm location did not predict longer LOS. Procedures performed under general anesthesia (OR: 0.98, 95% CI: 0.32-3.01, p = 0.98) and overall duration of procedure did not contribute to longer LOS, though the latter did trend towards significance.

There were 25 (7.6%) acute procedure-related complications, which did not differ significantly by LOS group (p = 0.17). Access site complications were observed in 9 patients (2.7%), transient neurologic deficit in 8 (2.4%) patients (all resolved prior to discharge; none with LOS <24 h), 3 patients with embolic ischemic strokes, 1 patient with bradycardia requiring pacemaker placement post-procedurally, 1 traumatic foley insertion, and 1 respiratory infection requiring antibiotics. There was one mortality from subacute (<4 h peri-procedurally) intraparenchymal hemorrhage following flow-diverting stent placement. Another developed a posterior-fossa and brainstem infarct on post-op day 2 following placement of flow-diverting stents for treatment of a giant fusiform vertebrobasilar aneurysm. The patient's hospital course was complicated by access site issues and wound breakdown following vascular surgery intervention. He was ultimately sent to hospice where he expired.

A total of 20 (6.1%) patients were readmitted within 30-days of discharge for any reason; 5 (5.8%) in the <24 h LOS group and 15 (6.1%) in the >24 h LOS group (p = 0.91). In the <24 h LOS group, there were 2 readmissions for GI bleed (both on dual antiplatelet therapy), 1 patient with a delayed intraparenchymal hemorrhage remote to the aneurysm, 1 patient who was readmitted for additional coil placement, and 1 who was involved in an MVA 2 weeks after discharge. In the >24 h LOS group, there were 5 cases of stent occlusion after discharge (4 who improved after angiographic revascularization and 1 with a complicated hospital course and mRS 5 at discharge). Additional reasons for readmissions in this group included transient confusion (2), delayed thalamic hemorrhage in a clopidogrel hyperresponder (1), embolic infarct (1), seizure (1), transient gait difficulty (1), pneumonia (1), mechanical fall (1), transient headache (1), and exacerbation of trigeminal neuralgia symptoms (1). Including these readmissions within 30 days as a complication, there were 45 total patients with complications (13.6%)

The type of anti-platelet or anti-coagulation therapy prior to treatment did not influence the overall LOS. Adjusting a patient's anti-platelet or anti-coagulation therapy following the procedure did not contribute to longer LOS in multivariable analysis. Access site complications peri-procedurally did not show significantly increased odds of LOS >24 h (OR: 2.96, 95% CI: 0.27-32.68, p = 0.38). Lastly, non-access site related complications had significantly higher likelihood (OR: 6.12, 95% CI: 1.38-27.07, p = 0.02) of longer LOS. Additional multivariable analysis was conducted to determine if there were predictors of any complication. This analysis demonstrated increased likelihood of complications during years 2016 (OR: 6.29, 95% CI: 1.11-35.58, p = 0.04), 2019 (OR: 3.77, 95% CI: 0.61-23.37, p = 0.01), and 2021 (OR: 19.17, 95% CI: 3.41-107.84, p = 0.001), but all other predictors were non-significant.

Discussion

Delays in the care of cerebral aneurysms due to various shortages have had a profound impact on our patients. In one study from Poland, the rate of mortality from cerebral aneurysms increased by 21% during the early pandemic stages. ⁹ Improving our peri-procedural strategem regarding cerebral aneurysms is critical in an era of decreasing hospital resources and bed availability. Same-day discharge after neuroangiography has been well-described as both safe and cost-minimizing.^18–21 As similar advances are made in the microsurgical aneurysm repair cohorts,^11,22,23 the neuroendovascular realm should seek to improve our discharge protocols as well. ²⁴

Early discharge did not increase the rate of readmission nor the overall complication rate. The two mortalities observed in this cohort and all 5 cases of in-stent thrombosis occurred in the later discharge group. This likely reflects the importance of physician discretion in ensuring safe discharge in fast-track patients (i.e. the surgeon is unlikely to discharge a patient early if they are concerned for complications or if the pathology/procedure warrants additional observation). Overall, this retrospective review demonstrates that early discharge did not increase likelihood of readmission or complications in our cohort.

The progressively decreasing LOS witnessed between 2018 to 2021 is likely a result of improving technology and techniques, in addition to our familiarity with the post-operative care of these patients. This is partly in combination with the aforementioned added hospital pressures to discharge patients early during the pandemic; though this effect begins in 2019, one year prior. Despite increasing discharge speeds, complications and hospital readmissions within 30 days did not increase when comparing the early (<24) and later (>24) discharge groups.

