ACADEMIC EMERGENCY MEDICINE 2011; 18:273–278 © 2011 by the Society for Academic Emergency Medicine
Objectives: The rate of recombinant tissue plasminogen activator (rtPA) use for stroke is low among the Chinese-Taiwanese. The study objective was to determine if less restrictive exclusion criteria for rtPA would increase eligibility and the rate of treatment.
Methods: This retrospective before-and-after study was conducted from 2006 to 2009. The authors compared stroke patients treated under the old rtPA exclusion criteria (January 2006 to December 2008) with those treated with less restrictive rtPA exclusion criteria (January to December 2009). Rates of eligibility and treatment and reasons for exclusion from rtPA between the two phases were assessed.
Results: Of 461 eligible patients during the study period, 333 were evaluated by the old criteria and 128 were evaluated by the less restrictive criteria. Minor or improving stroke was the primary exclusion reason in both groups (194/333, 58% vs. 70/128, 55%). Eligibility for rtPA was increased in the less restrictive criteria (35/127, 27%, 95% confidence interval [CI] = 20% to 36%) compared to the old criteria (40/333, 12%, 95% CI = 8.7% to 16%; p = 0.0001). Fewer patients were excluded due to old age in the less restrictive criteria (0/128, 0%) compared to the old criteria (37/333, 11%; p = 0.0001). The rate of consent refusal increased in the less restrictive criteria (27/128, 21%, 95% CI = 14% to 29%) compared to the old criteria (23/333, 6.9%, 95% CI = 4.4% to 10%; p < 0.0001). Rate of rtPA treatment was unchanged between the less restrictive criteria (8/128, 6.3%, 95% CI = 2.7% to 12%) and the old criteria (17/333, 5.1%, 95% CI = 3% to 8%; p = 0.63).
Conclusions: Increasing eligibility for rtPA does not increase the rate of treatment, possibly due to the high symptomatic intracerebral hemorrhage rate among Chinese-Taiwanese, which is a major concern among emergency physicians (EPs), neurologists, and patients. Dealing with perceived safety issues of rtPA is crucial before the rate of treatment can be increased.
Recombinant tissue plasminogen activator (rtPA) is currently the only approved agent for treating acute ischemic stroke.1–3 However, the usefulness of rtPA is largely limited by its narrow time window, which fails to provide significant impact on the overall ischemic stroke population.4–6 Even if patients present within 3 hours of stroke onset, the rates of eligibility and treatment with rtPA remain low.7–9 The exclusion criteria used, modified from the National Institute of Neurological Disorders and Stroke (NINDS) rtPA trial, are strict and exclude a large number of ischemic stroke patients who arrive at the hospital within 3 hours of onset. Based on new evidence from the most recent studies, less restrictive exclusion criteria were released by the Taiwan Stroke Society in November 2008.
A detailed understanding of the eligibility, rate of treatment, and reasons for exclusion from rtPA might contribute to increasing its use and provide useful information for emergency physicians (EPs) and neurologists in managing acute ischemic stroke patients who arrive at the hospital within 3 hours of stroke onset. Thus, this study aimed to determine whether less restrictive exclusion criteria for rtPA would increase the eligibility and rate of treatment. This study hypothesized that using less restrictive rtPA eligibility criteria for stroke patients would increase rtPA eligibility by over 10%.
This was a retrospective analysis of prospectively collected data studying the eligibility, rate of treatment, and reasons for exclusion from rtPA under the different exclusion criteria. Using a before-and-after study design, data were compared between the two phases. The local ethics committee approved this study.
Study Setting and Population
An intrahospital stroke code team was created in 2004 in the study center, which has a catchment area of 1.5 million inhabitants in the metropolitan area. Patients with signs and symptoms suggestive of acute ischemic stroke who arrived at the hospital within 3 hours of onset were enrolled. Their data were retrieved from the database, and patients with final discharge diagnoses other than ischemic stroke or transient ischemic attack were excluded. Only data from the period January 2006 to December 2009 were analyzed. Data from 2004–2005 were presented in a prior study.9
A trained neurologist with brain computed tomography (CT) or magnetic resonance imaging findings established the diagnosis of ischemic stroke. Reasons for exclusion from rtPA and the associated variables, including sex, age, time of symptom onset, hospital arrival, neurologist arrival, blood extraction, brain CT and electrocardiogram results, laboratory data, and National Institutes of Health Stroke Scale (NIHSS) score were prospectively registered. A stroke study nurse completed the standardized study form and the data were entered into a computerized database. A second study nurse reviewed the entered data for quality assurance. A review meeting was also held monthly to discuss and decide whether treatment or nontreatment was appropriate.
