Telestroke: Variations in Intravenous Thrombolysis by Spoke Hospitals Jeffrey A. Switzer, DO, MCTS,* Rajendra Singh, PhD,† Lars Mathiassen, PhD,‡ Jennifer L. Waller, PhD,x Robert J. Adams, MD, MS,k and David C. Hess, MDx
Background: Telestroke extends stroke expertise to underserved hospitals and facilitates treatment with tissue plasminogen activator (tPA). We investigated the variability in tPA treatment rates across 2 large telestroke networks—consisting of hubs at Georgia Regents Medical Center (GRMC) and Medical University of South Carolina (MUSC) and their affiliated spoke hospitals—to identify spoke-related factors predictive of greater tPA use. Methods: Observational study of tPA treatment rate at 32 spoke hospitals within the GRMC and MUSC telestroke networks. Spokes were characterized by primary stroke center status, local stroke nurse coordinator, local neurology support, hospital size, post-tPA management strategy, whether the spoke hospitals paid to participate in the network, and whether the hub or the spoke hospital initially proposed the telemedicine linkage for consultations with a remote stroke specialist. Primary outcome was tPA treatment rate adjusted for emergency department (ED) volume. Results: There was substantial variation in the adjusted tPA rate across spokes (range, .85-8.74 administrations/104 ED visits/ year). Only spokes with a stroke nurse coordinator (4.75/104 ED visits/year versus 2.84/104 ED visits/year, P 5 .03) were associated with higher tPA use. Conclusions: The application of telestroke has variable results on tPA delivery in spoke hospitals. However, the presence of a stroke nurse coordinator at the spoke facilitated treatment of ischemic stroke cases with tPA. Key Words: Telemedicine—stroke—tissue plasminogen activator—telestroke. Ó 2014 by National Stroke Association
Despite modest increases over the last few years, intravenous (IV) tissue plasminogen activator (tPA) remains underused in acute ischemic stroke.1 In small and rural
hospitals that often lack local neurologists, tPA administration is further reduced.2 Telestroke removes geographic barriers to stroke care, linking stroke experts
Department of Neurology, Medical College of Georgia, Georgia Regents University, Augusta, Georgia; †Department of Health Services Policy and Management, Arnold School of Public Health, University of South Carolina, Columbia, South Carolina; ‡Center for Process Innovation, J. Mack Robinson College of Business, Georgia State University, Atlanta, Georgia; xDepartment of Biostatistics and Epidemiology, Georgia Regents University, Augusta, Georgia; and kDepartment of Neurosciences, Medical University of South Carolina, Charleston, South Carolina. Received July 22, 2014; revision received September 9, 2014; accepted September 24, 2014. The study was funded by Genentech (Genentech ML28161). The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the article.
J.A.S. has served as a consultant for Genentech and REACH Health, Inc. R.S., L.M., and J.L.W. have no disclosures. D.C.H. is a cofounder of REACH Health Inc, has cofounder’s equity, and is on the REACH Health Inc Board of Directors. He serves on a Pfizer Data and Safety Monitoring Board and Adjudication Committee for CoVance and has received honoraria from Vindico CME. R.J.A. is a cofounder of REACH Health Inc and has cofounder’s equity. Address correspondence to Jeffrey A. Switzer, DO, MCTS, Department of Neurology, Medical College of Georgia, Georgia Regents University, 1122 15 Street, Augusta, GA 30912. E-mail:
[email protected]. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2014.09.028
Journal of Stroke and Cerebrovascular Diseases, Vol. -, No. - (---), 2014: pp 1-6
1
J.A. SWITZER ET AL.
2
with underserved hospitals, thereby improving tPA decision making, facilitating tPA administration, and improving functional outcome.3-5 Telestroke is cost effective from the societal and hospital perspective, and hub-and-spoke networks have become a widespread and expanding practice model.6-8 However, there are significant barriers to establishing and maintaining a telestroke program.9 To improve stroke care, telestroke must become part of routine decision making in the emergency department (ED) at the spoke hospitals. This requires ‘‘buy in’’ by administrators and local physicians, education of health-care personnel (including emergency medicine physicians, hospitalists, nurses, and prehospital providers), establishment of protocols of care, and user comfort with telemedicine, and the associated information and communication technologies. Given the costs to initiate and maintain a telestroke network, potential hubs and spokes need to understand the factors that predict clinical success (measured through increased tPA administration) to determine whether or not they will have a return on their investment. Previously, we conducted a qualitative analysis of the telestroke networks at the Georgia Regents Medical Center (GRMC) and Medical University of South Carolina (MUSC) by interviewing key stakeholders at the 2 hubs and selected spokes to gather a variety of perspectives on key enablers of telestroke assimilation.10 Our objective in this study was to examine the variation in usage of telestroke for decision making related to tPA administration across spokes and to identify spokerelated factors predictive of greater usage of the telestroke system.
