Introduction

Intravenous recombinant tissue plasminogen activator (rtPA) is the only approved medical reperfusion therapy for patients with acute ischemic stroke (AIS)1. Guidelines1,2 recommend this treatment to be initiated as soon as possible within 4.5 hours of stroke onset, as shorter onset-to-treatment time (OTT) translates into a better functional outcome3,4. OTT is composed of onset-to-door time (ODT) and door-to-needle time (DNT). However, it is difficult to accomplish improvement of ODT because campaigns aiming at raising public awareness of stroke symptoms have only limited impact on ODT5. Conversely, a focus on decreasing in-hospital DNT is feasible, valuable and arguably essential, for quality systems of stroke care1,6.

We used the dataset from the Enhanced Control of Hypertension and Acute Stroke Study (ENCHANTED)7,8 rtPA-dose comparison arm to determine (i) whether DNT improved over time among participants, (ii) the extent to which DNT in China differed from other regions, and (iii) the clinical predictors of DNT.

Results

There were 3219 (97.3%) patients with available data were included in these analyses. Table 1 shows the baseline characteristics of patients by DNT (≤60 vs. >60 mins). The majority of participants (1903 [59.1%]) had DNT >60 m. These were significantly more likely to be younger, Chinese, hypertensive, off-hour admission, being treated in a larger center, and have longer ODT. Multivariable analysis shows the independent predictors of longer DNT included recruitment from China (OR 9.75, 95%CI 7.70–12.35; P < 0.0001), shorter ODT (0.58, 0.53–0.65; P < 0.0001), fewer patients treated in the center (0.96, 0.95–0.97; P < 0.0001), higher diastolic blood pressure (BP) (1.07, 1.00–1.14; P = 0.037), off-hour admission (1.27, 1.07–1.50; P = 0.005), and CT or MRI angiogram not showing proximal occlusion (0.65, 0.52–0.82; P = 0.0002). Chinese patients were twice as likely (1147, 83.0%) to have DNT >60 m compared to non-Chinese patients (756, 41.2%). Shorter ODT and fewer patients treated in the center were independently associated with DNT >60 m for both Chinese and non-Chinese patients. Among non-Chinese patients, male patients (female vs. male: 0.74 [0.61–0.90]; P = 0.003) and Asians (other than Chinese) (0.62, 0.51–0.76; P < 0.0001) were less likely to have DNT >60 m (see Supplementary Tables S1 and S2).

Table 1 Baseline characteristics of patients with acute ischemic stroke, according to door-to-needle time above and below 60 minutes for use of intravenous thrombolysis treatment.
Full size table

Table 2 and Fig. 1 show that the overall mean DNT (min) was 71.8 (95%CI 70.4–73.2), and that it decreased significantly over the course of the trial (fourths): 77.9 (74.9–80.9), 69.3 (66.7–72.0), 69.1 (66.5–71.8) and 71.4 (68.7–74.2) (P for trend, 0.003) and this trend remained consistent after adjusting for confounders. In particular, there was a significant reduction of 33 minutes between the first and final fourth epocs in time across the trial in China (P for trend, 0.001), whilst there was little temporal difference in the other participating countries.

Table 2 Door-to-Needle time over time.
Full size table
Figure 1
figure 1

Correlation between time from the activation of each center in the ENCHANTED trial and door-to-needle time for (A) all patients, (B) Chinese patients, and (C) non-Chinese patients.

Full size image

Discussion

These analyses of a large clinical trial dataset involving AIS patients treated in a wide range of hospitals in multiple countries shows that DNT, whilst longer than recommended in best practice guidelines, progressively decreased in centers over the course of the study. In particular, centers in China achieved a 33 minutes reduction in DNT over three years of participation in ENCHANTED, whilst DNT remained consistent elsewhere. Aside from recruitment in China, the independent predictors of long DNT included shorter ODT, fewer patients treated at a center, higher diastolic BP, off-hour admission, and the absence of proximal occlusion on imaging.

