- Email: [email protected]
Factors Associated with the Administration of Tissue Plasminogen Activator for Acute Ischemic Stroke
Factors Associated with the Administration of Tissue Plasminogen Activator for Acute Ischemic Stroke Susumu Kunisawa, MD,* Daisuke Kobayashi, MPH,* Jason Lee, PhD,* Tetsuya Otsubo, PhD,* Hiroshi Ikai, MD, PhD,* Chiaki Yokota, MD, PhD,† Kazuo Minematsu, MD, PhD,‡ and Yuichi Imanaka, MD, PhD*
Background: The use of intravenous tissue plasminogen activator (t-PA) can be an effective treatment for acute ischemic stroke if administered promptly. Despite its clinical effectiveness, overall use in Japan remains low, and regional variations have been reported. Factors such as ambulance utilization and geographical distance from patients’ residences to hospitals may influence t-PA administration rates. The aim of this study is to identify factors associated with the administration of t-PA for acute ischemic stroke while adjusting for casemix using a large-scale administrative database in Japan. Methods: We analyzed acute ischemic stroke patients admitted to acute care hospitals between July 2010 and March 2011 using a nationwide database. A logistic regression model was used to analyze the factors influencing t-PA administration. Candidate factors included patient gender, age, stroke severity, direct distance between each patient’s residence and admitting hospital, and ambulance utilization. Results: Of the 10,615 ischemic stroke patients from 89 hospitals analyzed, 557 (5.2%) received t-PA treatment. Patients aged 75 years and older were found to be associated with decreased t-PA administration. In contrast, severe stroke and ambulance utilization were associated with increased t-PA administration. Distance was not significantly associated with the use of t-PA. Conclusions: Our findings suggest that ambulance utilization is an important factor for improving the likelihood of t-PA administration in patients with stroke and may underline a need for educational programs to the general public that promote the use of ambulances for suspected stroke patients. Key Words: Stroke—tissue plasminogen activator— health services accessibility—ambulances—risk adjustment. Ó 2014 by National Stroke Association
Introduction The use of intravenous tissue plasminogen activator (t-PA) was approved as a treatment for acute ischemic stroke in Japan in October 2005. Although utilization rates
From the *Department of Healthcare Economics and Quality Management, Graduate School of Medicine, Kyoto University, Kyoto, Japan; †Department of Cerebrovascular Medicine, National Cerebral and Cardiovascular Center, Suita, Osaka, Japan; and ‡National Cerebral and Cardiovascular Center, Suita, Osaka, Japan. Received May 10, 2013; revision received June 20, 2013; accepted June 21, 2013. Sources of funding: This study was supported in part by a Health and Labour Sciences Research Grant from the Ministry of Health, Labour and Welfare, Japan; an Intramural Research Grant for Cardio-
724
have gradually increased after its first application, the overall use remains low, and regional variations have been reported.1 Opportunities to administer t-PA are limited as guidelines recommend the use of intravenous t-PA
vascular Diseases from the National Cerebral and Cardiovascular Center; and a grant-in-aid for Scientific Research from the Japan Society for the Promotion of Science. Address correspondence to Yuichi Imanaka, MD, PhD, Department of Healthcare Economics and Quality Management, Graduate School of Medicine, Kyoto University, Yoshida Konoe-cho, Sakyo-ku, Kyoto City, Kyoto 606-8501, Japan. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2013.06.033
Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 4 (April), 2014: pp 724-731
FACTORS ASSOCIATED WITH USE OF t-PA
only to patients within a few hours of acute ischemic stroke onset.2-5 Because the use of t-PA is dependent on the elapsed time after stroke onset, emergency medical service infrastructure and geography are important factors to consider in an analysis of this treatment. Variations in these factors can directly affect the time duration from onset to treatment and may result in regional health disparities that need to be addressed from a public health perspective. Although the factors that influence t-PA administration have been previously investigated, these studies have generally been conducted using relatively small sample sizes.6-8 A previous study has reported an association between the use of ambulance transportation and acute stroke intervention in young adults,7 with the stated caveat that the limitations of small sample size and lack of stroke severity data may have affected the observed results. Additionally, the association between the distance from patient’s residence to hospital and t-PA utilization has also been reported.8-10 However, the results concerning the relationship between patient’s travel distance and delayed hospital arrival are inconclusive as other researchers have reported no association between these 2 factors.11 Understanding the effects of ambulance use and geographical distance between patient’s residence and hospital on the likelihood of t-PA use is essential to improve stroke care from political and social perspectives. To do so, an analysis based on a relatively large sample size comprising numerous hospitals of different sizes and infrastructures that are located throughout various regions would yield results with greater external validity. Since the introduction of the hospital reimbursement system known as the Diagnosis Procedure Combination (DPC) per-diem payment system in Japan in 2003, hospitals using this system routinely generate the DPC data for each patient per hospitalization for reimbursement purposes. From July 2010 onward, the DPC per-diem payment system has required the collection of additional information such as patient’s residence postal code, neurologic deficits, and date of stroke onset for patients admitted for stroke. These data are uniformly formatted, which allows researchers to collect, manage, and analyze large quantities of administrative data from numerous hospitals located nationwide. Furthermore, the recent inclusion of factors reflecting disease severity such as patient consciousness and physical impairment levels in this database supports analyses that take into account critical differences in patient casemix. An extensive investigation of the associations between t-PA use and the factors of ambulance use and geographical distance from patient’s residence to hospital in Japan has not yet been conducted. Therefore, the objective of this study was to investigate the influence of these and other factors on t-PA administration using a large administration database in Japan.
725
Methods Study Design and Setting We used a nationwide database maintained by the government and public research funds. This database comprises data from more than 300 hospitals distributed across Japan and includes a variety of public and private hospitals, with varying teaching statuses, case volumes, casemix, and specialties. In addition to hospital identifier codes, patient identifier codes, and DPC codes, the DPC data also include items such as principal diagnosis and other medical and demographic information such as gender, date of ischemic stroke onset, consciousness and physical impairment levels at admission, postal code of patient’s residence, and information regarding whether the patient arrived at the hospital by ambulance. The postal code for each patient’s residence and the address for the hospital to which they were admitted were geocoded to longitudinal and latitudinal co-ordinates using MarketPlanner 2.5.1 (PASCO Corporation, Tokyo, Japan) and Geocoding Tools & Utilities (Center for Spatial Information Science, Tokyo, Japan)12; the direct distances between the 2 were then calculated. We estimated the average error of the distance when using the postal code as a substitute for patients’ addresses to be less than 1 km, by applying the following formula for each city: sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðinhabitable areaÞ 3:14 3 ðnumber of postal codesÞ The average radius of each postal code was calculated to be .68 km (6.34 SD).
Data Selection Using the DPC database, we identified and extracted acute ischemic stroke cases discharged between July 2010 and March 2011. The criteria for identification of acute ischemic stroke were cases with (1) ‘‘010060’’ as the first 6 digits of the 14-digit DPC code; (2) an International Classification of Diseases, Tenth Revision, ‘‘I63$’’ as the principal diagnosis code; and (3) stroke onset occurring either on the day before or on the day of admission. We limited the analysis to patients residing within 50 km from the admitting hospital. Cases that were administered t-PA were identified by the number ‘‘4’’ in the 12th digit of the DPC code. Because not all hospitals are equipped to provide t-PA treatment,2 we further limited our analysis to include only hospitals that had administered t-PA to 1 or more patients during the study period. Only cases with complete information for the variables of interest were retained for analysis. Consequently, a total of 10,615 cases from 89 hospitals were included in the analysis.