Methods This study is a retrospective analysis of patients treated with IV tPA within the GRMC and MUSC telestroke networks. The telestroke program in Georgia was initiated in February 2003, when the Department of Neurology at GRMC signed a contract with the first spoke hospital. As of June 2012, 17 rural and community hospitals had become active spoke hospitals in the GRMC stroke program. Network spoke hospitals were located between 6 and 180 miles from the hub, covering 1831 beds (mean, 108 beds per hospital) and receiving 360,224 ED patients/year. Nine of the hospitals in the network were small (,100 beds) rural facilities, for which the hub subsidized the telestroke cart and technical support, and stroke consultation was provided at no cost to the spoke. In the remaining community and larger suburban hospitals, the spoke paid for technical costs and a per diem (spoke fee) to the GRMC Neurology Department for consultative coverage. The telestroke program at MUSC was initiated in May 2008, when the first spoke hospital signed a contract with MUSC for remote stroke consultations. By June 2012, 15 rural and community hospitals had
joined the South Carolina telestroke network. Altogether, these spoke hospitals have 2482 beds (mean, 165 beds per hospital), receive 471,875 ED patients/year and are located between 61 and 191 miles from the hub. All spokes in the South Carolina network paid a daily spoke fee to the hub for call coverage. The consultative process for GRMC and MUSC have been described in detail elsewhere.3,11 In brief, both networks use the same telemedicine platform, REACH 3.0 (Alpharetta, GA), with technology and 24 3 7 maintenance support provided by REACH Health, Inc. A mobile telestroke cart (which has a mounted adjustable camera and a high-resolution monitor) is located in the ED of each spoke hospital and provides a web-based 2-way video link. Audio is supplemented via an attached telephone. An emergency medicine physician at the spoke makes a decision to seek consultation and places a call to the GRMC or MUSC emergency call center respectively. The GRMC stroke consultants consist of 6 vascular neurologists and 1 emergency medicine physician, and MUSC consultants include 7 vascular neurologists and 1 emergency medicine physician. One physician (R.J.A.) has taken call for both networks. To register a patient, the ED staff at the spoke hospital enters the patient name, symptom onset time, and blood pressure. The patient’s computed tomography scan is uploaded to the REACH website for viewing by the consultant. A remote evaluation based on National Institutes of Health Stroke Scale is performed, and recommendations for or against IV tPA use are provided to the local emergency medicine physician including documentation of thrombolytic contraindication if indicated. Post-thrombolysis care instructions and a physician note are provided at consultation completion. Depending on patient and spoke hospital characteristics, patients are either transported to the hub (‘‘drip and ship’’) or maintained at the spoke (‘‘drip and keep’’) after the initiation of tPA treatment.12 Records of all consultations are stored by REACH Health, Inc. We queried each hub database for the total number of tPA treatments per spoke hospital. An average annual rate of tPA use at each spoke was calculated. To account for spoke hospital volume, we adjusted the annual rate by their reported ED volume (#tPA/year/104 ED volume). A few spokes required only partial telestroke coverage (for example, at nights only, or only certain days of the week). In these cases, we only considered tPA treatments that occurred as a result of a telestroke consultation with a hub specialist, and we further adjusted the tPA treatment rate by the period of partial coverage. We were interested in determining whether certain prespecified spoke characteristics influenced treatment rate. Therefore, spokes were categorized in terms of their stroke capabilities including primary stroke center (PSC) certification, designated stroke nurse coordinator,
Telestroke network
Spoke hospital
Date of initiation
Beds
PSC
Stroke coordinator
Local neurologist
Post-tPA management strategy
Technology, push/pull
Spoke fees
Adjusted annual tPA rate
GRMC GRMC GRMC GRMC GRMC GRMC GRMC GRMC GRMC GRMC GRMC GRMC GRMC GRMC GRMC GRMC GRMC MUSC MUSC MUSC MUSC MUSC MUSC MUSC MUSC MUSC MUSC MUSC MUSC MUSC MUSC MUSC
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
2/1/03 3/1/03 7/1/03 8/1/03 9/1/03 3/1/04 4/1/04 2/1/05 3/1/06 1/1/08 8/1/08 6/1/09 10/1/09 10/1/09 1/1/10 3/1/10 11/1/10 5/1/08 5/6/08 5/7/08 9/1/08 9/18/08 12/23/08 1/20/10 3/26/10 5/19/10 7/29/10 8/26/10 1/21/11 2/28/11 2/28/11 3/2/11