Our analysis demonstrated that there has been a significant reduction in DNT over time for patients in China, which was not matched by reductions in other regions. However, there is still a large gap between the DNT in China (mean DNT 97.5 min, which is consistent with some previous observational studies in China) and the quality performance DNT of <60 min recommended in international guidelines8,9,10,11. Although various quality improvement efforts have been undertaken in China in the last decade, resulting in dramatic improvements in acute stroke care, there remains significant gaps between guideline recommendations and clinical practice in this country9. The Target Stroke Initiative, launched by the American Heart Association and American Stroke Association in 2011, was successful at improving the timeliness of rtPA administration for AIS on a national scale in the United States10,12,13. A similar comprehensive program is clearly needed in China, with ongoing monitoring of DNT as an important benchmark for quality in stroke care1,10,11,12. Why DNT did not improve for non-China centers (mean 57.1 min) may relate to the already strong performance and low volumes compared with Chinese centers.

Our analyses identified some interesting factors influencing DNT. Firstly, our data confirm the paradox that shorter ODT is associated with longer DNT13,14,15. This suggests that the responsible medical officer may be influenced by the perception that there is ample time prior to the expiration of the thrombolysis window - this may be especially true in China12,16. Secondly, in line with previous reports14,17, we found that hospitals with higher annual volumes of thrombolysis activity achieved significantly shorter DNT. This may relate to high-volume hospitals having better processes in place for investigation and decision-making for rtPA administration17. Thirdly, our analysis of higher diastolic BP being associated with longer DNT supports the findings of Skolarus et al. who reported that pre-thrombolytic antihypertensive treatment for patients with elevated BP prolongs DNT in a secondary analysis of a clinical trial population18. Furthermore, our analysis demonstrates that off-hour admission was an independent predictor of longer DNT, which may be due to reduced hospital staffing and delayed imaging procedures during off-hour time19,20.

Key strengths of our study include the large sample size of patients, and the rigorous evaluation7,8. However, as ENCHANTED excluded those patients who were likely to die within the next 24 hours, the present findings may not be generalizable to patients with severe AIS. In addition, several factors which influence DNT, including hospital pre-notification by ambulances10 and time needed for informed consent and randomization, were not collected and thus excluded from the adjusted analyses.

In summary, DNT progressively declined among centers participating in the ENCHANTED trial, especially in China. However, as DNT remains longer than the 60 minute cut-point recommended in guidelines, national programs to improve the timeliness of delivery of rtPA are needed in China.

Methods

Design

The ENCHANTED trial is an international, multi-center, prospective, randomized, open-label, blinded-endpoint trial; the details of which are outlined elsewhere7,8,21. In brief, 3310 patients with a clinical diagnosis of AIS confirmed on brain imaging and fulfilling local criteria for thrombolysis treatment, including symptom onset within 4.5 hours, participated in the rtPA dose evaluation arm, where they were randomly assigned to receive low- (0.6 mg/kg; 15% as bolus, 85% as infusion over 1 hour) or standard-dose (0.9 mg/kg; 10% as   bolus, 90% as infusion over 1 hour) rtPA. The study protocol was approved by Ethics Committee of Peking University First Hospital and appropriate ethics committee at each participating center. Written informed consent was obtained from each patient or their appropriate surrogate. Participants of this study were managed according to standard guidelines and protocols at each hospital.

Procedures

Key demographic and clinical characteristics were recorded at the time of enrolment, with stroke severity measured using the National Institute of Health Stroke Scale (NIHSS) and Glasgow Coma Scale (GCS) at baseline, 24 hours, and at Day 7 (or earlier on discharge from hospital). Uncompressed digital images of all baseline and follow-up digital CT, MRI and angiogram images were collected in DICOM format on a CD-ROM identified only with the patient’s unique study number, and uploaded by a special purpose-built web-based system for central analysis at The George Institute of Global Health.

Statistical Analysis

Independent predictors of long DNT (>60 min) were determined using logistic regression models. Significant predictors (P < 0.05) from the univariate analysis were tested for their association with long DNT (>60 min) in a multivariable model. We reduced the full model by successively removing nonsignificant covariates until all the remaining predictors remained statistically significant (P < 0.05). Correlation between time from the activation of each center and DNT was assessed using Spearman Correlation. DNT was log-transformed to remove skewness. Generalized linear model was used to investigate secular trends in DNT across fourths of time since the start of each site with adjustment for all independent predictor of long DNT.

Data availability

The dataset analysed in the current study are available upon appropriate request sent to a corresponding author.