The modified exclusion criteria from the NINDS rtPA trial were used to evaluate the eligibility for rtPA from 2004 (see Data Supplement S1, available as supporting information in the online version of this paper). In November 2008, less restrictive exclusion criteria based on latest studies were released by the Taiwan Stroke Society (Table 1).10–16 The major change made was that of the definition of minor stroke, which was lowered from NIHSS <6 to <4. In addition, severe stroke (NIHSS >25) and old age (>80 years) were no longer absolute contraindications in using rtPA. Overall, the new exclusion criteria were less restrictive than the old criteria.
|Exclusion Criteria||Less Restrictive Criteria (November 2008)|
|Age <18 or >80 years||Decided by neurologist after discussion with patient and family|
|Minor (NIHSS <6) or rapidly improving stroke||Minor (NIHSS <4) or rapidly improving stroke|
|Severe stroke (NIHSS >25)||Decided by neurologist after discussion with patient and family|
|Seizure at onset of stroke||Not contraindicated if postictal paralysis was excluded|
|Patient has had major surgery or serious trauma within 10 days||Patient has had major surgery or serious trauma within 14 days|
|Prolonged or traumatic CPR (>2 minutes), delivery, or recent (<10 days) vessel puncture at a noncompressable site||Recent (<10 days) vessel puncture at a noncompressable site|
|Current use of oral anticoagulants (INR >1.3)||Current use of oral anticoagulants (INR >1.7)|
|History of gastrointestinal ulcer within 3 months||History of gastrointestinal or urological hemorrhage within 21 days|
|History of previous stroke combined with DM|
|Known bleeding diathesis within 6 months|
|History of CNS lesion (tumor, aneurysm, or intracranial or spine surgery)|
|History of severe and poorly controlled hypertension|
|AVM or aneurysm|
|Tumor with potential to bleed|
|History of AMI within 3 months|
|Evidence of aortic dissection|
The main outcome measure was the difference in rates of eligibility and treatment after implementing the less restrictive criteria. Secondary outcome measures included the difference in reasons for exclusion from rtPA.
Statistical analysis was performed using SPSS 12.0 (SPSS, Inc., Chicago, IL). A p-value of <0.05 was considered statistically significant. To evaluate differences between the two groups, the two-tailed t-test was used for continuous variables, and the chi-square test was used for categorical variables. Comparisons made between the two groups were also presented as differences in proportions with 95% confidence intervals (CIs). Results of continuous normally distributed variables were expressed as mean ± standard deviation. Fisher’s exact test was used when the numbers tested were too small for chi-square analysis.
Based on patients who presented within 3 hours, a sample of 461 patients provided an 80% power to detect a difference of 7.5% on eligibility for rtPA at the 0.05 significance level, with a baseline proportion of 12%. According to a previous study,17 a > 10% increment of eligibility for rtPA in ischemic stroke patients who arrived at the hospital within 3 hours was considered significant. The poststudy power analysis of the current study estimated that the sample size of this study was sufficient to detect a significant change in eligibility between groups.
The study population consisted of 461 patients, of whom 333 were evaluated by the old criteria and 128 were evaluated by the less restrictive criteria. Details of changes made to the old criteria are in Table 1, and data on which patients became eligible under the new criteria are in Table 2. Important demographic characteristics of the study population before and after the criteria change are in Table 3. Generally, there were no significant differences between the two groups except in sex, as there were significantly more males in 2009. The proportion of patients aged ≥80 years and age of patients who refused consent were similar in the two phases of the study.