Previous studies of unruptured intracranial aneurysm repair have demonstrated a 90-day readmission rate of 7.34%. ²⁵ Our rate of 6.1% is slightly lower but did not differ significantly between the shorter LOS (5.8%) and longer LOS (6.2%) groups. While access site complications did not differ significantly between those discharged before and after 24 h of observation, non-access site complications inevitably resulted in longer LOS which is expected. For future fast-track protocols for endovascular aneurysm embolization, patients who have acute complications should not be eligible for early discharge.

Of note, we did notice a difference in LOS based on treatment type, namely that patients who underwent stent-coiling or flow-diversion remained in the hospital longer, alluding to the added treatment complexity of these aneurysms. This finding is likely a surrogate for other factors that delayed discharge (e.g. post-procedure heparinization, activity not advanced, etc.) or surgeon preference, following more complicated procedures. As we develop expedited discharge protocols for our endovascular patients, we should either (1) exclude stent-coil or flow-diversion patients from protocolization or (2) monitor their outcomes from earlier discharge more stringently during prospective studies, in order to better assess the safety of earlier discharge for these patients going forward.

Prior authors have demonstrated the safety of same-day discharge with embolization using flow-diversion. ²⁶ However, our study demonstrated longer LOS for those treated with flow-diversion when compared to coil embolization. As we gain more confidence in newer technologies, achieving same-day discharge may become a reality for more straightforward flow-diversion cases. Furthermore, utilizing the radial approach for more interventions may help reduce access site issues for properly-selected patients and further improve discharge timing (as seen in our cohort).^27–30 In the current study, radial access trended towards shorter LOS, but did not demonstrate significance, though limited by sample size.

Changes to anti-platelet or anti-coagulation therapy occurred often due to alterations in planned therapy (e.g. stent-assisted coiling was planned for pre-operatively but no stent was needed and so anti-platelet therapy was stopped, an un-planned stent was required intra-procedurally, etc.). Alternatively, we routinely test for responsiveness to clopidogrel using P2Y12 platelet reactivity unit assays, which if patients are found to be subtherapeutic requires medication adjustment as well. In either scenario, adjustments to medication regimen did not affect our discharge timing.

In their review of same-day discharge after aneurysm embolization using flow-diversion, Zanaty et al. proposed pre- and post-operative checklists to help improve the safety of same-day discharge following aneurysm treatment. We modified their checklist based on our results (Table 3), de-emphasizing the role of anti-platelet and/or anticoagulant medications as our analysis did not demonstrate any effect from medications on our discharge practices (though this will be tracked prospectively in future studies). Our results lay the groundwork for the development of a fast-track protocol for patients undergoing neuroendovascular aneurysm treatment, which we have modified from a current institutional protocol used in spine patients (Supplement 1).

Table 3.

Checklist for patients eligible for same-day discharge, modified from zanaty et al. 2016 based on our results.

DISCHARGE CHECKLIST
No allergy to contrast
No neurological changes from baseline following embolization.
No evidence of access site complications.
No acute intra- or peri-procedural (e.g. distal embolic, hemorrhagic) complications.
Patient lives less than 2 h away.
Patient is independent.
Patient has family support at home.
Monitored for a minimum of 4 h post-procedurally.
No discharges between 9:00pm and 7:00am for same-day discharge.
Patient received appropriate pre-procedural counseling on discharge planning and is in agreement.
Physician discretion post-procedurally based on findings.

A source of bias is the increasing pressures for early discharge imposed by the pandemic and limitations on hospital resources, which was found to be significant in our analysis. However, this allowed us to study the safety of early discharge by providing a larger cohort of patients discharged earlier. This study is further limited by its retrospective nature and the constraints of data found during chart review. Our hope is that these data will serve as a model for other centers that wish to develop a fast-track discharge or ERAS model for their own cerebrovascular procedures and to provide a framework for prospective analyses of such protocols. Future endeavors will be to track cerebral aneurysm patient discharged through this discharge protocol in a prospective manner, not only to validate the safety of such a protocol but to demonstrate any potential benefits in patient outcomes and healthcare costs.

Conclusions

Learning lessons from the challenges posed to us during the viral pandemic – such as resource shortages – are important in maintaining hospital efficiency while simultaneously promoting patient safety. This study demonstrates that patients discharged within 24 h did not suffer from increased complications nor readmissions. Our data further suggest that developing fast-track discharge protocols, with the ultimate goal of discharging patients after a 4–6 h recovery period, is feasible and likely safe for our neuroendovascular patients. The second stage of this study will be to begin using these protocols and prospectively track these patients across multiple enlisted institutions.