72 47 50 10 56 65 20 52 71 191 236 40 190 196 180 163 192 131 140 453 220 124 25 45 288 121 79 231 116 105 50 354
No No No No No No No No No No Yes No Yes Yes Yes No No No No No No No No Yes Yes No No No No No No No
No No No No No No No No No No No No Yes Yes Yes No Yes No No Yes No No Yes Yes Yes Yes No No No No No Yes
None None None None None None None None None Community Neurohospitalist None Community Neurohospitalist Neurohospitalist Community Community Community Community Neurohospitalist Neurohospitalist None None Community Neurohospitalist Community None None Community Community Community Neurohospitalist
Drip and Ship Drip and Ship Drip and Ship Drip and Ship Drip and Ship Drip and Ship Drip and Ship Drip and Ship Drip and Ship Drip and Keep Drip and Keep Drip and Ship Drip and Keep Drip and Keep Drip and Keep Drip and Keep Drip and Keep Drip and Ship Drip and Ship Drip and Ship Drip and Keep Drip and Ship Drip and Ship Drip and Keep Drip and Keep Drip and Keep Drip and Ship Drip and Ship Drip and Ship Drip and Ship Drip and Ship Drip and Keep
Push Push Push Push Push Push Push Push Push Pull Pull Push Pull Pull Pull Pull Pull Push Push Push Push Push Push Push Pull Pull Pull Push Pull Pull Pull Pull
No No No No No No No No No Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
4.45 2.78 4.04 2.39 4.04 2.39 1.31 2.55 3.96 1.84 1.97 1.63 2.33 8.74 3.36 .85 1.78 2.62 6.60 3.31 1.49 2.26 6.40 4.85 6.40 3.10 4.27 4.77 1.56 3.90 .90 7.26
TELESTROKE: VARIATION ACROSS SPOKES
Table 1. Spoke characteristics and adjusted annual tPA rate (per 104 ED volume/year)
Abbreviations: ED, emergency department; GRMC, Georgia Regents Medical Center; MUSC, Medical University of South Carolina; PSC, primary stroke center; tPA, tissue plasminogen activator.
3
J.A. SWITZER ET AL.
4
the availability of a local neurologist for post-tPA consultation, hospital size (small hospitals defined as ,100 beds), and post-tPA management strategy (‘‘drip and ship’’ vs. ‘‘drip and keep’’). Stroke nurse coordinators were self-identified by spoke hospitals, and their responsibilities were not prespecified. The availability of local neurologists was also identified by the spokes and could be either community neurologists who provided hospital consultation or neurohospitalists employed by the spoke hospitals. We suspected that hospitals certified as PSCs, hospitals with stroke nurse coordinator support, hospitals that had access to local neurologists, hospitals with more than 100 beds, and those that did not transfer patients after tPA administration would have higher telestroke-mediated tPA usage. In addition, spokes were also categorized as to whether they initially solicited the hub for telestroke services (‘‘pull’’) or were approached by the hub (‘‘push’’) to join the network13 and whether the spoke paid the hub a physician support fee. Spokes that sought telestroke support and funded their participation in the network were anticipated to have higher tPA usage. All statistical analysis was performed using SAS 9.3 (SAS, Cary, NC). Statistical significance was assessed using an alpha level of .05. Descriptive statistics were calculated on all variables. To examine differences in tPA usage for various spoke characteristics, 2 sample t tests or 1-way analysis of variance was performed, depending on the number of categories in the spoke characteristic variable. For the 1-way analysis of variance analysis, a Tukey– Kramer multiple comparison procedure was used to examine post hoc pairwise differences while controlling the overall alpha level.
Results The characteristics of the 32 spoke hospitals in the GRMC and MUSC networks are detailed in Table 1. In the GRMC network, tPA was administered on 381 occasions (spoke range, 4-53; mean, 22) compared with 489 (spoke range, 3-92; mean, 33) in the MUSC network. The mean adjusted tPA usage across all spokes was 3.44 administrations/104 ED visits/year; more patients were treated at MUSC network spokes (mean, 3.98; standard deviation, 2.04) than GRMC (mean, 2.96; standard deviation, 1.82), but this difference was not significant (P 5 .15; Table 2). There was large variability in adjusted tPA use at spokes, with 10-fold range across hospitals (range, .858.74 administrations/104 ED visits/year). One hospital, GRMC-14, had significantly higher rates of tPA use than the other spokes (Fig 1). Hospitals with a stroke coordinator (P 5 .03) were associated with greater tPA use (Table 2). Primary stroke center designation (P 5 .25), adopting ‘‘drip-and-keep’’ as the primary post-tPA treatment strategy (P 5 .67), paying spoke fees (P 5 .24), hospital size (P 5 .70), and having local neurology support (P 5 .14) were not associated with greater tPA use at the spoke.
