Neurology

Trends in intravenous thrombolysis utilization for acute ischemic stroke based on hospital size: Paul Coverdell National Acute Stroke Program, 2010-2019

a b s t r a c t

Introduction: The rate of intravenous thrombolysis (IVT) utilization in acute ischemic stroke has been increasing, and this has coincided with improved Door-to-needle times (DNTs). Smaller hospitals have been observed to utilize IVT less frequently or even not at all. Using a multistate stroke registry, we sought to deter- mine the impact of hospital size on trends in IVT utilization for AIS.

Methods: Utilizing data from the Paul Coverdell National Acute Stroke Program (PCNASP), we studied trends in IVT for AIS patients between 2010 and 2019 based on hospital size. Hospitals were grouped into quartiles based on size. We studied the impact of hospital size on DNTs and overall IVT utilization.

Results: During the study period, there were 530,828 AIS patients (mean age 70.3 +- 0.02 years, 50.4% men) from 540 participating hospitals. We did not identify a significant trend in IVT utilization among hospitals within the first quartile (p = 0.1005), but there were significantly increased trends within the hospitals belonging to the second, third, and fourth quartiles (p < 0.001 for all). All quartiles were observed to have significantly increased trends in DNTs <=60 min (p < 0.0001), but only hospitals within the second, third, and fourth quartiles experi- enced significantly increased trends in DNTs <=45 min (p < 0.0001).

Conclusion: In our registry-based analysis, we observed an increased trend in IVT utilization for AIS among larger hospitals. There was an overall improvement in rates of DNTs <=60 min, but only larger hospitals were observed to have improved DNTs <=45 min.

(C) 2023 Published by Elsevier Inc.

  1. Introduction

Stroke continues to be a leading cause of death and serious disability in the United States [1]. The rate of Intravenous thrombolysis uti- lization in acute ischemic stroke has been increasing, and this has coincided with improved door-to-needle times (DNTs) [2,3]. Despite this observed trend, Kleindorfer and colleagues have discovered that more than half of the hospitals in the United States reported no use of IVT [4]. These tended to be smaller hospitals (average of approximately 95 beds) and those located in rural areas [4]. Small hospitals have been found to offer less optimal quality of stroke care, which has been correlated to lower IVT utilization rates; additionally, bed size has been reported to be inversely correlated with short-term mortality in some studies [5]. With overall increasing rates, we sought to determine

* Corresponding author at: Allina Health Institute of Neuroscience, Spine, and Pain, 310 North Smith Avenue, Suite 440, St. Paul, MN 55102, United States of America.

E-mail address: [email protected] (G. Asaithambi).

the impact of hospital size on trends in IVT utilization for AIS using a multistate stroke registry.

  1. Methods

Our study population included patients admitted with AIS from 2010 through 2019 within the Paul Coverdell National Acute Stroke Pro- gram (PCNASP). We only included patients with documented National Institutes of Health Stroke Scale Score (NIHSSS) at admission. The PCNASP is an ongoing acute stroke quality improvement program funded by the Centers for Disease Control and Prevention (CDC) and provides feedback to states on adherence to guidelines of care to im- prove care quality for patients hospitalized with stroke and Transient ischemic attack . Within our study period, hospitals across twelve states (Arkansas, California, Georgia, Iowa, Massachusetts, Michigan, Minnesota, New York, North Carolina, Ohio, Washington, and Wiscon- sin) participated in PCNASP. Hospital participation within each state is voluntary, and trained abstractors from participating hospitals collect

https://doi.org/10.1016/j.ajem.2023.02.009 0735-6757/(C) 2023 Published by Elsevier Inc.

detailed information on stroke and TIA admissions concurrent with or soon after hospital discharge using standard data definitions provided by the CDC [6,7]. This study was approved by the CDC Institutional Review Board.

We defined the rates of IVT utilization as percent of AIS admissions. We identified the proportion of patients who received IVT within 60 min of arrival and within 45 min of arrival (DNT). Hospital size was defined by number of Hospital beds at each participating site, and hospitals were divided into quartiles based on size. Baseline characteris- tics analyzed for the study included age, sex, and race/ethnicity. Clinical characteristics analyzed for the study included Stroke severity upon pre- sentation, as defined by NIHSSS, and proportion of patients arriving by ambulance. Comorbidities for patients studied included history of prior stroke, hypertension, dyslipidemia, coronary artery disease, heart failure, diabetes, atrial fibrillation, and history of smoking.

