Objective To investigate whether significant differences exist in hospital bills and patient outcomes between patients who undergo endovascular thrombectomy (EVT) alone and those who undergo EVT with pretreatment intravenous tissue plasminogen activator (IV tPA).
Methods We retrospectively grouped patients in an EVT database into those who underwent EVT alone and those who underwent EVT with pretreatment IV tPA (EVT+IV tPA). Hospital encounter charges (obtained via the hospital’s charge capture process), final patient bills (ie, negotiated final bills as per insurance/Medicare rates), demographic information, existing comorbidities, admission and discharge National Institutes of Health Stroke Scale (NIHSS) score, and functional independence data (modified Rankin Scale score 0–2) were collected. Univariate and multivariate statistical analyses were performed.
Results Of a total of 254 patients, 96 (37.8%) underwent EVT+IV tPA. Median NIHSS score at admission was significantly higher in the EVT+IV tPA group than in the EVT group (p=0.006). After adjusting for NIHSS admission score, patient bills and encounter charges in the EVT+IV tPA group were still found to be $3861.64 (95% CI $658.84 to $7064.45, p=0.02) and $158 071.29 (95% CI $134 641.50 to $181 501.08, p < 0.001) greater than in the EVT only group respectively. The EVT+IV tPA group had a higher complication rate of intracranial hemorrhage (ICH) (p=0.005). The EVT and EVT+IV tPA groups did not differ significantly in median discharge NIHSS score (p=0.56), functional independence rate at 90 days (p=0.96), or average length of hospital stay (p=0.21).
Conclusion Patients treated with EVT+IV tPA have greater hospital encounter charges and final hospital bills as well as higher rates of ICH than patients who undergo treatment with EVT only.
Statistics from Altmetric.com
Endovascular thrombectomy has been shown to be a safe and efficacious treatment that improves patients’ clinical outcomes and adds adjunctive benefit to intravenous (IV) thrombolysis.1–5 However, the efficacy of pre-thrombectomy IV thrombolytics (IV tissue plasminogen activator (tPA)) for patients already scheduled to receive endovascular treatment (EVT) has been a recent matter of contention. Hypothetically, IV tPA administration may increase rates of early recanalization (ie, recanalization before EVT treatment) as well as reduce time to recanalization with thrombectomy, thereby decreasing the chance for endovascular complications. Several studies have found similar rates of good outcome with EVT alone and EVT+IV tPA,6 7 while other research has shown that pre-thrombectomy IV tPA increases the rates of both symptomatic and asymptomatic intracranial hemorrhage (ICH)8 as well as placing procedural restraints on the use of certain other medications such as antiplatelet drugs during stenting procedures.9
Given that endovascular thrombectomy is the present standard of care for stroke patients with emergent large vessel occlusions (ELVO)1–5 10 and the rapidly increasing rates of mechanical thrombectomy procedures,11 both the clinical efficacy and potential financial costs of pretreatment IV tPA administration need to be reassessed.
This study evaluates the differences, if any, between hospital bills, encounter charges, as well as the rates of good patient outcome and safety outcome between patients who underwent EVT alone and patients who underwent EVT in conjunction with pre-thrombectomy IV tPA.
In this retrospective analysis of a prospectively collected and maintained high-volume single-center endovascular database, patients’ encounter charges and hospital bill data between January 2012 and August 2018 were collected from the hospital finances administration and the chief financial officer’s records. Our selection criteria for endovascular treatment include imaging and clinical criteria. Selected patients have a minimum National Institutes of Health Stroke Scale (NIHSS) score of 6 and ELVO on CT angiogram. For patients last known well >6 hours or wake up strokes, they also had CT perfusion with evidence of penumbra, cerebral blood volume <50 mL, and a mismatch ratio of >1.2. Patients who had a modified Rankin Scale (mRS) score <3 had the risks and benefits of EVT discussed with their family. All patients underwent endovascular thrombectomy and were divided into two groups based on whether they received IV tPA prior to EVT (EVT+IV tPA group) or received the EVT alone (EVT group).