|1. Patients older than 80 years|
|2. Patients with NIHSS 4 and 5|
|3. Patients with NIHSS >25|
|4. Patients with seizure at onset of stroke|
|5. Patients with prolonged or traumatic CPR (>2 minutes)|
|6. Patients with delivery|
|7. Patients taking oral anticoagulants with INR 1.3 to 1.7|
|8. Patients with history of gastrointestinal ulcer within 21 to 90 days|
|9. Patients with history of previous stroke combined with DM|
|10. Patients with known bleeding diathesis within 6 months|
|11. Patients with history of CNS lesion (tumor, aneurysm, intracranial or spine surgery)|
|12. Patients with history of severe and poor controlled hypertension|
|13. Patients with AVM or aneurysm|
|14. Patients who had tumor with potential to bleed|
|Time to presentation (minutes)||85.92 ± 86.12||80.70 ± 56.16||0.537|
|Age (yr)||65.67 ± 12.14||67.00 ± 12.92||0.304|
|Sex (male)||197 (59.2)||90 (70.3)||0.027|
|Diabetes||129 (39.8)||41 (33.1)||0.188|
|Hypertension||217 (67.0)||85 (68.6)||0.751|
|Prior stroke||92 (28.4)||38 (30.7)||0.639|
|Atrial fibrillation||61 (18.8)||26 (21.0)||0.608|
|Presenting sBP pressure (mm Hg)||161.33 ± 31.48||157.33 ± 38.38||0.259|
|Prior treatment with antithrombolytics||249 (76.9)||99 (79.8)||0.497|
|Presenting glucose level (mg/dL)||149.69 ± 80.71||155.34 ± 72.74||0.498|
|Patients age ≥80 years||41 (12.5)||21 (16.8)||0.229|
|Age of patients who refused consent (yr)||63.33 ± 13.51||64.42 ± 11.55||0.787|
Details regarding the rates of eligibility and treatment are in Figure 1. When examining all acute ischemic stroke patients, eligibility for rtPA significantly increased, from 40 of 2,176 (1.8%) to 35 of 874 (4%; difference in proportions 2.2%, 95% CI = 0.7% to 3.7%; p < 0.0001), but the rate of treatment was not significantly different between the two phases of the study (17 of 2,176 [0.8%] vs. 8 of 874 [0.9%], difference in proportions 0.1%, 95% CI = −0.7% to 0.9%; p = 0.710). When the subset of patients who arrived within 3 hours of onset was examined, eligibility for rtPA substantially increased from 40 of 333 (12%) to 35 of 128 (27.3%; difference in proportions 15.3%, 95% CI = 6.3% to 24.3%; p < 0.0001). However, the rate of treatment was not significantly different between the two groups (17 of 333 [5.1%] vs. 8 of 128 [6.3%], difference in proportions 1.2%, 95% CI = −4.2% to 6.5%; p = 0.6269).
In patients using the old exclusion criteria, main reasons for not using rtPA included minor or rapidly improving stroke (194 of 333, 58.3%), insufficient time to complete the studies (42 of 333, 12.6%), very old age (3 of 333, 11.1%), and refused consent (23 of 333, 6.9%). In patients using the new exclusion criteria, main reasons for not using rtPA included minor or rapidly improving stroke (70 of 128, 54.7%), refused consent (27 of 128, 21.1%), and insufficient time to complete studies (20 of 128, 15.6%; Table 4). Comparing the two groups, patients excluded by old age (>80 years) significantly decreased from 37 of 333 (11.1%) to 0 of 128 (0%) after using the less restrictive criteria (difference in proportions −11.1%, 95% CI = −15% to −7.2%; p < 0.0001). Patients excluded by previous stroke combined with diabetes also significantly decreased from 19 of 333 (5.7%) to 0 of 128 (0%; difference in proportions −5.7%, 95% CI = −8.7% to −2.7%; p = 0.0029). However, the numbers of patients who refused to receive rtPA significantly increased from 23 of 333 (6.9%) to 27 of 128 (21.1%; difference in proportions 14.2%, 95% CI = 6% to 22.3%; p < 0.0001).