Categorical variables were compared across groups using two-tailed Fisher’s exact or chi-square tests. Descriptive statistics were expressed as means with standard error (SE), medians with interquartile range (IQR), and frequency (percentages). Continuous variables were com- pared using the Wilcoxon-Mann-Whitney rank test or the Kruskal- Wallis test. We examined trends and obtained the p-values based on Cochran-Armitage test. We performed both univariate and multivariate logistic regression analyses using generalized estimating equations (GEE) to assess the effects of hospital size on outcomes for stroke pa- tients receiving IVT and present data as odds ratio (OR) and adjusted odds ratio (AOR) with confidence intervals (CI). Because patients were clustered within hospitals, the hospital was treated as a cluster variable in GEE models. All statistical analyses were performed using SAS soft- ware (version 9.4; SAS Institute, Cary, NC). The data that support

findings within this study are available upon reasonable request to the corresponding author.

  1. Results

From 2010 to 2019, we had available bed size information from 540 participating hospitals in PCNASP; the median hospital bed size was 237 (IQR 111, 404). Within these hospitals, there were 530,828 patients (mean age 70.3 +- 0.02 years, 50.4% men) with a clinical diagnosis of AIS presenting with documented NIHSS at admission. Hospitals belong- ing to the first quartile had <111 beds, second quartile had 111-236 beds, third quartile had 237-403 beds, and fourth quartile had >=404 beds. Baseline and clinical characteristics are shown in Table 1. Among all patients identified through 2010-2019, approximately 46.2% of patients presented to participating hospitals by ambulance, median NIHSSS upon presentation was 4 (1, 9), and approximately 10.7% of pa- tients received IVT. Among all studied patients approximately 6.3% had DNTs <=60 min, and approximately 3.7% had DNTs <=45 min.

We identified significantly increased trends in the utilization of IVT within the hospitals belonging to the second, third, and fourth quartiles (Table 2, p < 0.001 for all). Hospitals in the second quartile increased IVT utilization from 8.9% of patients in 2010 to 10.4% of patients in 2019. Hospitals in the third quartile increased IVT utilization from 9% of patients in 2010 to 11.2% of patients in 2019. Hospitals in the fourth quartile increased IVT utilization from 10.7% of patients in 2010 to 11.9% of patients in 2019. These increases in IVT utilization corresponded to significantly increased trends in DNTs <=60 min and

<= 45 min (Figs. 1 and 2) in hospitals within the second, third, and fourth

quartiles as well (p < 0.001 for all). While there was no significantly

Table 1

Baseline and Clinical Characteristics of Acute Ischemic Stroke Admissions by Hospital Bed Size, 2010-2019.

Variables

Overall n (%) or statistics

1st quartile (<111 beds)

2nd quartile (111-236 beds)

3rd quartile (237-403 beds)

4th quartile (>=404 beds)

(N = 530,828)

(n = 15,452)

(n = 81,474)

(n = 130,236)

(n = 303,666)

Mean age (SE), years

70.3 (0.02)

72.8 (0.11)

71.5 (0.05)

71.5 (0.04)

69.4 (0.03)

No. of males (%)

267,420 (50.4)

7270 (47.0)

39,848 (48.9)

64,457 (49.5)

155,845 (51.3)

Race/Ethnicity

Non-Hispanic White

383,397 (72.2)

12,853 (83.2)

62,953 (77.3)

99,400 (76.3)

208,191 (68.6)

Non-Hispanic Black

94,604 (17.8)

1771 (11.5)

11,477 (14.1)

19,797 (15.2)

61,559 (20.3)

Hispanic

17,775 (3.3)

269 (1.7)

2374 (2.9)

4284 (3.3)

10,848 (3.6)

Arrival by ambulance

245,018 (46.2)

7096 (45.9)

40,256 (49.4)

63,793 (49.0)

