BACKGROUND:Tenecteplase has been previously evaluated in large- and medium-sized vessel occlusion subgroups, but its effectiveness in more distal occlusions, particularly those involving the distal middle cerebral artery branches or the anterior cerebral artery and posterior cerebral artery territories, and across varying ischemic core and perfusion profiles, remains uncertain. METHODS:We performed a secondary analysis of TASTE (Tenecteplase Versus Alteplase for Stroke Thrombolysis Evaluation), a randomized clinical trial comparing tenecteplase and alteplase in patients presenting within 4.5 hours of symptom onset with perfusion imaging-confirmed stroke and evidence of target mismatch (penumbra/core ratio >1.8 and an absolute difference >15 mL). The primary outcome was the modified Rankin Scale (mRS) score of 0 to 1 at 90 days. We compared the effect of tenecteplase versus alteplase in subgroups based on occlusion site (proximal M2, distal M2, M3 and beyond, anterior cerebral artery, and posterior cerebral artery), ischemic core, core growth rate, and penumbra. The treatment effect of tenecteplase and alteplase was compared stratifying by the subgroup of interest, adjusting for age, baseline National Institutes of Health Stroke Scale score, and premorbid mRS score in modified Poisson regression models. RESULTS:Of the 680 patients enrolled, 492 were included in the primary analysis (median age, 73 [interquartile range, 63-82] years; male sex: 306/492 [62%]). Two hundred forty-two (49%) received tenecteplase, and 250 (51%) received alteplase. Tenecteplase was associated with a higher proportion of mRS score of 0 to 1 with distal (M3 and beyond) occlusions (tenecteplase: 62/81 [77%] versus alteplase: 59/93 [63%]; adjusted risk ratio, 1.23 [95% CI, 1.04-1.46]). Numerically higher rates of mRS score of 0 to 1 were observed with distal M2 (tenecteplase: 30/46 [65%] versus alteplase 28/48 [58%]; adjusted risk ratio, 1.14 [95% CI, 0.84-1.54]) and anterior cerebral artery occlusions (tenecteplase: 12/21 [57%] versus alteplase 5/13 [38%]; adjusted risk ratio, 1.46 [95% CI, 0.71-3.00]). No treatment differences were seen for proximal M2 or posterior cerebral artery occlusions (Pinteraction across all occlusion sites: 0.89). Across ischemic core, penumbra, and core growth rate subgroups, no difference in treatment effect was observed. CONCLUSIONS:Patients with distal middle cerebral artery occlusions who are treated with tenecteplase are more likely to achieve an mRS score of 0 to 1 at 90 days than those treated with alteplase. REGISTRATION:URL: https://www.anzctr.org.au; Unique identifier: ACTRN12613000243718.

Introduction: Meta-analyses of RCTs of intravenous alteplase (IVT) for stroke with unknown time of onset (SUTO), selected by advanced imaging have shown better functional outcomes than placebo or standard care, but with higher mortality and symptomatic intracerebral hemorrhage (SICH). A balanced assessment of benefits and harms is therefore needed. Hypothesis: Desirability of Outcome Ranking (DOOR) methods may provide a comprehensive assessment of benefits and harms that closely reflect the patient's experience in IVT for SUTO. Methods: The Evaluation of Unknown Onset Stroke Thrombolysis (EOS) collaboration included four RCTs comparing IVT with standard care or placebo. DOOR analysis was applied to individual patient-level data. Patients were assigned a DOOR rank (1-10) based on 90-day mRS, SICH, major extracranial bleeding, and lack of NIHSS improvement ≥2 points at 24 hr. Ranks were: (1) mRS 0–1 with no undesirable events; (2) mRS 0–1 with 1 event (SICH, major extracranial bleeding, or lack of NIHSS improvement); (3) mRS 0–1 with >1 events; (4) mRS 2–3 with no events; (5) mRS 2–3 with 1 event; (6) mRS 2–3 with >1 events; (7) mRS 4–5 with no events; (8) mRS 4–5 with 1 event; (9) mRS 4–5 with >1 events; (10) death. DOOR probability and win ratio were calculated. Results: Among 829 patients (420 IVT, 409 controls), DOOR probability was 0.55 (95% CI, 0.51–0.59; p = 0.011) and win ratio was 1.24 (95% CI, 1.22-1.25), indicating a 55% probability and a 1.24 times higher frequency of achieving a more desirable outcome with IVT compared with control (Fig 1, 2). A forest plot shows that IVT was associated with a lower proportion of patients with poor functional outcome (mRS 4–5: 15.0% vs. 21.5%) and a higher proportion of NIHSS improvement (57.6% vs. 48.9%), but a higher rate of death (6.4% vs. 3.4%) and SICH/major extracranial bleeding (3.3% vs. 0.5%) compared to controls (Fig 2). Additionally, a sequential dichotomization analysis of DOOR rankings revealed that the DOOR probability remained above 50% across most dichotomization thresholds. The greatest separation was observed at DOOR > 4 vs. ≤4, with a probability of 54.8% (95% CI, 51.4–58.1) (Fig 3). Conclusions: DOOR analysis demonstrated a more desirable overall outcome with IVT for SUTO. The results suggest that IVT was associated with more favorable DOOR ranking, particularly within the better DOOR ranking strata that reflect a favorable balance between clinical efficacy and safety.