S. KUNISAWA ET AL.
726
Levels of Consciousness and Physical Impairment The National Institutes of Health Stroke Scale is used in several countries to assess the severity of stroke patients. However, the DPC data do not include this information. Instead, we used a combination of consciousness and physical impairment levels. Consciousness levels were determined using the Japan Coma Scale (JCS). The JCS consists of 10 levels of consciousness, which can be categorized into the following 4 grades: grade 0 (alert); grade 1 (awake without any stimuli), comprising JCS levels 1-3; grade 2 (arousable), comprising JCS levels 10-30; and grade 3 (unarousable), comprising JCS levels 100300.13,14 In this study, JCS was included in analysis using these 4 grades. The level of physical impairment was determined using the modified Rankin Scale (mRS) scoring system. The mRS is principally an indicator of activities of daily living concerned with disability and dependence levels of patients after stroke.15 The DPC data include the mRS scores of patients at admission with any type of stroke because of the relative ease of assessment and the need for a common scale among the various types of stroke under the uniformly formatted DPC database. In this context, the mRS for acute stroke at admission assesses the level of physical impairment at the point of admission. Although the mRS is scored from 0 to 6, we selected patients with an mRS score of 1 or more and categorized them into 3 classes for our analysis: 1-2 (mild), 3-4 (moderate), and 5 (severe) as we assumed that these combined classes would allow for more distinct differentiation of the levels of severity. Finally, we created the following 8 patient categories for the analysis based on the paired JCS grades and mRS classes. The categories were (1) JCS grade 0 and mild mRS score; (2) JCS grade 0 and moderate mRS score; (3) JCS grade 0 and severe mRS score; (4) JCS grade 1 and mild mRS score; (5) JCS grade 1 and moderate mRS score; (6) JCS grade 1 and severe mRS score; (7) JCS grade 2; and (8) JCS grade 3. Because of the difficulties in assessing the mRS score levels of comatose patients, the highest 2 JCS grades were not paired with any specific mRS class.
Statistical Analysis We used multivariable logistic regression analysis to identify independent factors associated with t-PA administration. Independent variables in the regression model included patient gender, age, stroke severity, direct distance between each patient’s residence and admitting hospital, and ambulance utilization. We analyzed distance as an ordinal variable using the category with the shortest distance as the reference group to test the hypothesis that distance could affect t-PA administration after a certain threshold value. Statistical tests were 2 tailed and associations were considered significant at P less than .05. Statistical analyses were conducted using
SPSS software, version 19.0.0.2 (IBM, Tokyo, Japan). Discrimination of the regression model was evaluated with the c statistic.
Ethical Standard This study was approved by the Ethics Committee, Kyoto University Graduate School and Faculty of Medicine.
Results Characteristics of Study Subjects The baseline characteristics and medical information of the 10,615 patients are presented in Table 1. The mean age of the patients in the sample was 74.8 6 11.9 years, and 41.3% of the patients were women. Of the patients in the sample, 557 (5.2%) received t-PA during admission. These 557 patients who received t-PA treatment constitute approximately 9% of all stroke patients in Japan who had received t-PA treatment during the study period, which is estimated to be approximately 6400 patients (T. Ueno, Mitsubishi-Tanabe Pharmaceutical Corporation, written communication, May 2012). The unadjusted rate of t-PA administration was higher in patients with JCS grade 2 more severe mRS scores, and those who arrived by ambulance.
Factors Associated with t-PA Administration Table 2 shows the results of the multivariable logistic analysis. Patients aged 75 years or older were less likely to receive t-PA when compared with patients younger than 65 years. In contrast, administration of t-PA was more likely in patients with a medical status more severe than JCS grade 0 and a mild mRS score. Increasingly severe stroke symptoms were associated with increased t-PA administration, except for the most severe cases of JCS grade 3; the likelihood of t-PA administration for the cases of JCS grade 3 was lower than that of JCS grade 2. The direct distance between each patient’s residence and the admitting hospital did not show a significant association with t-PA administration, but arrival by ambulance was significantly associated with a substantial increase in the likelihood of t-PA administration (odds ratio, 4.31; 95% confidence interval, 3.27-5.69). The discrimination of the regression model was high, with a c statistic of .808.
Discussion By analyzing the administrative data of 10,615 cases from 89 hospitals in Japan, our study showed that 557 (5.2%) acute ischemic stroke patients received t-PA while hospitalized. Elderly patients were found to be associated with decreased t-PA administration. However, increasingly severe stroke symptoms (except for the most severe
FACTORS ASSOCIATED WITH USE OF t-PA
727
Table 1. Characteristics, medical information, and administration of t-PA of acute ischemic stroke patients t-PA administration Variables
Yes
%
Total
Gender Male 321 5.2 6226 Female 236 5.4 4389 Age (y) ,65 111 5.2 2139 65-74 138 5.2 2670 75-84 215 5.8 3704 .85 93 4.4 2102 Consciousness level at admission* Grade 0 (alert) 93 1.8 5187 Grade 1 (awake) 270 6.9 3920 Grade 2 (arousable) 147 14.7 997 Grade 3 (unarousable) 47 9.2 511 Physical impairment at admissiony Mild (1-2) 36 1.1 3249 Moderate (3-4) 216 4.5 4788 Severe (5) 305 11.8 2578 Direct distance from patient home to the hospital ,2 km 113 5.5 2054 2-3 km 139 5.2 2671 4-5 km 83 4.8 1728 6-7 km 55 5.0 1099 8-9 km 38 4.7 808 10-14 km 60 5.3 1141 15-19 km 37 7.2 513 20-29 km 26 6.3 412 30-49 km 6 3.2 189 Arrival by ambulance No 62 1.2 4986 Yes 495 8.8 5629 Total 557 5.2 10,615 Abbreviations: JCS, Japan Coma Scale; mRS, modified Rankin Scale; t-PA, tissue plasminogen activator. *Consciousness level was scored by the JCS. yPhysical impairment was assessed using the mRS.
cases) and ambulance utilization were associated with increased t-PA administration. On the surface, it may seem that the utilization rate of tPA in this sample is low. However, it should be noted that simple comparisons with the utilization rates in previous studies may not be appropriate as study samples can differ between analyses.6-8,16 In this study, we analyzed acute ischemic stroke patients, all of whom are potentially eligible for t-PA administration, although there may have been some patients who were contraindicated for this treatment. When compared with younger patients, administration of t-PA was found to be less likely in patients older than 75 years. The Japanese guidelines advise cautious administration to patients older than 75 years, and the European guidelines recommend administration to patients older
4,5
than 80 years only if they fulfill certain criteria. Our findings were consistent with these indications but in contrast with those observed in a German study.17 On the other hand, aging has also been found to be associated with delays in seeking medical help and time to hospital arrival.18,19 In this context, our results may also partly reflect this delay caused by aging. Gender had no association with the likelihood of t-PA administration in our analysis, but this is inconsistent among previous studies.7,20-22 Our findings showed that consciousness and physical impairment levels were associated with t-PA administration, which is also consistent with recommendations in guidelines. Guidelines do not recommend the use of t-PA treatment for patients with minor neurologic symptoms.2,3 This was compatible with our findings that patients with more severe stroke symptoms were more likely to receive t-PA when compared with patients with minor symptoms. On the other hand, guidelines also recommend caution in t-PA use for patients with major neurologic deficits, such as the JCS grade 3 (patients in an unarousable coma)2,3; accordingly, the odds ratio for t-PA administration was correspondingly lower for patients with the JCS grade 3 relative to the JCS grade 2 (patients in an arousable coma) in this study. The association between severity of stroke symptoms and tPA administration may in part be indicative of how quickly the patients or their families acted in recognizing a stroke and seeking appropriate care. As previous authors have shown, an increase in the severity of stroke symptoms is associated with shorter time to treatment.3,10,18,19,23,24 The observed influences of the severity of stroke symptoms on t-PA administration may, therefore, have multiple interpretations. Distance did not show a significant association with t-PA administration in our sample. The variations in results regarding this association in previous studies were likely to be dependent on factors such as social and health-care systems, geography, and traffic networks such as highway systems. Moreover, although we employed direct distance for our analysis, actual driving times may differ. However, a previous study in the Kumamoto region of Japan showed that the direct distance was significantly correlated with the transportation time by ambulance.11 Nonetheless, it may be important for future analyses to take into account the hour of the day, the day of the week, and the traffic density to understand regional situations better. As there are many mountainous regions in Japan, driving times may also be affected depending on whether patient routes include passages through flatlands or mountain corridors. Additionally, we analyzed the direct distance from each patient’s residence to the admitting hospital. Because stroke onset does not always occur at home, the location of onset may introduce a degree of bias to actual distances traveled. However, as a previous study has shown that 79.5% of stroke onset occurs