133,873 (44.1)

Median NIHSSS (IQR)

4 (1, 9)

3 (1, 7)

3 (1, 7)

3 (1, 8)

4 (1,10)

Received IVT (%)

56,888 (10.7)

946 (6.1)

7741 (9.5)

13,618 (10.5)

34,583 (11.4)

DNT <= 60 min (%)

33,275 (6.3)

382 (2.5)

3929 (4.8)

7338 (5.6)

21,626 (7.1)

DNT <= 45 min (%)

19,644 (3.7)

167 (1.1)

2124 (2.6)

3961 (3.0)

13,392 (4.4)

Medical history (%)

Prior stroke

132,463 (25.0)

3945 (25.5)

20,157 (24.7)

33,696 (25.9)

74,665 (24.6)

Hypertension

403,316 (76.0)

11,906 (77.1)

62,643 (76.9)

100,680 (77.3)

228,087 (75.1)

Dyslipidemia

260,317 (49.0)

7789 (50.4)

41,316 (50.7)

67,504 (51.8)

143,708 (47.3)

CAD

127,526 (24.0)

3769 (24.4)

19,403 (23.8)

32,723 (25.1)

71,631 (23.6)

Heart failure

54,580 (10.3)

1537 (9.9)

7677 (9.4)

13,551 (10.4)

31,815 (10.5)

Diabetes

177,356 (33.4)

5154 (33.4)

27,839 (34.2)

44,510 (34.2)

99,853 (32.9)

Atrial fibrillation

102,528 (19.3)

3087 (20.0)

15,566 (19.1)

25,832 (19.8)

58,043 (19.1)

History of smoking

104,707 (19.7)

2804 (18.1)

14,887 (18.3)

24,460 (18.8)

62,556 (20.6)

Trends in Hospital Admissions of acute ischemic stroke patients Across all Four Quartiles (%) Percentages in each quartile reflect the proportion of admissions by year

2010

19,391 (3.7)

749 (3.9)

2653 (13.7)

4278 (22.1)

11,711 (60.4)

2011

23,931 (4.5)

1112 (4.6)

3113 (13.0)

5428 (22.7)

14,278 (59.7)

2012

27,466 (5.2)

1180 (4.3)

4063 (14.8)

6026 (21.9)

16,197 (59.0)

2013

43,370 (8.2)

1403 (3.2)

6687 (15.4)

9871 (22.8)

25,409 (58.6)

2014

53,895 (10.2)

1553 (2.9)

8437 (15.7)

13,090 (24.3)

30,815 (57.2)

2015

57,131 (10.8)

1564 (2.7)

8793 (15.4)

14,103 (24.7)

32,671 (57.2)

2016

62,730 (11.8)

1590 (2.5)

9798 (15.6)

15,988 (25.5)

35,354 (56.4)

2017

72,869 (13.7)

1932 (2.7)

10,962 (15.0)

18,479 (25.4)

41,496 (56.9)

2018

81,453 (15.3)

2053 (2.5)

13,052 (16.0)

20,757 (25.5)

45,591 (56.0)

2019

88,592 (16.7)

2316 (2.6)

13,916 (15.7)

22,216 (25.1)

50,144 (56.6)

Abbreviations: CAD = coronary artery disease; DNT = door to needle time; IQR = interquartile range; IVT = intravenous thrombolysis; NIHSSS=National Institutes of Health Stroke Scale score; SE = standard error.

Table 2

Trends in IVT utilization by year based on hospital size, 2010-2019.

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

P for trend

1st quartile

No. of patients

749

1112

1180

1403

1553

1564

1590

1932

2053

2316

(<111 beds)

Received IVT N(%)

DNT <=60 min

47 (6.3)

8 (1.1)

57 (5.1)

16 (1.4)

67 (5.7)

24 (2.0)

80 (5.7)

31 (2.2)

90 (5.8)

32 (2.1)

115 (7.4)

41 (2.6)

96 (6.0)

32 (2.0)

114 (5.9)

51 (2.6)

102 (5.0)

51 (2.5)

178 (7.7)

96 (4.1)

0.1005

<0.001

N(%)

DNT <=45 min

1 (0.1)

