We recruited adult (≥18 years) patients of either sex, who had an AMIS with NIHSS score ≤ 3 or a TIA with moderate-to-high risk of stroke recurrence (ABCD² Score ≥4) within 14 days of symptom onset.

The inclusion criteria were as follows: (1) ≥18 years of any sex or ethnicity; (2) diagnosis of non-cardiogenic AMIS/TIA within 14 days of stroke symptom onset, wherein AMIS was defined as ischemic stroke with an NIHSS score ≤ 3 at the time of enrollment, and TIA was defined as a transient episode of neurological dysfunction without acute infarction with moderate-to-high risk of stroke recurrence (defined as an ABCD score ≥ 4 at the time of enrollment); (3) stable vital signs, with normal cardiac (Class I-II in New York Heart Association Functional Classification), hepatic (normal ranges in blood liver function tests) and renal functions (normal ranges in blood renal function tests); (4) ability consent by themselves or by a legally authorized representative; and (5) agreement to conduct regular RIC by themselves or others.

On the other hand, subjects who met any of the following exclusion criteria were excluded from the study: (1) diagnosis of brain hemorrhage or other pathologies, such as vascular malformation, tumor, abscess, or other non-vascular diseases, based on brain computed tomography (CT) or magnetic resonance imaging (MRI); (2) modified Rankin Scale (mRS) score >2 before the index event; (3) administration of intravenous thrombolytic therapy (Alteplase or Urokinase) or endovascular treatment for the index event; (4) contraindication to aspirin or clopidogrel (e.g., known allergy, severe asthma, heart failure); (5) indication for anticoagulation therapy (cardiac source of embolus); (6) hemorrhagic tendency of any reason (including but not limited to hemostatic disorder, platelet count <100 × 10⁹/L, history of hepatic dysfunction, among others); (7) any hemorrhagic transformation on brain scans (MRI or CT); (8) gastrointestinal bleed or major surgery within 3 months before the index event; (9) stroke or TIA due to medical procedure or other iatrogenic cause; (10) any upper extremity soft tissue disease, vascular injury, or peripheral blood vessel disease, which is contraindicated for RIC; (11) hypertension with a systolic blood pressure ≥ 200 mmHg despite medical treatment at the time of enrolment; (12) planned revascularization (any angioplasty or vascular surgery) within the following 3 months; (13) scheduled surgery or intervention within the following 3 months that may affect the study procedure; (14) life expectancy ≤ 6 months; (15) pregnant at the time of the study; and (16) ongoing investigational drug or device use by other studies at the time of the study.

Eligible subjects were identified from the inpatient service or stroke emergency center of each study site, where a well-trained research physician confirmed the diagnosis of AMIS (NIHSS score ≤ 3) or TIA (ABCD² score ≥ 4). If the patient met the criteria and provided written informed consent by themselves or legal proxies, they were instructed how to perform RIC using an electric auto-control device (patent number ZL200820123637.X, China), which can be performed by the patient or family members. RIC consisted of five cycles of 5-min inflation at 200 mmHg and 5-min deflation of cuffs on bilateral upper limbs twice a day (45 min) for 90 consecutive days. Antiplatelet strategies based on the physician's best judgment were as follows: aspirin alone (100–300 mg daily), clopidogrel alone (75 mg daily), or a combination of aspirin and clopidogrel. Study visits were conducted on the day of enrollment (day 1) at the inpatient wards or emergency rooms, and the patients were followed up on day 30 ±7 and 90 ±14 in the outpatient department.

In this study, the primary efficacy outcome was the number (percentage) of patients who had a recurrent ischemic stroke or TIA within 90 days after the index event. The secondary efficacy outcome measures included: (1) number (percentage) of patients who had a second ischemic stroke or TIA within 1 month after the index event; (2) number (percentage) of patients with a new cerebrovascular and coronary artery event within 1 and 3 months, including hemorrhagic stroke, myocardial infarction, and deaths from cardiovascular causes, from all causes, and from the index event; (3) NIHSS score change (continuous) from the baseline to 1 and 3 months; (4) mRS score (continuous), dichotomized by percentage with score ≤ 1 vs. ≥2, at 1 and 3 months; (5) Barthel index score (continuous), dichotomized by percentage with score ≥95 vs. <95, at 1 and 3 months; and (6) hand grip strength change (continuous) from baseline to 1 and 3 months on the affected side in patients with upper-limb motor deficit who did not have a recurrent vascular event.

Additionally, compliance was assessed with simultaneous records, which were delivered by the electric auto-control device through 4G signals, and the feasibility outcomes were defined as the number (percentage) of patients who completed ≥50% or <50% of the long-term, regular, 45-min RIC.

Furthermore, safety outcomes were defined as the risks of expected device-related local or systemic adverse events, including the number (percentage) of patients with upper limb pain assessed by the visual analog scale; redness, swelling, or skin petechiae; palpitation; and dizziness. Any new condition (symptoms, injuries, significant abnormal laboratory values) that was not present at the beginning of the study was also documented as an unexpected adverse event, and the determination of whether these events were associated with the RIC device was adjudicated by research physicians. Serious adverse events (SAEs) in this study included death, life-threatening conditions, inpatient hospitalization or prolongation of existing hospitalization requiring medical/surgical intervention to prevent permanent impairment or damage, and other serious medical events.

The null hypothesis was that the recurrence rate of ischemic stroke/TIA within 3 months would be greater than the largest regression probability of recurrence (P₀ = 12%). The alternative hypothesis was that the recurrence rate of ischemic stroke/TIA within 3 months would be less than the smallest regression probability of recurrence (P_A= 6%) (14). To test our hypothesis, with an assumption of a type 1 error of 5%, a power of 80%, and an attrition rate of 10%, the sample size was estimated to be 165 (15).