Baseline variables and clinical, radiographic, and safety outcome rates were also included. Baseline variables included age, gender, ethnicity, admission NIHSS score, time from last seen well to presentation, anticoagulation and antiplatelet rates prior to the stroke, and patients’ past medical history of hypertension, atrial fibrillation, coronary artery disease, hyperlipidemia, smoking status, diabetes mellitus, and a history of stroke/transient ischemic attack. Safety outcomes included rates of hemorrhagic transformation, mass effect, symptomatic ICH, and asymptomatic ICH. All patients with symptomatic ICH displayed a 4+ point NIHSS score worsening within 24 hours post-procedure. Clinical outcomes included discharge and 90-day mRS scores, 24 hours post-procedure and discharge NIHSS scores, length of hospital stay, and mortality rates. The radiographic outcome was the rate of successful reperfusion (modified Thrombolysis in Cerebral Infarction (mTICI) ≥2b). Good clinical outcomes were defined as mRS scores of 0–2 at 90 days. Discharge mRS was substituted for patients with missing 90-day follow-up mRS scores (n=5). The study was approved by the local Institutional Review Board.
Continuous variables were reported as mean±SD while categorical variables were reported as median, frequency, and/or range. Initially a univariate analysis was carried out between the EVT and EVT+IV tPA groups for each variable to obtain p values. While continuous variables were analyzed with the independent samples t-test, ordinal categorical variables were analyzed using the Mann–Whitney U test and binary categorical variables were analyzed using the χ2 test. Variables with p values <0.05 significance level were then analyzed further with ANCOVA and bivariate logistic regression. ANCOVA was used for developing adjusted means in encounter charges and hospital bills while bivariate logistic regression was used to produce odds ratios and confidence intervals for safety outcome and clinical outcome rates. All statistical analyses were conducted with IBM SPSS.
Of the 254 patients who underwent EVT, 96 (37.8%) received both EVT and IV tPA. The average age of all patients was 70.7±12.9 years. Patients had an average time to presentation of 3.6±3.0 hours while their average NIHSS score on admission had a median value of 17 (range 0–36). The average encounter charge was $275 658.25±153 935.23 while the average hospital bill was $28 545.10±10 442.56. Successful recanalization (mTICI 2b/3) was obtained in 229 (90.2%) patients while rates of good clinical outcomes (mRS 0–2) were found in 93 (36.6%) patients. The patient mortality rate was 29.9% up to 90 days. Seventy-nine patients (31.1%) were found to have ICH, 64 (81.0%) of which were asymptomatic. A proportion of patients (approximately 10–15%) had a baseline mRS score of 2–3 which would not feasibly improve even with successful recanalization.
Univariate analysis of hospital bills, encounter charges, baseline patient data, safety outcome rates, and patient outcomes between the EVT and EVT+IV tPA groups are shown in tables 1 and 2. The mean±SD encounter charges for the EVT and EVT+IV tPA groups were $227 825.52±99 026.33 and $355 213.29±191 092.40, respectively. EVT+IV tPA encounter charges were $127 388.24 (95% CI $91 458.59 to $163 317.41, p<0.0001) greater than for the EVT group. The mean±SD hospital bills for the EVT and EVT+IV tPA groups were $27 057.00±9690.84 and $31 002.55±11 582.86, respectively. EVT+IV tPA hospital bills were $3945.44 (95% CI $1283.63 to $6607.25, p=0.004) greater than for the EVT group. The only significant baseline variable found was the admission NIHSS score, with a median value of 16 in the EVT group and 19 in the EVT+IV tPA group (p=0.006). None of the baseline comorbid conditions showed any statistically significant proportional differences between the EVT group and the EVT+IV tPA group. Significant safety outcome rates between the EVT group and EVT+IV tPA group included hemorrhagic transformation (p=0.05), mass effect (p=0.04), and asymptomatic ICH (p=0.02), all of which were proportionally greater in the EVT+IV tPA group. Clinical outcomes including recanalization rate (p=0.85), NIHSS at 24 hours (p=0.25), NIHSS at discharge (p=0.56), functional independence up to 90 days (p=0.96), and mortality rate (p=0.15) failed to show any significant proportional difference between the EVT and EVT+IV tPA groups.