S. KUNISAWA ET AL.
728
Table 2. Results of a multiple logistic regression model with t-PA administration for acute ischemic stroke patients as the dependent variable
Independent variables Gender Male (reference) Female Age (y) ,65 (reference) 65-74 75-84 .85
OR (95% CI)
P value
.98 (.81-1.19)
.875
.78 (.59-1.02) .65 (.50-.83) .33 (.24-.45)
.068 .001 ,.001
Consciousness level and physical impairment level at admission Consciousness*
Physical impairmenty
Grade 0 (alert) Mild (reference) Moderate Severe
.08 (.04-.15) .25 (.16-.38)
,.001 ,.001
1.05 (.64-1.72) .26 (.15-.46) .73 (.51-1.05) 1.30 (.91-1.87) 1.72 (1.21-2.45)
.851 ,.001 .091 .146 .003
Grade 1 (awake) Mild Moderate Severe Grade 2 (arousable) Grade 3 (unarousable)
Direct distance between patient home and the hospital ,2 km (reference) 2-3 km 4-5 km 6-7 km 8-9 km 10-14 km 15-19 km 20-29 km 30-49 km Arrival by ambulance
.91 (.69-1.18) .78 (.57-1.05) .79 (.56-1.12) .75 (.51-1.11) .84 (.60-1.17) 1.08 (.72-1.62) .95 (.60-1.50) .49 (.21-1.16) 4.31 (3.27-5.69)
.469 .098 .182 .154 .295 .700 .825 .103 ,.001
Abbreviations: CI, confidence interval; JCS, Japan Coma Scale; mRS, modified Rankin Scale; OR, odds ratio; t-PA, tissue plasminogen activator. C statistic of this model was .808. *Consciousness level was scored by the JCS. yPhysical impairment was assessed using the mRS.
at home,25 we believe this factor would have little effect on our results. On the other hand, our data are not population based but composed of patients treated in hospitals that have administered t-PA to 1 or more patients during admission. It is, therefore, conceivable that patients who live too far away from a hospital and have travel times that render them ineligible for t-PA administration were not included in our study. Although our results indicate that distance did not affect whether patients who visited the sample hospitals would receive t-PA, further studies are needed to confirm this relationship to provide a clearer
understanding of regional issues. The Japanese government is currently developing a national-level database that will be available for analysis in the near future. Our findings showed a strong association between ambulance utilization and t-PA administration. The influence of ambulance utilization has 3 components: time before calling for an ambulance, time during transportation, and time between arrival and initial diagnosis or treatment. Arrival by ambulance has been shown to reduce the delay between arrival and initial diagnostic brain computed tomography.26,27 This may be because
FACTORS ASSOCIATED WITH USE OF t-PA
patients who arrive by ambulance avoid registration time on arrival at the hospital and the physician sees the patient immediately when receiving the ambulance26,27 or because hospital prenotification by the ambulance shortens preparation time for in-hospital diagnostic tests and treatments. In this way, rapid transportation by ambulance can shorten the time delay between stroke onset and treatment. Although the use of an ambulance may possibly shorten the transportation time, the influence of a shorter travel time may not be a major component to the total time delay as reported in previous studies26,27 and it is possible that the time spent in activities other than transportation may dilute the influence of distance.26,27 A sense of crisis can reduce the time delay from stroke onset to the decision to call emergency services, and ambulance utilization can partly reflect a situation in which the patient or someone around them has recognized the stroke.28,29 Recognition of stroke and its potential severity may, therefore, lead to an ambulance call. However, previous studies have reported that many patients wait to see if symptoms improve before deciding to go to the hospital or call for an ambulance.6,29,30 In this perspective, public education may increase t-PA administration by encouraging the early recognition of an emergency. In Australia, ambulance dispatches for stroke increased during their launch of the National Stroke Foundation stroke awareness campaign.31 In Japan, a television stroke education campaign was employed to improve public knowledge about early stroke symptoms.32 As ambulance use for patients with symptoms of suspected stroke is recommended, it may be necessary to further promote the importance of calling for an ambulance to the general public.3 Accordingly, the promotion of ambulance use without hesitation in all suspected stroke cases may be a critical step to improve the quality of stroke care. Although the Japanese government has provided guidelines for the use of ambulance services that include stroke-related symptoms,33 official announcements have also been made to discourage the unjustified and indiscriminate use of ambulances, which has increased in the recent years.34 Because of these seemingly contradictory instructions, people may hesitate to call for an ambulance— especially if their stroke symptoms are relatively minor—in an attempt to reduce this burden on ambulance services. However, our findings emphasize the need to ensure that patients are educated in the identification of possible stroke symptoms and promote unhesitating ambulance use for suspected stroke cases to improve prognoses. In Japan, as mentioned earlier, there are regional variations in t-PA administration.1 Although our study showed no association between the distance and t-PA administration, an in-depth investigation of regional disparities in stroke care requires further analyses to explore the
729
impact of patients whose residences are located far away from hospitals equipped to provide t-PA administration; such patients would be unlikely to be admitted to these hospitals and would, therefore, be unable to receive tPA treatment. Although the education of the general public is important to promote ambulance use for suspected stroke cases, other issues also have to be considered when developing and implementing new policies or quality improvement projects. These issues include the need for considerable improvements in the quality of care at the ambulance35,36 and hospital levels.23,37 Moreover, the processes of medical treatment for stroke patients in the hospital can change depending on whether there is an intention to administer t-PA.38 Systemic total improvement projects, such as the TLL Temple Foundation Stroke Project, have been shown to increase the chances for t-PA administration39 and may improve patient outcomes.25 The time limitation for t-PA administration eligibility has been widely discussed,40-42 and a European guideline has recently extended its recommendation from 3 hours to 4.5 hours.5 The data in this study reflect the situation in which the Japanese guidelines recommend a 3-hour limitation,2 but this time limitation has also been changed in Japan. Thus, future studies or comparisons between countries should take into account these discrepancies and further changes in the criteria. A limitation of the database used in this study is that it did not use random sampling of hospitals. However, the data were obtained from various types of hospitals in different regions in Japan. The use of such widely varied sources may reduce the degree of selection bias as the data would be able to account for variations because of the hospital types, sizes, practice patterns, and geographical factors.
Conclusions We analyzed a large administrative database to shed light on the factors associated with t-PA administration, although taking into account variations in patient casemix. Older patients were found to be associated with decreased t-PA administration, whereas increasingly severe stroke symptoms and ambulance utilization were associated with increased likelihood of t-PA administration. Our findings may underline a need for the promotion of ambulance use for suspected stroke patients to the general public. Although our study indicated that the distance between patient’s residence and hospital was not a significant factor, further population-based studies are needed to confirm this relationship.
References 1. Okada Y, Minematsu K, Ogawa A, et al. Nation-wide survey of use of intravenous rt-PA (alteplase) therapy during the first four years after approval—for overcoming
S. KUNISAWA ET AL.
730
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16. 17.
18. 19.
