Automated Cerebral Hemorrhage Detection Using RAPID

BACKGROUND AND PURPOSE

Intracranial hemorrhage (ICH) is an important event that is diagnosed on head NCCT. Increased NCCT utilization in busy hospitals may limit timely identification of ICH. RAPID ICH is an automated hybrid 2D-3D convolutional neural network application designed to detect ICH that may allow for expedited ICH diagnosis. We determined the accuracy of RAPID ICH for ICH detection and ICH volumetric quantification on NCCT.

MATERIALS AND METHODS

NCCT scans were evaluated for ICH by RAPID ICH. Consensus detection of ICH by 3 neuroradiology experts was used as the criterion standard for RAPID ICH comparison. ICH volume was also automatically determined by RAPID ICH in patients with intraparenchymal or intraventricular hemorrhage and compared with manually segmented ICH volumes by a single neuroradiology expert. ICH detection accuracy, sensitivity, specificity, positive predictive value, negative predictive value, and positive and negative likelihood ratios by RAPID ICH were determined.

RESULTS

We included 308 studies. RAPID ICH correctly identified 151/158 ICH cases and 143/150 ICH-negative cases, which resulted in high sensitivity (0.956, CI: 0.911-0.978), specificity (0.953, CI: 0.907-0.977), positive predictive value (0.956, CI: 0.911-0.978), and negative predictive value (0.953, CI: 0.907-0.977) for ICH detection. The positive likelihood ratio (20.479, CI 9.928-42.245) and negative likelihood ratio (0.046, CI 0.023-0.096) for ICH detection were similarly favorable. RAPID ICH volumetric quantification for intraparenchymal and intraventricular hemorrhages strongly correlated with expert manual segmentation (correlation coefficient r = 0.983); the median absolute error was 3 mL.

CONCLUSIONS

RAPID ICH is highly accurate in the detection of ICH and in the volumetric quantification of intraparenchymal and intraventricular hemorrhages.

Thrombectomy for Stroke in the Public Health Care System of Brazil

BACKGROUND

Randomized trials involving patients with stroke have established that outcomes are improved with the use of thrombectomy for large-vessel occlusion. These trials were performed in high-resource countries and have had limited effects on medical practice in low- and middle-income countries.

METHODS

We studied the safety and efficacy of thrombectomy in the public health system of Brazil. In 12 public hospitals, patients with a proximal intracranial occlusion in the anterior circulation that could be treated within 8 hours after the onset of stroke symptoms were randomly assigned in a 1:1 ratio to receive standard care plus mechanical thrombectomy (thrombectomy group) or standard care alone (control group). The primary outcome was the score on the modified Rankin scale (range, 0 [no symptoms] to 6 [death]) at 90 days.

RESULTS

A total of 300 patients were enrolled, including 79 who had undergone thrombectomy during an open-label roll-in period. Approximately 70% in the two groups received intravenous alteplase. The trial was stopped early because of efficacy when 221 of a planned 690 patients had undergone randomization (111 to the thrombectomy group and 110 to the control group). The common odds ratio for a better distribution of scores on the modified Rankin scale at 90 days was 2.28 (95% confidence interval [CI], 1.41 to 3.69; P = 0.001), favoring thrombectomy. The percentage of patients with a score on the modified Rankin scale of 0 to 2, signifying an absence of or minor neurologic deficit, was 35.1% in the thrombectomy group and 20.0% in the control group (difference, 15.1 percentage points; 95% CI, 2.6 to 27.6). Asymptomatic intracranial hemorrhage occurred in 51.4% of the patients in the thrombectomy group and 24.5% of those in the control group; symptomatic intracranial hemorrhage occurred in 4.5% of the patients in each group.

CONCLUSIONS

In this randomized trial conducted in the public health care system of Brazil, endovascular treatment within 8 hours after the onset of stroke symptoms in conjunction with standard care resulted in better functional outcomes at 90 days than standard care alone. (Funded by the Brazilian Ministry of Health; RESILIENT ClinicalTrials.gov number, NCT02216643.).

Abstract WMP1: Implementation of a Telestroke Program Based on a Smartphone Application

Background: The increasing demand and shortage of experts to evaluate and treat acute stroke patients has led to the development of remote communication tools to aid stroke management. We aimed to describe the experience of the Brazilian Stroke Network in the implementation of a telemedicine stroke program using a low-cost smartphone application system (JOIN App, Allm, Japan) for rapid sharing of clinical and neuroimaging patient data.