4 (0.4)

10 (0.8)

12 (0.9)

9 (0.6)

13 (0.8)

16 (1.0)

21 (1.1)

32 (1.6)

49 (2.1)

NA

2nd quartile

N(%)

No. of patients

2653

3113

4063

6687

8437

8793

9798

10,962

13,052

13,916

(111-236

Received IVT

236 (8.9)

259 (8.3)

344 (8.5)

616 (9.2)

727 (8.6)

735 (8.4)

950 (9.7)

1065 (9.7)

1368

1441

<0.001

beds)

N(%)

(10.5)

(10.4)

DNT <=60 min

47 (1.8)

70 (2.2)

108 (2.7)

254 (3.8)

314 (3.7)

365 (4.2)

503 (5.1)

604 (5.5)

793 (6.1)

871 (6.3)

<0.001

N(%)

DNT <=45 min

17 (0.6)

25 (0.8)

35 (0.9)

119 (1.8)

140 (1.7)

174 (2.0)

263 (2.7)

340 (3.1)

461 (3.5)

550 (4.0)

<0.001

3rd quartile

N(%)

No. of patients

4278

5428

6026

9871

13,090

14,103

15,988

18,479

20,757

22,216

(237-403

Received IVT

385 (9.0)

437 (8.1)

526 (8.7)

1014

1334

1451

1748

1893

2331

2499

<0.001

beds)

N(%)

(10.3)

(10.2)

(10.3)

(10.9)

(10.2)

(11.2)

(11.2)

DNT <=60 min N(%)

DNT <=45 min

91 (2.1)

40 (0.9)

131 (2.4)

56 (1.0)

167 (2.8)

71 (1.2)

447 (4.5)

207 (2.1)

646 (4.9)

337 (2.6)

791 (5.6)

391 (2.8)

969 (6.1)

502 (3.1)

1125 (6.1)

601 (3.3)

1393 (6.7)

787 (3.8)

1578 (7.1)

969 (4.4)

<0.001

<0.001

4th quartile

No. of patients

11,711

14,278

16,197

25,409

30,815

32,671

35,354

41,496

45,591

50,144

(>=404 beds)

Received IVT

1250

1471

1742

2802

3235

3476

4037

5066

5522

5982

<0.001

N(%)

(10.7)

(10.3)

(10.8)

(11.0)

(10.5)

(10.6)

(11.4)

(12.2)

(12.1)

(11.9)

DNT <=60 min N(%)

DNT <=45 min

480 (4.1)

198 (1.7)

619 (4.3)

261 (1.8)

827 (5.1)

395 (2.4)

1513 (6.0)

764 (3.0)

1846 (6.0)

996 (3.2)

2209 (6.8)

1346 (4.1)

2704 (7.6)

1698 (4.8)

3483 (8.4)

2268 (5.5)

3801 (8.3)

2563 (5.6)

4144 (8.3)

2903 (5.8)

<0.001

<0.001

N(%)

Abbreviations: DNT = door to needle times; IVT = intravenous thrombolysis.

increased trend for IVT utilization among hospitals within the first quar- tile (p = 0.1005), these hospitals did experience a significantly in- creased trend in DNTs <=60 min from 1.1% of patients in 2010 to 4.1% of patients in 2019 (p < 0.001). The proportion of patients with DNT

<=45 min at hospitals within the first quartile was too small to identify a trend.

Patients receiving IVT were less likely to have presented to hospi- tals in the first (AOR 0.53 [95% CI 0.49-0.57]), second (AOR 0.83 [95% CI 0.81-0.85]), and third (AOR 0.91 [95% CI 0.89-0.93]) quartiles when

compared to hospitals belonging to the fourth quartile. This corresponded to lower likelihood of DNT <=60 min among patients who presented to hospitals in the first (AOR 0.34 [95% CI 0.30-0.38]), second (AOR 0.65 [95% CI 0.63-0.68]), and third

(AOR 0.76 [95% CI 0.74-0.78]) quartiles (Table 3). Similarly, patients with DNT <=45 min were less likely to have presented to hospitals in the first (AOR 0.24 [95% CI 0.21-0.28]), second (AOR 0.57 [95% CI

0.55-0.60]), and third (AOR 0.66 [95% CI 0.64-0.69]) quartiles

(Table 3).