20.
regional gaps. Jpn J Stroke 2010;32:365-372 [in Japanese with English abstract]. Shinohara Y, Yanagihara T, Abe K, et al. II. Cerebral infarction/transient ischemic attack (TIA). J Stroke Cerebrovasc Dis 2011;20:S31-S73. Adams HP Jr, del Zoppo G, Alberts MJ, et al. Guidelines for the early management of adults with ischemic stroke: a guideline from the American Heart Association/American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups. Stroke 2007;38:1655-1711. Ringleb PA, Bousser MG, Ford G, et al. Guidelines for management of ischaemic stroke and transient ischaemic attack 2008. Cerebrovasc Dis 2008;25:457-507. The European Stroke Organization (ESO) Executive Committee and the ESO Writing Committee. Guidelines for management of ischaemic stroke and transient ischaemic attack 2008. 2009. Available at: http://www.eso-stroke. org/pdf/ESO_Guideline_Update_Jan_2009.pdf. Accessed April 30, 2013. Barber PA, Zhang J, Demchuk AM, et al. Why are stroke patients excluded from TPA therapy? An analysis of patient eligibility. Neurology 2001;56:1015-1020. Nagaraja N, Bhattacharya P, Norris G, et al. Arrival by ambulance is associated with acute stroke intervention in young adults. J Neurol Sci 2012;316:168-169. Acharya AB, Nyirenda JC, Higgs GB, et al. Distance from home to hospital and thrombolytic utilization for acute ischemic stroke. J Stroke Cerebrovasc Dis 2011;20:295-301. Srivastava AK, Prasad K. A study of factors delaying hospital arrival of patients with acute stroke. Neurol India 2001;49:272-276. Hong ES, Kim SH, Kim WY, et al. Factors associated with prehospital delay in acute stroke. Emerg Med J 2011; 28:790-793. Tanaka Y, Nakajima M, Hirano T, et al. Factors influencing pre-hospital delay after ischemic stroke and transient ischemic attack. Intern Med 2009;48:1739-1744. Center for Spatial Information Science. Geocoding tools & utilities. Available at: http://newspat.csis.u-tokyo.ac. jp/geocode/. Accessed April 30, 2013. Ohta T, Kikuchi H, Hashi K, et al. Nizofenone administration in the acute stage following subarachnoid hemorrhage. Results of a multi-center controlled double-blind clinical study. J Neurosurg 1986;64:420-426. The Joint Committee on Guidelines for the Management of Stroke for the English Version. Japanese guidelines for the management of stroke 2009: appendix. J Stroke Cerebrovasc Dis 2011;20:S181-S196. Van Swieten J, Koudstaal P, Visser M, et al. Interobserver agreement for the assessment of handicap in stroke patients. Stroke 1988;19:604-607. Hills NK, Johnston SC. Why are eligible thrombolysis candidates left untreated? Am J Prev Med 2006;31:S210-S216. Singer OC, Hamann GF, Misselwitz B, et al. Time trends in systemic thrombolysis in a large hospital-based stroke registry. Cerebrovasc Dis 2012;33:316-321. Chang KC, Tseng MC, Tan TY. Prehospital delay after acute stroke in Kaohsiung, Taiwan. Stroke 2004;35:700-704. Gargano JW, Wehner S, Reeves MJ. Presenting symptoms and onset-to-arrival time in patients with acute stroke and transient ischemic attack. J Stroke Cerebrovasc Dis 2011;20:494-502. De Silva DA, Ebinger M, Davis SM. Gender issues in acute stroke thrombolysis. J Clin Neurosci 2009;16:501-504.
21. Allen NB, Myers D, Watanabe E, et al. Utilization of intravenous tissue plasminogen activator for ischemic stroke: are there sex differences? Cerebrovasc Dis 2009; 27:254-258. 22. Turtzo LC, McCullough LD. Sex differences in stroke. Cerebrovasc Dis 2008;26:462-474. 23. Lichtman JH, Watanabe E, Allen NB, et al. Hospital arrival time and intravenous t-PA use in US Academic Medical Centers, 2001-2004. Stroke 2009;40:3845-3850. 24. Morinaga M, Ogita M, Kato M, et al. Shigaken no ichichiku ni okeru nousocchu no kyukyuhanso no jittai [Ambulance for ischemic stroke in a region of Shiga prefecture]. J Nursing, Shiga Univ Med Sci 2010;8:51-54 [in Japanese]. Available at: http://hdl.handle.net/10422/ 189. Accessed April 30, 2013. 25. Kimura K, Kazui S, Minematsu K, et al. Analysis of 16,922 patients with acute ischemic stroke and transient ischemic attack in Japan. A hospital-based prospective registration study. Cerebrovasc Dis 2004;18:47-56. 26. Morris DL, Rosamond W, Madden K, et al. Prehospital and emergency department delays after acute stroke: the Genentech Stroke Presentation Survey. Stroke 2000; 31:2585-2590. 27. Yoneda Y, Mori E, Uehara T, et al. Referral and care for acute ischemic stroke in a Japanese tertiary emergency hospital. Eur J Neurol 2001;8:483-488. 28. Iguchi Y, Wada K, Shibazaki K, et al. First impression at stroke onset plays an important role in early hospital arrival. Intern Med 2006;45:447-451. 29. Lundelin K, Graciani A, Garcia-Puig J, et al. Knowledge of stroke warning symptoms and intended action in response to stroke in Spain: a nationwide populationbased study. Cerebrovasc Dis 2012;34:161-168. 30. Palomeras E, Fossas P, Quintana M, et al. Emergency perception and other variables associated with extrahospital delay in stroke patients in the Maresme region (Spain). Eur J Neurol 2008;15:329-335. 31. Bray JE, Mosley I, Bailey M, et al. Stroke public awareness campaigns have increased ambulance dispatches for stroke in Melbourne, Australia. Stroke 2011;42:2154-2157. 32. Miyamatsu N, Kimura K, Okamura T, et al. Effects of public education by television on knowledge of early stroke symptoms among a Japanese population aged 40 to 74 years: a controlled study. Stroke 2012;43:545-549. 33. Fire and Disaster Management Agency. Making the most of the ambulance service—When do we need an ambulance? 2011 Available at: http://www.fdma.go.jp/html/ life/kyuukyuusya_manual/pdf/2011/english.pdf. Accessed April 30, 2013. 34. Fire and Disaster Management Agency. Kyukyusha no tekisei riyo ni tsuite (about the proper utilization of ambulances, Japanese). 2009 Available at: http://www. fdma.go.jp/ugoki/h1907/190702-1_30.pdf. Accessed April 30, 2013. 35. Mosley I, Nicol M, Donnan G, et al. The impact of ambulance practice on acute stroke care. Stroke 2007; 38:2765-2770. 36. Quain DA, Parsons MW, Loudfoot AR, et al. Improving access to acute stroke therapies: a controlled trial of organised pre-hospital and emergency care. Med J Aust 2008;189:429-433. 37. Van Wijngaarden JDH, Dirks M, Niessen LW, et al. Do centres with well-developed protocols, training and infrastructure have higher rates of thrombolysis for acute ischaemic stroke? QJM 2011;104:785-791. 38. Sato S, Uehara T, Toyoda K, et al. Impact of the approval of intravenous recombinant tissue plasminogen activator
FACTORS ASSOCIATED WITH USE OF t-PA therapy on the processes of acute stroke management in Japan: the Stroke Unit Multicenter Observational (SUMO) Study. Stroke 2009;40:30-34. 39. Morgenstern LB, Staub L, Chan W, et al. Improving delivery of acute stroke therapy: The TLL Temple Foundation Stroke Project. Stroke 2002;33:160-166. 40. Hacke W, Kaste M, Bluhmki E, et al. Thrombolysis with alteplase 3 to 4.5 hours after acute ischemic stroke. N Engl J Med 2008;359:1317-1329.
731 41. Del Zoppo GJ, Saver JL, Jauch EC, et al. Expansion of the time window for treatment of acute ischemic stroke with intravenous tissue plasminogen activator: a science advisory from the American Heart Association/American Stroke Association. Stroke 2009;40:2945-2948. 42. Shobha N, Buchan AM, Hill MD, et al. Thrombolysis at 34.5 hours after acute ischemic stroke onset—evidence from the Canadian Alteplase for Stroke Effectiveness Study (CASES) registry. Cerebrovasc Dis 2011;31:223-228.