Methods: We evaluated the initial experience of the telestroke program using a smartphone app measuring its feasibility, safety and speed in the acute stroke decision-making process, with a particular focus on intravenous thrombolysis. The App was implemented in hospitals without neurologists available for acute stroke evaluation and was connected with a stroke team to support the decision for thrombolysis. We analyzed the times of acute treatment and safety through the rate of symptomatic intracranial hemorrhage (sICH) in 24hours. The program was implemented as part of the partnership with the Angels Initiative in Brazil supported by Boehringer Ingelheim to improve stroke care across the globe.

Results: The telestroke program started in May 2019 and since then 7 hospitals, in 5 Brazilian states, were included. Only 2 of these hospitals had previous experience with stroke thrombolysis. In all hospitals, the patients were assisted by emergency physicians with eventual support of local neurologists. The telestroke program was activated for 58 patients. The median age was 68yo (IQR58-72), 47% were female and the baseline NIHSS was 11 (IQR8-16). In 69% of the activation the diagnosis was ischemic stroke (IS), 7% hemorrhagic, 5% TIA and 28% other diagnosis. Thrombolysis was suggested in 48% of all evaluations (70% of IS patients). A total of 53% of IS were actually treated with thrombolys. The response-time of the stroke experts was 2.6 minutes (2-6), door-to-CT scan 20min (15-31) and door-needle time 59min (27-81). None of the patients had sICH.

Conclusion: In a telestroke program using a smartphone App, thrombolysis performed by emergency physicians was feasible and safe. This mobile low cost technology can increase the possibility of patients with stroke to receive treatment in regions without neurologists across the globe.

Validation of a Smartphone Application in the Evaluation and Treatment of Acute Stroke in a Comprehensive Stroke Center

Background and Purpose- The increasing demand and shortage of experts to evaluate and treat acute stroke patients has led to the development of remote communication tools to aid stroke management. We aimed to evaluate the JOIN App smartphone system-a low-cost tool for rapid clinical and neuroimaging data sharing to expedite decision-making in stroke. Methods- Consecutive acute ischemic stroke patients treated at a University Hospital in Brazil from December 2014 to December 2015 were evaluated. The analysis included all patients presenting with acute ischemic stroke who underwent initial evaluation by neurology residents followed by JOIN teleconsultation with a stroke neurologist on call for management decisions. An expert panel of stroke neurologists and neuroradiologists revised all cases using a standard Picture Archiving and Communication System imaging workstation within 24 hours and analyzed the decision made with remote assistance during the emergency setting. Results- A total of 720 stroke codes were evaluated with 442 acute ischemic stroke qualifying. Seventy-eight (18%) patients were treated with intravenous thrombolysis. The main reasons for tPA (tissue-type plasminogen activator) exclusion were symptom onset >4.5 hours (n=295; 67%) and hypodense middle cerebral artery territory area >1/3 (n=31; 7%). The agreement rates between Picture Archiving and Communication System versus JOIN-based thrombolysis decisions were 100% for the stroke (unblinded) and 99.3% for the neuroradiologist (blinded) experts. The use of the application resulted in a significant reduction in the door-to-needle times across the pre- versus postimplementation periods (median, 90 [interquartile range, 75-106] versus 63 [interquartile range, 61-117] minutes; P=0.03). The rates of 90-day excellent outcomes (modified Rankin Scale, 0-1) were 51.3%; 90-day mortality, 2.6%; and symptomatic intracranial hemorrhage, 3.8%. Conclusions- The JOIN smartphone system allows rapid sharing of clinical and imaging data to facilitate decisions for stroke treatment. The remote application-based decisions seem to be as accurate as the physical presence of stroke experts and might lead to faster times to treatment. This system represents an easily implementable low-cost telemedicine solution for centers that cannot afford the full-time presence of stroke specialists.