Fig. 1. Door to needle times within 60 minutes among patients receiving intravenous thrombolysis based on hospital size, 2010-2019.

Fig. 2. Door to needle times within 45 minutes among patients receiving intravenous thrombolysis based on hospital size, 2010-2019.

  1. Discussion

Over the past two decades, ongoing intervention programs have focused on providing all patients with sufficient access to time- sensitive reperfusion treatment options for AIS [8,9]. Extensive work with communities, including mass media strategies have shown the potential in helping increase acute stroke treatments [9]. The wider im- plementation of telestroke services has also helped increase access to care [8]. In 2011, approximately 20% of Americans lacked timely access to IVT-capable hospitals; however, with increase in hospital seeking stroke quality care certification, approximately 96% of the United States population are within 60 min of an emergency department with any acute stroke capabilities [8]. Even further, 99% of Americans have access to hospitals within 120 min for either IVT or Endovascular therapy by ground or Air ambulance [10].

In 2013, the Brain Attack Coalition endorsed the establishment of a certification program for acute stroke ready hospitals (ASRH) [11]. ASRHs are typically hospitals with fewer than 100 beds and can be found in suburban or rural locations [12]. By establishing Performance metrics for ASRHs, these centers can form the base of any regional stroke system of care [11]. Alberts and colleagues suggested that the

Table 3

Effect of hospital size on IVT utilization and time to treatment, 2010-2019.

OR (95% CI)

AOR (95% CI)

Received IVT

1st quartile 2nd quartile

0.51 (0.47-0.54)

0.82 (0.80-0.84)

0.53 (0.49-0.57)

0.83 (0.81-0.85)

3rd quartile 0.91 (0.89-0.93) 0.91 (0.89-0.93)

4th quartile Ref Ref

DNT <=60 min 1st quartile 0.33 (0.30-0.37) 0.34 (0.30-0.38)

2nd quartile 0.66 (0.64-0.68) 0.65 (0.63-0.68)

3rd quartile 0.78 (0.76-0.8) 0.76 (0.74-0.78)

4th quartile Ref Ref

DNT <=45 min 1st quartile 0.24 (0.20-0.28) 0.24 (0.21-0.28)

2nd quartile 0.58 (0.55-0.61) 0.57 (0.55-0.6)

3rd quartile 0.68 (0.66-0.7) 0.66 (0.64-0.69)

4th quartile Ref Ref

Abbreviations: AOR = adjusted odds ratio; CI = confidence interval; DNT = door to needle time; IVT = intravenous thrombolysis; OR = odds ratio.

* Adjusted for age, sex, race/ethnicity, arrival by ambulance, National Institutes of Health Stroke Scale score.

development of an ASRH network could increase the chances of AIS pa- tients receiving Timely interventions and higher chances of favorable outcomes [11]. Tsai and colleagues evaluated Minnesota-based data and discovered that early adoption of ASRH status led to a significantly increased trend in IVT utilization for AIS; however, this trend was not observed among hospitals that never achieved ASRH status [13].

With improving access to care, we know that rates of IVT utilization are increasing concurrently with improved DNTs [2,3]. Hospital size has been shown to be an important predictor of quality of care for stroke patients [5]; therefore, we elected to study the impact of hospital size on rates of IVT utilization as well as timeliness of treatment. Our results show that hospitals within the second, third, and fourth quartiles expe- rienced significantly increased trends in IVT utilization rates from 2010 to 2019. Timeliness to treatment also significantly increased within these quartiles with increased rates of DNTs <=60 min and <= 45 min. Smaller hospitals belonging to the first quartile did not experience a significant trend in IVT utilization rates during the study period, but a significantly increased trend in DNTs <=60 min was observed. There was no improvement in rates of DNTs <=45 min among first quartile hos- pitals. Previous analyses suggest that achieving optimal DNTs are likely secondary to varying levels of infrastructure and resources required for acute stroke care [3], which confirms why we identified that larger hospitals likely achieved better DNTs. Additionally, stroke centers with at least 100 stroke cases per year were found to have greater annual decreases in DNTs [3].