Baseline categorical variables were listed as numbers (percentage), baseline normally distributed continuous variables were reported as means (standard deviation, SD), and non-normally distributed continuous variables were reported as medians (interquartile range, IQR).

We used multivariate logistic regression models to assess the association between the odds ratio (OR) of recurrent ischemic stroke/TIA occurring within 3 months with a compliance rate of RIC (≥50 vs. <50%), adjusting the antiplatelet strategy (dual vs. single antiplatelet), age (≥60 vs. <60 years), and sex (male vs. female) parameters. Moreover, we used repeated-measures analysis with mixed models to analyze NIHSS or handgrip strength changes associated with the RIC compliance rate (≥50 vs. <50%), antiplatelet strategy (dual vs. single antiplatelet), and follow-up visits (baseline, 1 month, and 3 months). To follow the parsimonious principle in variable selection for regression models, we only included the most crucial covariates that may affect recurrence or neurological behavior results. A P-value of < 0.05 was considered as statistically significant, and all statistical analyses were performed using SAS software, version 9.4 (SAS Institute, Carry, NC, USA).

From December 2016 to August 2017, 285 patients met the study criteria at the four enrollment centers, and 167 of them provided signed informed consent and were enrolled in the study. Ultimately, a total of 162 patients were included in the final analysis, whereas 118 patients declined to participate for various reasons, including skepticism of the study device or clinical trials, inability to commit to follow-up visits, and other non-specific reasons. Notably, five patients were excluded for the following reasons: one patient lost to follow-up due to a car accident; one patient received Alteplase meeting the exclusion criteria; one patient did not meet the inclusion criteria (had stroke >14 days ago); and two patients received cerebral artery stent within 3 months meeting the exclusion criteria (Figure 1).

The median age of the included patients was 58 years, and 74.7% (121/162) of them were male. 106 (65.4%) patients received dual antiplatelet therapy with the combination of aspirin and clopidogrel, 45 (27.8%) patients received aspirin and 11 (6.8%) received clopidogrel. The index event was AMIS in 153 patients (94.4%). The median NIHSS score was 1. ABCD² score was 5 in the patients with TIA. Regarding medical history, 69.8% (113/162) of the included patients had hypertension, 36.4% (59/162) had diabetes, and 51.2% (83/162) were current or previous smokers. Furthermore, the median time from the onset of the qualifying event to RIC intervention was 7 days. Baseline blood biochemical test results of the included patients are listed in Table 1.

Ischemic stroke/TIA occurred in six patients (3.7%) within 3 months. It rejects the null hypothesis and supports the alternative hypothesis. (i.e. the recurrence rate of ischemic stroke/TIA within 3 months is less than the smallest regression probability of recurrence of 6%). Among the six patients, three patients received dual antiplatelet therapy (Table 2), two patients received aspirin, and one patient received clopidogrel alone. Regarding compliance to RIC, Three patients are being compliant while the other 3 patients are not compliant.

Ischemic stroke/TIA occurred in one patient (0.9%) under dual antiplatelet therapy, and one patient (1.8%) under single antiplatelet therapy (aspirin, to be specific) within 1 month. None of the included patients had any other vascular events or hemorrhagic stroke. We collected neurological outcomes data from 143 patients who visited the outpatient department in person, showing that 74.8% (107/143) of patients had a favorable outcome (mRS score of 0 or 1) at 3 months, with a median mRS score of 1 (0–2) and Barthel index of 100 (100–100). The mean change in NIHSS score in the study patients was −0.8 and −1.0 within 1 and 3 months, respectively. The mean change in grip strength at the affected upper limbs was 4.3 and 4.6 Kg within 1 and 3 months, respectively (Table 2).

Upper limb pain was reported to be the most common (44/162, 27.2%) local adverse event, followed by upper limb petechia (26/162, 16.0%), heart palpitation (5/162, 3.1%), and superficial venous thrombosis within the upper limb (1/162, 0.6%). No bleeding events occurred. No SAEs were observed.

Only 68 patients (42.0%) completed ≥ 50% of the long-term, regular, 45-min RIC sessions using real-time 4G signal data (Table 2). They were 43 men (63%) and the median age of the 68 patients was 61 years. Three (4.4%) patients had the primary outcomes within three months, and none had recurrence within 1 month (Table 2).

No significant difference of recurrent ischemic stroke/TIA within 3 months were detected on comparison of the subgroups completing ≥50% of the 45-min RIC tasks and those who did not (OR: 1.7, 95% CI: 0.3–8.7, P = 0.55), after adjusting for age (≥60 vs. <60 years), sex (male vs. female), and antiplatelet strategy (dual vs. single antiplatelet) (Figure 2). In the multiple-variable model, age (≥60 vs. <60 years), sex (male vs. female) or antiplatelet strategy (dual vs. single antiplatelet) was not associated with the risks of having a recurrence.

In the repeated-measures models consisting variables of antiplatelet strategy (dual vs. single antiplatelet), the compliance rate of 45-min RIC (≥50 vs. <50%), and follow-up visits (baseline, 1 month and 3 months), NIHSS at 3 months was significantly improved, as compared to baseline (−1.0 ± 0.1, 95% CI: −1.2 to −0.9; P < 0. 001), although antiplatelet strategy and compliance rate of 45-min RIC were not associated with NIHSS change in this model (P = 0.60 and 0.76, respectively). Similarly, handgrip strength was estimated to increase by 4.6 ± 0.8 kg from baseline to 3 months in patients with unilateral upper limb weakness resulting from the index event without any vascular events (95% CI 3.0–6.0; P < 0. 001) (Table 2); however, antiplatelet strategy and compliance rate of 45-min RIC were not associated with the handgrip strength change in this model (P = 0.24 and 0.98, respectively).