Priorities to reduce the burden of stroke in Latin American countries

The large and increasing burden of stroke in Latin American countries, and the need to meet the UN and WHO requirements for reducing the burden from non-communicable disorders (including stroke), brought together stroke experts and representatives of the Ministries of Health of 13 Latin American countries for the 1st Latin American Stroke Ministerial meeting in Gramado, Brazil, to discuss the problem and identify ways of cooperating to reduce the burden of stroke in the region. Discussions were focused on the regional and country-specific activities associated with stroke prevention and treatment, including public stroke awareness, prevention strategies, delivery and organisation of care, clinical practice gaps, and unmet needs. The meeting culminated with the adoption of the special Gramado Declaration, signed by all Ministerial officials who attended the meeting. With agreed priorities for stroke prevention, treatment, and research, an opportunity now exists to translate this Declaration into an action plan to reduce the burden of stroke.

Abstract TP314: Comparison of Patient-centered Outcome Measures of the International Consortium for Health Outcomes Measurements (ICHOM) Between Private and Public Healthcare System in Brazil

Introduction: An international experts panel was assembled to create a standard set of outcome measures for use in both low and high-income countries. Additionally, the panel intended to represent the most relevant outcomes for subsequent cost-effectiveness analysis. For implementing the tool, it was necessary to make it feasible in different healthcare systems. The objective of this study was to implement and compare the ICHOM outcome measures between two different stroke centers: a university public hospital and a private hospital.

Methods: the medical data of all patients with stroke diagnosis consecutively admitted in the two hospitals were registered. ICHOM outcomes in the 3-month follow-up were measured by interviews in person, by phone or by e-mail, and the results were compared between the hospitals.

Results: 90-day outcome measurements were available for 328 patients (169 in the private hospital and 159 in the public), corresponding to 87% of the patients admitted with stroke diagnosis in these hospitals. The mean age was 68 years in the private vs. 65 in the public, with a mean NIHSS of 7, and 90% had ischemic strokes (18% received IV thrombolysis in the private setting vs. 16% in the public). Public hospital inpatients had a greater number of comorbidities. The mortality rate was 14% in both hospitals, and functional independence (mRS 0-2) occurred in 51% of private setting patients and 39% of the public. The greater proportion of patients needing help for dressing or toileting, as well as having language issues, was in the public setting. Better outcomes were measured in the private hospital regarding the resumption of the social role (64% vs 40%), satisfaction with social activities (65% vs 57%) and good quality of life (66% vs 57%). In both hospitals, 34% had access to physiotherapy after hospital discharge.

Conclusions: In patients with acute stroke from private and public hospitals the thrombolysis eligibility are similar, without increasing the mortality rates in the public setting. However, better functional outcomes were found in private patients, probably due to less comorbidities, better prevention and post-discharge rehabilitation care.

Abstract TP312: Assessment of the Data Collection Strategies for Stroke Patient-centered Outcomes: Implementation of the International Consortium for Health Outcomes in Brazil

Introduction: In 2015, international experts defined a standard set of patient-centered outcome measures after stroke as part of the International Consortium for Health Outcomes (ICHOM). It aims to make outcomes comparable around the world, allowing the value-based healthcare delivery. In developing countries, it is still a challenge to implement outcome measurement for stroke patients after hospital discharge, due to different structures and healthcare network characteristics. The study objective is to assess the implementation strategies of the outcome measures of ICHOM stroke in 6 Brazilian hospitals.

Methods: For 1 year, we progressively implemented ICHOM outcome measurements in 6 stroke centers: 1 private, 1 public university and 4 primary stroke centers with telemedicine guidance. The data collection strategies included: in person interview, interview by phone or a link to an online form sent by e-mail.

Results: 429 patients were included in the 90-day follow-up interview. In person interview was made for 80 patients (mean interview time of 5 minutes), by phone in 285 (mean interview time 8 minutes) and by e-mail in 64. There was loss of follow-up in 15% of in-person interviews, 28% on phone and 62% on e-mail. In the private hospital, where there is better data registry, only 18% of the patients were lost to follow-up, while 54% of telemedicine patients weren’t found for the interview, mainly because of lack of phone number register or even wrong number. Higher succeeded attempts occurred on evening calls compared to business hours’ calls.

Conclusion: The data collection for outcome measurements of stroke patients in Brazil is still a challenge. The barrier for phone interviews was data registry error, and the correct phone number being registered allows the follow-up. Though requiring further validation, e-mail contact seems to be a good initial strategy, decreasing the need for phone calls and saving time. The phone calls are the option for patients who do not answer e-mails.