Background: The use of intravenous tissue plasminogen activator (t-PA) can be an effective treatment for acute ischemic stroke if administered promptly. Despite its clinical effectiveness, overall use in Japan remains low, and regional variations have been reported. Factors such as ambulance utilization and geographical distance from patients’ residences to hospitals may influence t-PA administration rates. The aim of this study is to identify factors associated with the administration of t-PA for acute ischemic stroke while adjusting for casemix using a large-scale administrative database in Japan. Methods: We analyzed acute ischemic stroke patients admitted to acute care hospitals between July 2010 and March 2011 using a nationwide database. A logistic regression model was used to analyze the factors influencing t-PA administration. Candidate factors included patient gender, age, stroke severity, direct distance between each patient’s residence and admitting hospital, and ambulance utilization. Results: Of the 10,615 ischemic stroke patients from 89 hospitals analyzed, 557 (5.2%) received t-PA treatment. Patients aged 75 years and older were found to be associated with decreased t-PA administration. In contrast, severe stroke and ambulance utilization were associated with increased t-PA administration. Distance was not significantly associated with the use of t-PA. Conclusions: Our findings suggest that ambulance utilization is an important factor for improving the likelihood of t-PA administration in patients with stroke and may underline a need for educational programs to the general public that promote the use of ambulances for suspected stroke patients. Key Words: Stroke—tissue plasminogen activator— health services accessibility—ambulances—risk adjustment. Ó 2014 by National Stroke Association
Introduction The use of intravenous tissue plasminogen activator (t-PA) was approved as a treatment for acute ischemic stroke in Japan in October 2005. Although utilization rates
From the *Department of Healthcare Economics and Quality Management, Graduate School of Medicine, Kyoto University, Kyoto, Japan; †Department of Cerebrovascular Medicine, National Cerebral and Cardiovascular Center, Suita, Osaka, Japan; and ‡National Cerebral and Cardiovascular Center, Suita, Osaka, Japan. Received May 10, 2013; revision received June 20, 2013; accepted June 21, 2013. Sources of funding: This study was supported in part by a Health and Labour Sciences Research Grant from the Ministry of Health, Labour and Welfare, Japan; an Intramural Research Grant for Cardio-
724
have gradually increased after its first application, the overall use remains low, and regional variations have been reported.1 Opportunities to administer t-PA are limited as guidelines recommend the use of intravenous t-PA
vascular Diseases from the National Cerebral and Cardiovascular Center; and a grant-in-aid for Scientific Research from the Japan Society for the Promotion of Science. Address correspondence to Yuichi Imanaka, MD, PhD, Department of Healthcare Economics and Quality Management, Graduate School of Medicine, Kyoto University, Yoshida Konoe-cho, Sakyo-ku, Kyoto City, Kyoto 606-8501, Japan. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2014 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2013.06.033
Journal of Stroke and Cerebrovascular Diseases, Vol. 23, No. 4 (April), 2014: pp 724-731
FACTORS ASSOCIATED WITH USE OF t-PA
only to patients within a few hours of acute ischemic stroke onset.2-5 Because the use of t-PA is dependent on the elapsed time after stroke onset, emergency medical service infrastructure and geography are important factors to consider in an analysis of this treatment. Variations in these factors can directly affect the time duration from onset to treatment and may result in regional health disparities that need to be addressed from a public health perspective. Although the factors that influence t-PA administration have been previously investigated, these studies have generally been conducted using relatively small sample sizes.6-8 A previous study has reported an association between the use of ambulance transportation and acute stroke intervention in young adults,7 with the stated caveat that the limitations of small sample size and lack of stroke severity data may have affected the observed results. Additionally, the association between the distance from patient’s residence to hospital and t-PA utilization has also been reported.8-10 However, the results concerning the relationship between patient’s travel distance and delayed hospital arrival are inconclusive as other researchers have reported no association between these 2 factors.11 Understanding the effects of ambulance use and geographical distance between patient’s residence and hospital on the likelihood of t-PA use is essential to improve stroke care from political and social perspectives. To do so, an analysis based on a relatively large sample size comprising numerous hospitals of different sizes and infrastructures that are located throughout various regions would yield results with greater external validity. Since the introduction of the hospital reimbursement system known as the Diagnosis Procedure Combination (DPC) per-diem payment system in Japan in 2003, hospitals using this system routinely generate the DPC data for each patient per hospitalization for reimbursement purposes. From July 2010 onward, the DPC per-diem payment system has required the collection of additional information such as patient’s residence postal code, neurologic deficits, and date of stroke onset for patients admitted for stroke. These data are uniformly formatted, which allows researchers to collect, manage, and analyze large quantities of administrative data from numerous hospitals located nationwide. Furthermore, the recent inclusion of factors reflecting disease severity such as patient consciousness and physical impairment levels in this database supports analyses that take into account critical differences in patient casemix. An extensive investigation of the associations between t-PA use and the factors of ambulance use and geographical distance from patient’s residence to hospital in Japan has not yet been conducted. Therefore, the objective of this study was to investigate the influence of these and other factors on t-PA administration using a large administration database in Japan.
725
Methods Study Design and Setting We used a nationwide database maintained by the government and public research funds. This database comprises data from more than 300 hospitals distributed across Japan and includes a variety of public and private hospitals, with varying teaching statuses, case volumes, casemix, and specialties. In addition to hospital identifier codes, patient identifier codes, and DPC codes, the DPC data also include items such as principal diagnosis and other medical and demographic information such as gender, date of ischemic stroke onset, consciousness and physical impairment levels at admission, postal code of patient’s residence, and information regarding whether the patient arrived at the hospital by ambulance. The postal code for each patient’s residence and the address for the hospital to which they were admitted were geocoded to longitudinal and latitudinal co-ordinates using MarketPlanner 2.5.1 (PASCO Corporation, Tokyo, Japan) and Geocoding Tools & Utilities (Center for Spatial Information Science, Tokyo, Japan)12; the direct distances between the 2 were then calculated. We estimated the average error of the distance when using the postal code as a substitute for patients’ addresses to be less than 1 km, by applying the following formula for each city: sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðinhabitable areaÞ 3:14 3 ðnumber of postal codesÞ The average radius of each postal code was calculated to be .68 km (6.34 SD).
Data Selection Using the DPC database, we identified and extracted acute ischemic stroke cases discharged between July 2010 and March 2011. The criteria for identification of acute ischemic stroke were cases with (1) ‘‘010060’’ as the first 6 digits of the 14-digit DPC code; (2) an International Classification of Diseases, Tenth Revision, ‘‘I63$’’ as the principal diagnosis code; and (3) stroke onset occurring either on the day before or on the day of admission. We limited the analysis to patients residing within 50 km from the admitting hospital. Cases that were administered t-PA were identified by the number ‘‘4’’ in the 12th digit of the DPC code. Because not all hospitals are equipped to provide t-PA treatment,2 we further limited our analysis to include only hospitals that had administered t-PA to 1 or more patients during the study period. Only cases with complete information for the variables of interest were retained for analysis. Consequently, a total of 10,615 cases from 89 hospitals were included in the analysis.