With respect to the admission NIHSS score (controlling for EVT+IV tPA group status), significant odds ratios included 1.04 (95% CI 1.008 to 1.091, p=0.017) for hemorrhagic transformation and 1.05 (95% CI 1.014 to 1.093, p=0.008) for asymptomatic ICH. With respect to EVT+IV tPA (controlling for NIHSS score), the only significant OR regarding safety outcome data was 1.88 (95% CI 1.023 to 3.446, p=0.042) for asymptomatic ICH. With respect to EVT+IV tPA (controlling for NIHSS score), no significant differences in clinical outcomes were found (ie, rate of recanalization (p=0.872), rate of functional independence (p=0.424), NIHSS score at 24 hours (p=0.613), NIHSS score at discharge (p=0.844), mortality rate (p=0.203), and length of hospital stay (p=0.309)).
ANCOVA was performed to yield adjusted mean hospital bills and encounter charges for both the EVT and EVT+IV tPA groups (table 3) while bivariate logistic regression was performed using EVT+IV tPA group status and significant predictor variables found during univariate analysis to develop odds ratios for rates of hemorrhagic transformation, mass effect, and asymptomatic ICH (table 4). The ANCOVA test as well as all of the bivariate logistical analyses used group status (ie, EVT vs EVT+IV tPA) and admission NIHSS scores as predictor variables. After adjusting for NIHSS score, the adjusted EVT+IV tPA mean hospital bill was found to be $3861.64 (95% CI $658.84 to $7064.45, p=0.02) greater than the EVT hospital bill, while the EVT+IV tPA mean encounter charge was found to be $158 071.29 (95% CI $134 641.50 to $181 501.08, p<0.001) greater than the EVT encounter charge.
This study has shown that patients undergoing EVT with pretreatment IV tPA have larger hospital bills and encounter charges than those who undergo EVT only. However, it is important to note that these costs typically comprise many different components. One of the traditionally large components of these costs—namely, length of hospital stay—was not, however, found to be significantly different between the EVT and EVT+IV tPA groups (p=0.21). Rai et al, performing a similar EVT vs EVT+IV tPA efficacy study, found that length of hospital stay, IV tPA+EVT treatment group status, as well as rates of recanalization were independent predictors of higher total and direct costs.12 After finding that differences in recanalization rates and length of hospital stay were negligible between EVT and EVT+IV tPA groups, Rai et al similarly concluded that the variable most associated with significant differences between direct costs for EVT and EVT +IV tPA groups was pretreatment with IV tPA.12
Our study found that clinical outcomes—namely, functional independence up to 90 days (mRS 0–2), NIHSS at 24 hours, NIHSS at discharge, and mortality rate— were not significantly different between the two groups, seeming to provide evidence that EVT+IV tPA is associated with lower costs than EVT alone. Several meta-analyses which found negligible differences between EVT and EVT+IV tPA functional independence and mortality rates up to 90 days substantiate these initial findings.13 14 Several benefit-cost studies regarding EVT alone versus IV tPA alone have similarly shown EVT alone to be more cost effective, particularly in large vessel occlusions.15 16
The rates of several safety outcomes—namely, asymptomatic ICH, hemorrhagic transformation, and mass effect—were found to be significantly higher in the EVT+IV tPA group. However, these differences were more attributable to the higher admission NIHSS scores in the EVT+IV tPA group than to EVT+IV tPA group status alone. EVT+IV tPA status was only significantly associated with the incidence of asymptomatic ICH (OR 1.88, 95% CI 1.023 to 3.446; p=0.042). Broeg-Morvay et al found similar results when comparing patients undergoing mechanical thrombectomy against patients undergoing mechanical thrombectomy with IV tPA. They found that the mechanical thrombectomy treated patient group had significantly lower rates of asymptomatic ICH (p=0.023) and also no significant differences in functional independence post-procedure compared with patients undergoing mechanical thrombectomy with IV tPA.17 It should be noted, however, that while the rates of asymptomatic ICH have indeed been shown to be higher in EVT compared with other reperfusion therapies,18 whether asymptomatic ICH, as its name suggests, definitively has an insignificant association with functional outcome is still a matter of contention.19