|Reasons for Exclusion From rtPA||Old Criteria (n = 333)||New Criteria (n = 128)||Differences in Proportions (%)||95% CI (%)|
|Age <18 or >80 years||37 (11.1)||0 (0)||−11.1||−15 to −7.2|
|Insufficient time to complete studies||42 (12.6)||20 (15.6)||3||−4.8 to 10.8|
|Minor or rapidly improving stroke||194 (58.3)||70 (54.7)||−3.6||−14.2 to 7.1|
|Severe stroke||14 (4.2)||2 (1.6)||−2.6||−6 to 0.9|
|Seizure at onset of stroke||3 (0.9)||0 (0)||−0.9||−2.5 to 0.7|
|Within 3 months—serious head trauma or previous stroke||6 (1.8)||1 (0.8)||−1||−3.7 to 1.6|
|History of previous stroke combined with DM||19 (5.7)||0 (0)||−5.7||−8.7 to −2.7|
|Patient has received heparin within 48 hours and has an elevated aPTT||1 (0.3)||0 (0)||−0.3||−1.1 to 0.6|
|Platelet count <100 × 109/L||0 (0)||2 (1.6)||1.6||−1.1 to 4.3|
|Intracranial tumor||2 (0.6)||0 (0)||−0.6||−2 to 0.8|
|Systolic blood pressure >185 mm Hg or diastolic blood pressure >110 mm Hg||19 (5.7)||4 (3.1)||−2.6||−7 to 1.9|
|Blood glucose <50 or >400 mg/dL||3 (0.9)||1 (0.8)||−0.1||−2 to 1.8|
|Known bleeding diathesis within 6 months||1 (0.3)||0 (0)||−0.3||−1.1 to 0.6|
|Current use of oral anticoagulants||3 (0.9)||2 (1.6)||0.7||−2.3 to 3.6|
|History of CNS lesion (tumor, aneurysm, intracranial or spine surgery)||2 (0.6)||0 (0)||−0.6||−2 to 0.8|
|History of SAH or ICH||5 (1.5)||2 (1.6)||0.1||−2.5 to 2.6|
|History of severe and poor controlled hypertension||3 (0.9)||0 (0)||−0.9||−2.5 to 0.7|
|Patient has had major surgery or serious trauma within 10 days||1 (0.3)||0 (0)||−0.3||−1.1 to 0.6|
|Hepatic failure, liver cirrhosis, portal hypertension, or acute hepatitis||1 (0.3)||0 (0)||−0.3||−1.1 to 0.6|
|Hemorrhagic retinopathy||1 (0.3)||0 (0)||−0.3||−1.1 to 0.6|
|History of gastrointestinal ulcer within 3 months||6 (1.8)||3 (2.3)||0.5||−3 to 4|
|AVM or aneurysm||1 (0.3)||0 (0)||−0.3||−1.1 to 0.6|
|Tumor with potential to bleed||1 (0.3)||0 (0)||−0.3||−1.1 to 0.6|
|Others (ESRD, CHF, poor general condition)||5 (1.5)||5 (3.9)||2.4||−1.7 to 6.5|
|Refused consent||23 (6.9)||27 (21.1)||14.2||6 to 22.3|
|Patients received IV rtPA||17 (5.1)||8 (6.3)||1.2||−4.2 to 6.5|
This study demonstrates that using less restrictive eligibility criteria for rtPA significantly increases the number of eligible patients, but does not change the rate of treatment with rtPA. Changing the eligibility criteria decreased the number of patients excluded by old age (>80 years) and previous stroke combined with diabetes. Nonetheless, the rates of eligibility and treatment used as the outcome measures in this study are not patient-oriented. Thus, the study is unable to determine if using less restrictive criteria affects important patient-oriented outcomes, including neurologic function or mortality. Future studies must address these issues.