S. KUNISAWA ET AL.
726
Levels of Consciousness and Physical Impairment The National Institutes of Health Stroke Scale is used in several countries to assess the severity of stroke patients. However, the DPC data do not include this information. Instead, we used a combination of consciousness and physical impairment levels. Consciousness levels were determined using the Japan Coma Scale (JCS). The JCS consists of 10 levels of consciousness, which can be categorized into the following 4 grades: grade 0 (alert); grade 1 (awake without any stimuli), comprising JCS levels 1-3; grade 2 (arousable), comprising JCS levels 10-30; and grade 3 (unarousable), comprising JCS levels 100300.13,14 In this study, JCS was included in analysis using these 4 grades. The level of physical impairment was determined using the modified Rankin Scale (mRS) scoring system. The mRS is principally an indicator of activities of daily living concerned with disability and dependence levels of patients after stroke.15 The DPC data include the mRS scores of patients at admission with any type of stroke because of the relative ease of assessment and the need for a common scale among the various types of stroke under the uniformly formatted DPC database. In this context, the mRS for acute stroke at admission assesses the level of physical impairment at the point of admission. Although the mRS is scored from 0 to 6, we selected patients with an mRS score of 1 or more and categorized them into 3 classes for our analysis: 1-2 (mild), 3-4 (moderate), and 5 (severe) as we assumed that these combined classes would allow for more distinct differentiation of the levels of severity. Finally, we created the following 8 patient categories for the analysis based on the paired JCS grades and mRS classes. The categories were (1) JCS grade 0 and mild mRS score; (2) JCS grade 0 and moderate mRS score; (3) JCS grade 0 and severe mRS score; (4) JCS grade 1 and mild mRS score; (5) JCS grade 1 and moderate mRS score; (6) JCS grade 1 and severe mRS score; (7) JCS grade 2; and (8) JCS grade 3. Because of the difficulties in assessing the mRS score levels of comatose patients, the highest 2 JCS grades were not paired with any specific mRS class.
Statistical Analysis We used multivariable logistic regression analysis to identify independent factors associated with t-PA administration. Independent variables in the regression model included patient gender, age, stroke severity, direct distance between each patient’s residence and admitting hospital, and ambulance utilization. We analyzed distance as an ordinal variable using the category with the shortest distance as the reference group to test the hypothesis that distance could affect t-PA administration after a certain threshold value. Statistical tests were 2 tailed and associations were considered significant at P less than .05. Statistical analyses were conducted using
SPSS software, version 19.0.0.2 (IBM, Tokyo, Japan). Discrimination of the regression model was evaluated with the c statistic.
Ethical Standard This study was approved by the Ethics Committee, Kyoto University Graduate School and Faculty of Medicine.
Results Characteristics of Study Subjects The baseline characteristics and medical information of the 10,615 patients are presented in Table 1. The mean age of the patients in the sample was 74.8 6 11.9 years, and 41.3% of the patients were women. Of the patients in the sample, 557 (5.2%) received t-PA during admission. These 557 patients who received t-PA treatment constitute approximately 9% of all stroke patients in Japan who had received t-PA treatment during the study period, which is estimated to be approximately 6400 patients (T. Ueno, Mitsubishi-Tanabe Pharmaceutical Corporation, written communication, May 2012). The unadjusted rate of t-PA administration was higher in patients with JCS grade 2 more severe mRS scores, and those who arrived by ambulance.
Factors Associated with t-PA Administration Table 2 shows the results of the multivariable logistic analysis. Patients aged 75 years or older were less likely to receive t-PA when compared with patients younger than 65 years. In contrast, administration of t-PA was more likely in patients with a medical status more severe than JCS grade 0 and a mild mRS score. Increasingly severe stroke symptoms were associated with increased t-PA administration, except for the most severe cases of JCS grade 3; the likelihood of t-PA administration for the cases of JCS grade 3 was lower than that of JCS grade 2. The direct distance between each patient’s residence and the admitting hospital did not show a significant association with t-PA administration, but arrival by ambulance was significantly associated with a substantial increase in the likelihood of t-PA administration (odds ratio, 4.31; 95% confidence interval, 3.27-5.69). The discrimination of the regression model was high, with a c statistic of .808.
Discussion By analyzing the administrative data of 10,615 cases from 89 hospitals in Japan, our study showed that 557 (5.2%) acute ischemic stroke patients received t-PA while hospitalized. Elderly patients were found to be associated with decreased t-PA administration. However, increasingly severe stroke symptoms (except for the most severe
FACTORS ASSOCIATED WITH USE OF t-PA
727
Table 1. Characteristics, medical information, and administration of t-PA of acute ischemic stroke patients t-PA administration Variables
Yes
%
Total
Gender Male 321 5.2 6226 Female 236 5.4 4389 Age (y) ,65 111 5.2 2139 65-74 138 5.2 2670 75-84 215 5.8 3704 .85 93 4.4 2102 Consciousness level at admission* Grade 0 (alert) 93 1.8 5187 Grade 1 (awake) 270 6.9 3920 Grade 2 (arousable) 147 14.7 997 Grade 3 (unarousable) 47 9.2 511 Physical impairment at admissiony Mild (1-2) 36 1.1 3249 Moderate (3-4) 216 4.5 4788 Severe (5) 305 11.8 2578 Direct distance from patient home to the hospital ,2 km 113 5.5 2054 2-3 km 139 5.2 2671 4-5 km 83 4.8 1728 6-7 km 55 5.0 1099 8-9 km 38 4.7 808 10-14 km 60 5.3 1141 15-19 km 37 7.2 513 20-29 km 26 6.3 412 30-49 km 6 3.2 189 Arrival by ambulance No 62 1.2 4986 Yes 495 8.8 5629 Total 557 5.2 10,615 Abbreviations: JCS, Japan Coma Scale; mRS, modified Rankin Scale; t-PA, tissue plasminogen activator. *Consciousness level was scored by the JCS. yPhysical impairment was assessed using the mRS.
cases) and ambulance utilization were associated with increased t-PA administration. On the surface, it may seem that the utilization rate of tPA in this sample is low. However, it should be noted that simple comparisons with the utilization rates in previous studies may not be appropriate as study samples can differ between analyses.6-8,16 In this study, we analyzed acute ischemic stroke patients, all of whom are potentially eligible for t-PA administration, although there may have been some patients who were contraindicated for this treatment. When compared with younger patients, administration of t-PA was found to be less likely in patients older than 75 years. The Japanese guidelines advise cautious administration to patients older than 75 years, and the European guidelines recommend administration to patients older
4,5
than 80 years only if they fulfill certain criteria. Our findings were consistent with these indications but in contrast with those observed in a German study.17 On the other hand, aging has also been found to be associated with delays in seeking medical help and time to hospital arrival.18,19 In this context, our results may also partly reflect this delay caused by aging. Gender had no association with the likelihood of t-PA administration in our analysis, but this is inconsistent among previous studies.7,20-22 Our findings showed that consciousness and physical impairment levels were associated with t-PA administration, which is also consistent with recommendations in guidelines. Guidelines do not recommend the use of t-PA treatment for patients with minor neurologic symptoms.2,3 This was compatible with our findings that patients with more severe stroke symptoms were more likely to receive t-PA when compared with patients with minor symptoms. On the other hand, guidelines also recommend caution in t-PA use for patients with major neurologic deficits, such as the JCS grade 3 (patients in an unarousable coma)2,3; accordingly, the odds ratio for t-PA administration was correspondingly lower for patients with the JCS grade 3 relative to the JCS grade 2 (patients in an arousable coma) in this study. The association between severity of stroke symptoms and tPA administration may in part be indicative of how quickly the patients or their families acted in recognizing a stroke and seeking appropriate care. As previous authors have shown, an increase in the severity of stroke symptoms is associated with shorter time to treatment.3,10,18,19,23,24 The observed influences of the severity of stroke symptoms on t-PA administration may, therefore, have multiple interpretations. Distance did not show a significant association with t-PA administration in our sample. The variations in results regarding this association in previous studies were likely to be dependent on factors such as social and health-care systems, geography, and traffic networks such as highway systems. Moreover, although we employed direct distance for our analysis, actual driving times may differ. However, a previous study in the Kumamoto region of Japan showed that the direct distance was significantly correlated with the transportation time by ambulance.11 Nonetheless, it may be important for future analyses to take into account the hour of the day, the day of the week, and the traffic density to understand regional situations better. As there are many mountainous regions in Japan, driving times may also be affected depending on whether patient routes include passages through flatlands or mountain corridors. Additionally, we analyzed the direct distance from each patient’s residence to the admitting hospital. Because stroke onset does not always occur at home, the location of onset may introduce a degree of bias to actual distances traveled. However, as a previous study has shown that 79.5% of stroke onset occurs