Important strengths of our study include the large number of patients from a multi-state registry during regular care delivery. The voluntary nature of the PCNASP limits generalizability, and hospitals more oriented toward quality improvement may be more likely to participate in the registry. Some participating hospitals who did not have quality improvement resources may not have participated in some years. Therefore, the trends we observed with PCNASP data may not be indicative of overall national trends and should be interpreted with caution.

  1. Conclusion

We observed an increased trend in IVT utilization for AIS among larger hospitals participating in PCNASP. Hospitals of all sizes in

PCNASP were noted to have improved rates of DNTs <=60 min, but only larger hospitals improved in rates of DNTs <=45 min. Ongoing quality improvement efforts should focus on smaller hospitals to continue increasing rates of IVT utilization for AIS.

Declaration of Competing Interest

GA has nothing to disclose. XT has nothing to disclose. KL has nothing to disclose.

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

References

  1. Tsao CW, Aday AW, Almarzooq ZI, Alonso A, Beaton AZ, Bittencourt MS, et al. Heart disease and stroke statistics-2022 update: a report from the American Heart Associ- ation. Circulation. 2022;145:e153-639.
  2. Asaithambi G, Tong X, Lakshminarayan K, Coleman King SM, George MG. Current trends in the acute treatment of ischemic stroke: analysis from the Paul Coverdell National Acute Stroke Program. J Neurointerv Surg. 2020;12:574-8.
  3. Tong X, Wiltz JL, George MG, Odom EC, Coleman King SM, Chang T, et al. A decade of improvement in door-to-needle time among acute ischemic stroke patients, 2008- 2017. Circ Cardiovasc Qual Outcomes. 2018;11:e004981.
  4. Kleindorfer D, Xu Y, Moomaw CJ, Khatri P, Adeoye O, Hornung R. US geographic dis- tribution of rt-PA utilization for acute ischemic stroke. Stroke. 2009;40:3580-4.
  5. Reeves MJ, Gargano J, Maier KS, Broderick JP, Frankel M, LaBresh KA, et al. Patient- level and hospital-level determinants of the quality of acute stroke care: a multilevel modeling approach. Stroke. 2010;41:2924-31.
  6. George MG, Tong X, McGruder H, Yoon P, Rosamond W, Winquist A, et al. Paul coverdell national acute stroke registry surveillance – four states, 2005-2007. MMWR Surveill Summ. 2009;58:1-23.
  7. Centers for Disease Control and Prevention (CDC). Use of a registry to improve stroke care – seven states, 2005-2009. MMWR Morb Mortal Wkly Rep. 2011;60 (7):206-10.
  8. Zachrison KS, Cash RE, Adeoye O, Boggs KM, Schwamm LH, Mehrotra A, et al. Esti- mated population access to acute stroke and Telestroke Centers in the US, 2019. JAMA Netw Open. 2022;5:e2145824.
  9. Morgenstern LB, Staub L, Chan W, Wein TH, Bartholomew LK, King M, et al. Improv- ing delivery of acute stroke therapy: the TLL Temple foundation stroke project. Stroke. 2002;33:160-6.
  10. Adeoye O, Albright KC, Carr BG, Wolff C, Mullen MT, Abruzzo T, et al. Geographic ac- cess to acute stroke Care in the United States. Stroke. 2014;45:3019-24.
  11. Alberts MJ, Wechsler LR, Lee Jensen ME, Latchaw RE, Crocco TJ, George MG, et al. For- mation and function of acute stroke-ready hospitals within a stroke system of care recommendations from the brain attack coalition. Stroke. 2013;44:3382-93.
  12. Higashida R, Alberts MJ, Alexander DN, Crocco TJ, Demaerschalk BM, Derdeyn CP, et al. Interactions within stroke systems of Care. A policy statement from the American Heart Association/American Stroke Association. Stroke. 2013;44:2961-84.
  13. Tsai AW, Lindgren PG, Asaithambi G. Trends in acute stroke treatment among rural hospitals in Minnesota: utility of acute stroke ready hospital designation. Stroke. 2018;49:AWP276.