Although the EVT+IV tPA patient group in our study had arrived with significantly larger admission NIHSS scores than the EVT patient group, after controlling for admission NIHSS score we found no significant correlation between receiving pre-thrombectomy IV tPA and improved clinical outcome rates in our study. While our study did not find significant variation in clinical outcomes between the EVT and EVT+IV tPA patient groups, Tsivgoulis et al found on meta-analysis of six randomized clinical trials (n=1916 patients) that patients undergoing EVT alone had an increased likelihood (OR 2.47, 95% CI 1.32 to 4.63; p=0.001) of having a higher rate of functional independence at 3 months than patients who had EVT+IV tPA (OR 1.83, 95% CI 1.37 to 2.44; p<0.001).20 Such findings illustrate the necessity of reassessing pre-thrombectomy IV tPA administration guidelines.
Time to presentation has often been correlated with increased rates of functional independence and other positive clinical outcomes. Although the arrival time for patients undergoing EVT was found to be approximately 1.21 hours longer than that of the EVT+IV tPA group, our study did not find this variable to be statistically significant enough to be associated with either EVT or EVT+IV tPA group status (p=0.08). Rai et al, on the other hand, reported a significantly increased time to presentation for EVT patients compared with EVT+IV tPA patients (5:19±4:30 hours and 1:46±0:52 hours, respectively) but found that, on generously controlling for time to presentation (ie, selecting only those <4.5 hours from both EVT and EVT+IV tPA groups), patients in the EVT and EVT+IV tPA groups had insignificant differences in rates of recanalization and favorable outcome. Whether the clinical efficacy of EVT compared with EVT+IV tPA somehow ‘offsets’ the negative effects of the increased time to presentation in the EVT group found in both of these studies warrants further investigation.
The most important limitation in this study is its retrospective nature. While our study supports the finding that EVT is associated with lower costs than EVT+IV tPA, hospital bills and encounter charges are multifactorial and vary across hospitals. The fact that the EVT+IV tPA group had larger admission NIHSS scores which in turn contributed largely to safety outcome rates was a finding that required rigorous controlling. Furthermore, the ubiquity of our finding that an association exists between higher admission NIHSS score and pre-thrombectomy IV tPA administration is not definitively known. Further studies investigating the ubiquity of this association are warranted. Larger multicenter prospective studies would be necessary to corroborate the results of our study.
IV tPA before EVT has been shown to be highly associated with increased hospital bills and encounter charges without any improvement in clinical outcome compared with EVT alone. Additionally, patients in the EVT+IV tPA were shown to have a significantly increased risk of asymptomatic ICH. These results alone warrant a thorough reassessment of the efficacy as well as the guidelines of pre-thrombectomy IV tPA administration.
Contributors AEH provided research question, analyzed the data, and revised the paper. HK developed the statistical analyses, drafted the paper, and revised the paper. LG provided the hospital bill data. WT revised the paper. AS revised and edited the paper. AQ revised the paper.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests AEH Consultant for Medtronic, Microvention, Penumbra, Stryker, Genentech and GE Healthcare.
Provenance and peer review Not commissioned; externally peer reviewed.
Patient consent for publication Not required.
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.