S. KUNISAWA ET AL.
728
Table 2. Results of a multiple logistic regression model with t-PA administration for acute ischemic stroke patients as the dependent variable
Independent variables Gender Male (reference) Female Age (y) ,65 (reference) 65-74 75-84 .85
OR (95% CI)
P value
.98 (.81-1.19)
.875
.78 (.59-1.02) .65 (.50-.83) .33 (.24-.45)
.068 .001 ,.001
Consciousness level and physical impairment level at admission Consciousness*
Physical impairmenty
Grade 0 (alert) Mild (reference) Moderate Severe
.08 (.04-.15) .25 (.16-.38)
,.001 ,.001
1.05 (.64-1.72) .26 (.15-.46) .73 (.51-1.05) 1.30 (.91-1.87) 1.72 (1.21-2.45)
.851 ,.001 .091 .146 .003
Grade 1 (awake) Mild Moderate Severe Grade 2 (arousable) Grade 3 (unarousable)
Direct distance between patient home and the hospital ,2 km (reference) 2-3 km 4-5 km 6-7 km 8-9 km 10-14 km 15-19 km 20-29 km 30-49 km Arrival by ambulance
.91 (.69-1.18) .78 (.57-1.05) .79 (.56-1.12) .75 (.51-1.11) .84 (.60-1.17) 1.08 (.72-1.62) .95 (.60-1.50) .49 (.21-1.16) 4.31 (3.27-5.69)
.469 .098 .182 .154 .295 .700 .825 .103 ,.001
Abbreviations: CI, confidence interval; JCS, Japan Coma Scale; mRS, modified Rankin Scale; OR, odds ratio; t-PA, tissue plasminogen activator. C statistic of this model was .808. *Consciousness level was scored by the JCS. yPhysical impairment was assessed using the mRS.
at home,25 we believe this factor would have little effect on our results. On the other hand, our data are not population based but composed of patients treated in hospitals that have administered t-PA to 1 or more patients during admission. It is, therefore, conceivable that patients who live too far away from a hospital and have travel times that render them ineligible for t-PA administration were not included in our study. Although our results indicate that distance did not affect whether patients who visited the sample hospitals would receive t-PA, further studies are needed to confirm this relationship to provide a clearer
understanding of regional issues. The Japanese government is currently developing a national-level database that will be available for analysis in the near future. Our findings showed a strong association between ambulance utilization and t-PA administration. The influence of ambulance utilization has 3 components: time before calling for an ambulance, time during transportation, and time between arrival and initial diagnosis or treatment. Arrival by ambulance has been shown to reduce the delay between arrival and initial diagnostic brain computed tomography.26,27 This may be because
FACTORS ASSOCIATED WITH USE OF t-PA
patients who arrive by ambulance avoid registration time on arrival at the hospital and the physician sees the patient immediately when receiving the ambulance26,27 or because hospital prenotification by the ambulance shortens preparation time for in-hospital diagnostic tests and treatments. In this way, rapid transportation by ambulance can shorten the time delay between stroke onset and treatment. Although the use of an ambulance may possibly shorten the transportation time, the influence of a shorter travel time may not be a major component to the total time delay as reported in previous studies26,27 and it is possible that the time spent in activities other than transportation may dilute the influence of distance.26,27 A sense of crisis can reduce the time delay from stroke onset to the decision to call emergency services, and ambulance utilization can partly reflect a situation in which the patient or someone around them has recognized the stroke.28,29 Recognition of stroke and its potential severity may, therefore, lead to an ambulance call. However, previous studies have reported that many patients wait to see if symptoms improve before deciding to go to the hospital or call for an ambulance.6,29,30 In this perspective, public education may increase t-PA administration by encouraging the early recognition of an emergency. In Australia, ambulance dispatches for stroke increased during their launch of the National Stroke Foundation stroke awareness campaign.31 In Japan, a television stroke education campaign was employed to improve public knowledge about early stroke symptoms.32 As ambulance use for patients with symptoms of suspected stroke is recommended, it may be necessary to further promote the importance of calling for an ambulance to the general public.3 Accordingly, the promotion of ambulance use without hesitation in all suspected stroke cases may be a critical step to improve the quality of stroke care. Although the Japanese government has provided guidelines for the use of ambulance services that include stroke-related symptoms,33 official announcements have also been made to discourage the unjustified and indiscriminate use of ambulances, which has increased in the recent years.34 Because of these seemingly contradictory instructions, people may hesitate to call for an ambulance— especially if their stroke symptoms are relatively minor—in an attempt to reduce this burden on ambulance services. However, our findings emphasize the need to ensure that patients are educated in the identification of possible stroke symptoms and promote unhesitating ambulance use for suspected stroke cases to improve prognoses. In Japan, as mentioned earlier, there are regional variations in t-PA administration.1 Although our study showed no association between the distance and t-PA administration, an in-depth investigation of regional disparities in stroke care requires further analyses to explore the
729
impact of patients whose residences are located far away from hospitals equipped to provide t-PA administration; such patients would be unlikely to be admitted to these hospitals and would, therefore, be unable to receive tPA treatment. Although the education of the general public is important to promote ambulance use for suspected stroke cases, other issues also have to be considered when developing and implementing new policies or quality improvement projects. These issues include the need for considerable improvements in the quality of care at the ambulance35,36 and hospital levels.23,37 Moreover, the processes of medical treatment for stroke patients in the hospital can change depending on whether there is an intention to administer t-PA.38 Systemic total improvement projects, such as the TLL Temple Foundation Stroke Project, have been shown to increase the chances for t-PA administration39 and may improve patient outcomes.25 The time limitation for t-PA administration eligibility has been widely discussed,40-42 and a European guideline has recently extended its recommendation from 3 hours to 4.5 hours.5 The data in this study reflect the situation in which the Japanese guidelines recommend a 3-hour limitation,2 but this time limitation has also been changed in Japan. Thus, future studies or comparisons between countries should take into account these discrepancies and further changes in the criteria. A limitation of the database used in this study is that it did not use random sampling of hospitals. However, the data were obtained from various types of hospitals in different regions in Japan. The use of such widely varied sources may reduce the degree of selection bias as the data would be able to account for variations because of the hospital types, sizes, practice patterns, and geographical factors.
Conclusions We analyzed a large administrative database to shed light on the factors associated with t-PA administration, although taking into account variations in patient casemix. Older patients were found to be associated with decreased t-PA administration, whereas increasingly severe stroke symptoms and ambulance utilization were associated with increased likelihood of t-PA administration. Our findings may underline a need for the promotion of ambulance use for suspected stroke patients to the general public. Although our study indicated that the distance between patient’s residence and hospital was not a significant factor, further population-based studies are needed to confirm this relationship.
References 1. Okada Y, Minematsu K, Ogawa A, et al. Nation-wide survey of use of intravenous rt-PA (alteplase) therapy during the first four years after approval—for overcoming
S. KUNISAWA ET AL.
730
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16. 17.
18. 19.