Discussion Underutilization of tPA for acute ischemic stroke remains a problem in many hospitals. In this analysis of 2 large, longstanding telestroke networks, we found that tPA treatment rates vary widely across spoke hospitals. Although telestroke increases access to a stroke specialist in underserved locations, it does not ensure a uniformly
Table 2. Mean tPA treatment rate per 104 ED volume/year Variable Network Primary stroke center Stroke coordinator Post-tPA management strategy Hospital size Local neurology support
Spoke fees Push or pull
Level
Mean
SD
P value
GRMC MUSC No Yes No Yes Drip and Keep Drip and Ship #100 beds .100 beds None Community neurologist Neurohospitalist No Yes Pull Push
2.96 3.98 3.17 4.61 2.84 4.75 3.64 3.31 3.28 3.56 3.37 2.76 4.65 2.95 3.66 3.45 3.44
1.82 2.04 1.74 2.61 1.47 2.34 2.45 1.61 1.54 2.28 1.41 1.76 2.81 1.10 2.24 2.45 1.57
.15 .25 .03 .67 .70 .14
.24 .99
Abbreviations: ED, emergency department; GRMC, Georgia Regents Medical Center; MUSC, Medical University of South Carolina; SD, standard deviation; tPA, tissue plasminogen activator.
TELESTROKE: VARIATION ACROSS SPOKES
5
Figure 1. Distribution of adjusted annual tissue plasminogen activator (tPA) rate. Red lines are the mean 6 2 standard deviations.
higher rate of tPA treatment as illustrated by several spokes in both networks with very low tPA rates. Spoke hospitals with stroke nurse coordinators demonstrated higher rates of thrombolysis. By working with prehospital providers, the coordinators may increase the referral of patients with stroke symptoms to their ED. In addition, they may ensure that an evaluation algorithm is in place and is followed routinely, and that their stroke care team (including ED personnel, radiology staff, and prehospital providers) is educated and comfortable with telestroke technology and the consultative process. Although we do not know what factors determined whether a hospital had a stroke coordinator, the presence of a coordinator indicated a senior leadership commitment to stroke that trickled down to other health-care personnel, ensuring closer monitoring of the frequency of consultation by local administration, or direct encouragement of telestroke usage. However, many of the hospitals that lacked a designated coordinator were small (,100 beds, ,15,000 annual ED volume) and had limited financial capacity. Lower treatment rates in these settings may therefore simply reflect the challenges of telestroke at very small hospitals that do not evaluate many acute stroke cases. Perhaps surprisingly, other spoke factors, including PSC certification, did not significantly increase tPA use. This is in contrast to other studies,14-16 and may reflect the size of our sample and heterogeneity of the hospitals. Alternatively, this data may emphasize the importance of the nurse coordinator’s role in stroke center operations. Although many small hospitals may not be able to afford a full-time, dedicated nurse coordi-
nator, many of the coordinators in our networks carried out multiple functions at their hospitals in addition to stroke care coordination. Many spoke hospitals (11 of 32) in the 2 networks relied on a community neurologist for consultations during the day and telestroke at nights and/or weekends. However, the presence of a community neurologist at the spoke did not impact tPA treatment rates during the periods of telestroke coverage (Table 2). Conversely, there was a nonsignificant increase in tPA treatment at spokes with in-house neurohospitalists. In fact the hospital with the highest tPA rate, GRMC-14, has an in-house neurohospitalist on duty 7 days a week from 8 am to 6 pm and uses telestroke for after hours consultation. In the Finnish telestroke model, the combination of onsite daytime neurologists and night-time teleconsultation also resulted in a high rate of tPA treatment and favorable outcomes.17 For community hospitals, the combination of an in-house neurohospitalists by day and telestroke support at night may be an ideal model to support round-the-clock tPA treatment and become primary stroke centers. As discussed earlier, ‘‘push’’ refers to the scenario in which the hub contacts a spoke hospital and encourages it to adopt the telemedicine platform and related stroke model of care. In contrast, with ‘‘pull,’’ a spoke hospital contacts the hub and seeks out the telestroke service. Most of the early spokes in Georgia and South Carolina were recruited to their respective networks (ie, push), whereas most of the later spokes made the initial approach (ie, pull). We assumed that the spokes with pull would show higher tPA usage (because a
J.A. SWITZER ET AL.