Time from onset longer than 3 hours, or unknown symptom duration, remains the most important cause of patient exclusion from rtPA. In this study, only 15.1% of ischemic stroke patients arrived within the 3-hour time window, which is similar to the results of previous studies (15% to 22%).2,7,8 More than 80% of the patients were excluded by the narrow time window, which precluded rtPA from having a significant effect on the overall ischemic stroke population. Improving the stroke code system and enhancing the use of emergency medical services, as well as widespread public education to increase awareness of warning signs and availability of rtPA, are useful options to increase the transport of patients within the narrow time window. Telemedicine networks will likewise be helpful in extending the use of rtPA.18
However, with all of these efforts, the rates of eligibility and treatment are still very low. In this study, only 0.8% of acute ischemic stroke patients admitted to the hospital received rtPA, which means that 5.4% of patients who arrived within 3 hours, received rtPA. Previous studies showed that 2% of acute ischemic stroke patients admitted to the hospital, and 20% to 27% of patients who arrived within 3 hours received rtPA. Our study showed that the eligibility rate was similar to these data (2.5% vs. 2% and 16.3% vs. 20% to 27%), but the actual rate of treatment was much lower (0.8% vs. 2% and 5.4% vs. 20% to 27%).2,7
Because the proportion of ischemic stroke patients arriving within 3 hours is close to that of previous studies, it is imperative to know the reasons why the rate of treatment of rtPA is so low. From 2004 to 2005,9 the main reasons for exclusion were minor or improving stroke (46.2%), hypertension (35.2%), and insufficient time to complete the studies (24.2%). From 2006 to 2008, the main reasons for exclusion were minor or rapidly improving stroke (58.3%), insufficient time to complete the studies (12.6%), and very old age (11.1%). From these data, improving the stroke code system and widespread public education of availability of rtPA may contribute to decreasing the number of patients excluded.
Aside from failure to meet the narrow time window, minor or rapidly improving stroke is another important cause of patient exclusion from rtPA.8 This may be explained by the fact that, compared to patients with moderate or severe stroke, patients with minor stroke have much higher chances of seeking medical attention, especially in patients without change of consciousness or aphasia. Patients with moderate or severe stroke can probably be only found by others means, thus diminishing their chances of arriving within 3 hours.
It is also possible that the low rate of treatment of rtPA in the study population is related to the exclusion criteria used. The initial exclusion criteria were modified from the NINDS rtPA trial and used in the study hospital since 2004. These exclusion criteria were strict to avoid intracerebral hemorrhage, at the cost of lowering eligibility. Because of this, less restrictive exclusion criteria were released by the Taiwan Stroke Society in November 2008, based on new evidence.10–16 Since 2009, the new criteria have been used to evaluate stroke patients who arrive at the hospital within 3 hours of stroke onset. As expected, eligibility for rtPA has increased, but the rate of treatment has not. Although patients excluded by old age and previous stroke combined with diabetes have significantly decreased, patients who refused to receive rtPA have significantly increased. This is possibly due to the high symptomatic intracerebral hemorrhage rate (10.4%) and mortality (12.8%) after using rtPA among the Chinese-Taiwanese.19
The high symptomatic intracerebral hemorrhage and mortality rates can influence the attitude of EPs and neurologists toward the use of rtPA. Although recent studies show that using rtPA in very old patients (>80 years) or mild stroke (NIHSS 4–6) may still have benefits,11–14 EPs and neurologists tend not to emphasize the benefits of rtPA in these conditions to avoid inadvertently misleading the patients or their families by being encouraged, which may raise legal problems later if devastating intracerebral hemorrhage occurs. It is likely physicians are skeptical of these new recommendations and resist encouraging patients to provide consent to treatment. Although such influence has not been studied and proven here, it is a very possible driving force on why treatment rates do not increase with eligibility rates.20
A limitation of this study is the inadequate statistical power to detect a difference in the rate of treatment between the two groups. It is possible that the 1% difference is actually statistically significant. However, statistical significance is different from clinical significance, and the latter is the one that matters. In the study hospital, an average of 115 ischemic stroke patients arrive within 3 hours of onset annually. An increase of 1% in treatment with rtPA suggests one additional patient receiving rtPA per year, which is not clinically significant despite being statistically significant. Thus, it is likely safe to conclude that the less restrictive criteria do not have any significant effect on the rate of treatment.
Another limitation is that the ideal study should determine if the new criteria have an influence on patient-oriented outcomes, such as neurologic function and mortality. These issues can be used as reference for future directions.
In this study, increasing the rate of eligibility for recombinant tissue plasminogen activator treatment of acute stroke through less strict criteria did not increase the rate of treatment, which is still far from satisfying. This is possibly due to the high symptomatic intracerebral hemorrhage rate and mortality after using rtPA among the Chinese-Taiwanese, which is a major concern for EPs, neurologists, and patients. Lowering symptomatic intracerebral hemorrhage rate may be crucial before the rate of treatment can be increased.