20.
regional gaps. Jpn J Stroke 2010;32:365-372 [in Japanese with English abstract]. Shinohara Y, Yanagihara T, Abe K, et al. II. Cerebral infarction/transient ischemic attack (TIA). J Stroke Cerebrovasc Dis 2011;20:S31-S73. Adams HP Jr, del Zoppo G, Alberts MJ, et al. Guidelines for the early management of adults with ischemic stroke: a guideline from the American Heart Association/American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups. Stroke 2007;38:1655-1711. Ringleb PA, Bousser MG, Ford G, et al. Guidelines for management of ischaemic stroke and transient ischaemic attack 2008. Cerebrovasc Dis 2008;25:457-507. The European Stroke Organization (ESO) Executive Committee and the ESO Writing Committee. Guidelines for management of ischaemic stroke and transient ischaemic attack 2008. 2009. Available at: http://www.eso-stroke. org/pdf/ESO_Guideline_Update_Jan_2009.pdf. Accessed April 30, 2013. Barber PA, Zhang J, Demchuk AM, et al. Why are stroke patients excluded from TPA therapy? An analysis of patient eligibility. Neurology 2001;56:1015-1020. Nagaraja N, Bhattacharya P, Norris G, et al. Arrival by ambulance is associated with acute stroke intervention in young adults. J Neurol Sci 2012;316:168-169. Acharya AB, Nyirenda JC, Higgs GB, et al. Distance from home to hospital and thrombolytic utilization for acute ischemic stroke. J Stroke Cerebrovasc Dis 2011;20:295-301. Srivastava AK, Prasad K. A study of factors delaying hospital arrival of patients with acute stroke. Neurol India 2001;49:272-276. Hong ES, Kim SH, Kim WY, et al. Factors associated with prehospital delay in acute stroke. Emerg Med J 2011; 28:790-793. Tanaka Y, Nakajima M, Hirano T, et al. Factors influencing pre-hospital delay after ischemic stroke and transient ischemic attack. Intern Med 2009;48:1739-1744. Center for Spatial Information Science. Geocoding tools & utilities. Available at: http://newspat.csis.u-tokyo.ac. jp/geocode/. Accessed April 30, 2013. Ohta T, Kikuchi H, Hashi K, et al. Nizofenone administration in the acute stage following subarachnoid hemorrhage. Results of a multi-center controlled double-blind clinical study. J Neurosurg 1986;64:420-426. The Joint Committee on Guidelines for the Management of Stroke for the English Version. Japanese guidelines for the management of stroke 2009: appendix. J Stroke Cerebrovasc Dis 2011;20:S181-S196. Van Swieten J, Koudstaal P, Visser M, et al. Interobserver agreement for the assessment of handicap in stroke patients. Stroke 1988;19:604-607. Hills NK, Johnston SC. Why are eligible thrombolysis candidates left untreated? Am J Prev Med 2006;31:S210-S216. Singer OC, Hamann GF, Misselwitz B, et al. Time trends in systemic thrombolysis in a large hospital-based stroke registry. Cerebrovasc Dis 2012;33:316-321. Chang KC, Tseng MC, Tan TY. Prehospital delay after acute stroke in Kaohsiung, Taiwan. Stroke 2004;35:700-704. Gargano JW, Wehner S, Reeves MJ. Presenting symptoms and onset-to-arrival time in patients with acute stroke and transient ischemic attack. J Stroke Cerebrovasc Dis 2011;20:494-502. De Silva DA, Ebinger M, Davis SM. Gender issues in acute stroke thrombolysis. J Clin Neurosci 2009;16:501-504.
21. Allen NB, Myers D, Watanabe E, et al. Utilization of intravenous tissue plasminogen activator for ischemic stroke: are there sex differences? Cerebrovasc Dis 2009; 27:254-258. 22. Turtzo LC, McCullough LD. Sex differences in stroke. Cerebrovasc Dis 2008;26:462-474. 23. Lichtman JH, Watanabe E, Allen NB, et al. Hospital arrival time and intravenous t-PA use in US Academic Medical Centers, 2001-2004. Stroke 2009;40:3845-3850. 24. Morinaga M, Ogita M, Kato M, et al. Shigaken no ichichiku ni okeru nousocchu no kyukyuhanso no jittai [Ambulance for ischemic stroke in a region of Shiga prefecture]. J Nursing, Shiga Univ Med Sci 2010;8:51-54 [in Japanese]. Available at: http://hdl.handle.net/10422/ 189. Accessed April 30, 2013. 25. Kimura K, Kazui S, Minematsu K, et al. Analysis of 16,922 patients with acute ischemic stroke and transient ischemic attack in Japan. A hospital-based prospective registration study. Cerebrovasc Dis 2004;18:47-56. 26. Morris DL, Rosamond W, Madden K, et al. Prehospital and emergency department delays after acute stroke: the Genentech Stroke Presentation Survey. Stroke 2000; 31:2585-2590. 27. Yoneda Y, Mori E, Uehara T, et al. Referral and care for acute ischemic stroke in a Japanese tertiary emergency hospital. Eur J Neurol 2001;8:483-488. 28. Iguchi Y, Wada K, Shibazaki K, et al. First impression at stroke onset plays an important role in early hospital arrival. Intern Med 2006;45:447-451. 29. Lundelin K, Graciani A, Garcia-Puig J, et al. Knowledge of stroke warning symptoms and intended action in response to stroke in Spain: a nationwide populationbased study. Cerebrovasc Dis 2012;34:161-168. 30. Palomeras E, Fossas P, Quintana M, et al. Emergency perception and other variables associated with extrahospital delay in stroke patients in the Maresme region (Spain). Eur J Neurol 2008;15:329-335. 31. Bray JE, Mosley I, Bailey M, et al. Stroke public awareness campaigns have increased ambulance dispatches for stroke in Melbourne, Australia. Stroke 2011;42:2154-2157. 32. Miyamatsu N, Kimura K, Okamura T, et al. Effects of public education by television on knowledge of early stroke symptoms among a Japanese population aged 40 to 74 years: a controlled study. Stroke 2012;43:545-549. 33. Fire and Disaster Management Agency. Making the most of the ambulance service—When do we need an ambulance? 2011 Available at: http://www.fdma.go.jp/html/ life/kyuukyuusya_manual/pdf/2011/english.pdf. Accessed April 30, 2013. 34. Fire and Disaster Management Agency. Kyukyusha no tekisei riyo ni tsuite (about the proper utilization of ambulances, Japanese). 2009 Available at: http://www. fdma.go.jp/ugoki/h1907/190702-1_30.pdf. Accessed April 30, 2013. 35. Mosley I, Nicol M, Donnan G, et al. The impact of ambulance practice on acute stroke care. Stroke 2007; 38:2765-2770. 36. Quain DA, Parsons MW, Loudfoot AR, et al. Improving access to acute stroke therapies: a controlled trial of organised pre-hospital and emergency care. Med J Aust 2008;189:429-433. 37. Van Wijngaarden JDH, Dirks M, Niessen LW, et al. Do centres with well-developed protocols, training and infrastructure have higher rates of thrombolysis for acute ischaemic stroke? QJM 2011;104:785-791. 38. Sato S, Uehara T, Toyoda K, et al. Impact of the approval of intravenous recombinant tissue plasminogen activator
FACTORS ASSOCIATED WITH USE OF t-PA therapy on the processes of acute stroke management in Japan: the Stroke Unit Multicenter Observational (SUMO) Study. Stroke 2009;40:30-34. 39. Morgenstern LB, Staub L, Chan W, et al. Improving delivery of acute stroke therapy: The TLL Temple Foundation Stroke Project. Stroke 2002;33:160-166. 40. Hacke W, Kaste M, Bluhmki E, et al. Thrombolysis with alteplase 3 to 4.5 hours after acute ischemic stroke. N Engl J Med 2008;359:1317-1329.
731 41. Del Zoppo GJ, Saver JL, Jauch EC, et al. Expansion of the time window for treatment of acute ischemic stroke with intravenous tissue plasminogen activator: a science advisory from the American Heart Association/American Stroke Association. Stroke 2009;40:2945-2948. 42. Shobha N, Buchan AM, Hill MD, et al. Thrombolysis at 34.5 hours after acute ischemic stroke onset—evidence from the Canadian Alteplase for Stroke Effectiveness Study (CASES) registry. Cerebrovasc Dis 2011;31:223-228.