6
spoke would likely be more committed to the telestroke program if it solicited the hub and sought to become a paying telestroke customer). However, we found no difference in treatment rates between push and pull spokes (Table 2), and therefore, this concept may be less important than in other health-care technologies. We acknowledge that our methodology has several limitations. First, we did not have data on the ED ischemic stroke volume at spokes, and therefore, we adjusted tPA usage by their overall ED volume. It is possible that the proportion of ischemic stroke volume relative to overall volume differs across spoke hospitals, particularly if prehospital providers would target certain spokes and provide EDs prehospital stroke notification, interventions that improve thrombolysis rates and timing. Second, our analysis focused on tPA administered at spokes only under telestroke guidance and not total tPA that may have been given. For spoke hospitals with round-the-clock telestroke coverage, we could not confirm that tPA was never administered outside of telestroke consultation, and we suspect that any independent tPA usage was a rare event. Third, there may be a ‘‘learning curve’’ to telestroke-guided thrombolysis. However, all spokes were operational for more than a year at the time of data analysis, and the last spoke added to the MUSC network had among the highest adjusted rates of tPA usage. Fourth, we did not predefine the roles of stroke coordinators at the spokes and responsibilities likely varied across hospitals. Future research should confirm and characterize which coordinator functions facilitate tPA administration. Finally, we do not have data on tPA rates from some spokes before joining a telestroke network and cannot determine if telestroke increased tPA usage. In conclusion, tPA usage varies significantly across spokes within telestroke networks, but may be influenced by certain spoke-related characteristics. Specific hub and spoke strategies to maximize tPA use within the expanding model of telestroke should be further explored. Acknowledgment: J.A.S. had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
References 1. Bohannon RW, Lee N. Association of physical functioning with same-hospital readmission after stroke. Am J Phys Med Rehabil 2004;83:434-438. 2. Kleindorfer D, Xu Y, Moomaw CJ, et al. US geographic distribution of rt-PA utilization by hospital for acute ischemic stroke. Stroke 2009;40:3580-3584. 3. Switzer JA, Hall C, Gross H, et al. A web-based telestroke system facilitates rapid treatment of acute ischemic stroke patients in rural emergency departments. J Emerg Med 2009;36:12-18. 4. Meyer BC, Raman R, Hemmen T, et al. Efficacy of siteindependent telemedicine in the STRokE DOC trial: a randomised, blinded, prospective study. Lancet neurol 2008;7:787-795. 5. Schwab S, Vatankhah B, Kukla C, et al. Long-term outcome after thrombolysis in telemedical stroke care. Neurology 2007;69:898-903. 6. Nelson RE, Saltzman GM, Skalabrin EJ, et al. The costeffectiveness of telestroke in the treatment of acute ischemic stroke. Neurology 2011;77:1590-1598. 7. Hendrikse J, Petersen ET, Golay X. Vascular disorders: insights from arterial spin labeling. Neuroimaging Clin N Am 2012;22:259-269. x-xi. 8. Pruss H, Dalmau J, Harms L, et al. Retrospective analysis of NMDA receptor antibodies in encephalitis of unknown origin. Neurology 2010;75:1735-1739. 9. Switzer JA, Demaerschalk BM. Overcoming challenges to sustain a telestroke network. J Stroke Cerebrovasc Dis 2012;21:535-540. 10. Singh R, Mathiassen L, Switzer JA, et al. Assimilation of web-based urgent stroke evaluation: a qualitative study of two networks. JMIR Med Inform 2014;2. 11. Lazaridis C, DeSantis SM, Jauch EC, et al. Telestroke in South Carolina. J Stroke Cerebrovasc Dis 2013;22:946-950. 12. Hess DC, Switzer JA. Stroke telepresence: removing all geographic barriers. Neurology 2011;76:1121-1123. 13. Zmud RW. ‘‘An examination of ‘‘push-pull’’ theory applied to process innovation in knowledge work’’. ManageSci 1984;30:727-738. 14. Gropen TI, Gagliano PJ, Blake CA, et al. Quality improvement in acute stroke: the New York State Stroke Center Designation Project. Neurology 2006;67:88-93. 15. Stradling D, Yu W, Langdorf ML, et al. Stroke care delivery before vs after JCAHO stroke center certification. Neurology 2007;68:469-470. 16. Prabhakaran S, McNulty M, O’Neill K, et al. Intravenous thrombolysis for stroke increases over time at primary stroke centers. Stroke 2012;43:875-877. 17. Sairanen T, Soinila S, Nikkanen M, et al. Two years of Finnish telestroke: thrombolysis at spokes equal to that at the hub. Neurology 2011;76:1145-1152.