BACKGROUND:Bundled-care protocols, which include rapid lowering of blood pressure, have improved functional outcomes in patients with intracranial haemorrhage (ICH). Specifically, the intensive blood pressure reduction in acute cerebral hemmorhage trail (INTERACT) 3 trial aimed to achieve rapid lowering of systolic blood pressure (SBP) to 140 mmHg within an hour of arrival. We analyse factors that influenced the timely delivery of hyperacute ICH management. AIM:We aimed to identify the factors that delay the timely delivery of antihypertensive therapies for hyperacute ICH. METHODS:A retrospective analysis of acute ICH patients admitted to Monash Health (January-December 2023) was performed to evaluate time to achieve SBP <140 mmHg. A Gamma regression model was used to assess the relationship of predictors, which included patient factors (age, GCS, admission SBP), process factors (times to antihypertensive medication, emergency department (ED) bed, computed tomography (CT) scan), and system factors (Code Stroke status, ED resuscitation bed assignment, care goals). Median times for antihypertensive medication delivery, neurosurgical review, neurosurgical intervention and anticoagulation reversal were calculated. RESULTS:Of 209 patients (median age 74 (IQR 63-84), GCS 14 (IQR 6-15), National Institute of Health Stroke Severity 13 (interquartile range (IQR) 4-27), admission SBP 160 mmHg (IQR 137.5-187)), 52% were treated as a Code Stroke. Median time to CT was 33 min (IQR 23-47.5) with Code Stroke versus 121 min (IQR 36-254) without. Patients who received non-palliative therapies (74%) achieved SBP < 140 in 193 min (IQR 91-552) and 25% achieved this in 60 min. Higher admission SBP, longer times to ED bed and absence of Code Stroke were associated with increased time to reach 140 mmHg SBP (P < 0.01). Other dependent variables did not reach significance. CONCLUSION:SBP on arrival, delayed access to ED beds and Code Stroke activation were key factors associated with achieving early BP lowering. These findings underscore the need for protocolised, rapid-response systems to optimise ICH management.

BACKGROUND:Evidence is lacking on whether support and education programs, facilitated by mobile technology, benefit people with stroke. AIMS:The primary aim was to determine the effectiveness of a co-designed, eHealth self-management intervention for reducing unplanned hospital presentations. METHODS:Prospective, multicenter randomized controlled trial with blinded outcome assessment and intention-to-treat analysis. Randomization (1:1, stratified by age and level of disability at baseline) within 2 weeks of hospital discharge. Participants aged ⩾18 years with modified Rankin Scale (mRS) score 0-4 were recruited from 11 Australian hospitals. To maintain blinding, all participants co-developed three to five self-management goals with a trained clinician-researcher, using a standardized approach. Following randomization, participants were allocated to receive personalized, goal-centered electronic messages (intervention) or administrative messages (active control) over 12 weeks via text or email. The groups were not informed about the number of messages they would receive and were unaware that only the intervention group had their messages tailored to their goals. The primary outcome was unplanned hospital presentations (emergency department/admission) within 90 days post-randomization. Secondary outcomes included goal attainment, self-efficacy, and various health outcomes. We used generalized mixed-effects regression model with a logistic link for categorical outcomes and Cohen's d (0.2 considered small effect, 0.5 moderate-1.0 large) to assess within-group change. RESULTS:The trial was impacted by COVID-19 and ceased with 556/890 planned participants; 465 were randomized (n = 234 control, n = 231 experimental), a median of 10 days post-discharge (median age 67.2 years, 67.1% male, median 3 goals/participant). Primary outcome assessments for 222 control and 218 experimental participants showed no between-group differences for unplanned hospital presentations (odds ratio (OR): 1.32; 95% confidence interval (CI): 0.52-3.34). Compared to controls at 90 days, a larger proportion of the experimental group had >3 unmet needs and slight to severe disability (both <8% difference). Both groups showed non-significant improvements in goal attainment at 90 days post-randomization; overall attainment was stronger among intervention participants (within-group change Cohen's d intervention 0.76 vs 0.61). CONCLUSION:We found no evidence that tailored, electronic messaging over 12 weeks, in addition to structured goal-setting reduced unplanned hospital presentations. Future research is needed to understand the role of disability and unmet needs, and factors that influence goal attainment. TRIAL REGISTRATION:ACTRN12618001468213, U1111-1206-7237.

Introduction To define incident heart failure (HF) risk in ischaemic stroke/transient ischaemic attack (TIA) survivors. The secondary aims were to define the association of HF with all-cause mortality in stroke survivors and to describe their cardiac magnetic resonance (CMR) findings. Methods This was a prospective cohort study using the UK Biobank (UKBB) cohort of individuals aged 40 to 69 years old. We excluded individuals with prior HF and stratified them by history of ischaemic stroke/TIA. The main outcome was incident HF as defined by hospital admissions coded for heart failure. Secondary outcomes were all-cause mortality, myocardial infarction, and CMR findings. Results We included 405 406 individuals (age 56.5 years, 45.6% males). Over 13.7 years, 15 565 individuals experienced incident HF. Stroke survivors had an overall HR of 3.6 [95% confidence interval (CI) 3.3-3.8, P < .0001] for HF hospitalization and an adjusted HR of 1.4 (95% CI 1.3-1.5, P < .0001). The risk of HF hospitalization was greater than the risk of myocardial infarction (12.6% vs 5.4%). Stroke survivors with HF had a lower LVEF and higher LV mass than those without HF. Incident HF in stroke survivors was associated with a HR of 1.8 (95% CI 1.6-1.9, P < .0001) for mortality. Conclusion Incident HF is common in stroke survivors and strongly associates with mortality. The risk of HF varies greatly depending on underlying risk factors. Exploratory analyses suggest that stroke survivors with HF may have a lower ejection fraction phenotype. Future trials of HF preventive therapy in high-risk stroke survivors are warranted.

Abstract Background and aims Multiple randomized control trials have shown that thrombectomy for ischemic stroke resulting from large vessel occlusion (LVO) reduces disability and is cost effective. There remains uncertainty whether thrombectomy for patients with pre-stroke disability (modified Rankin Score/mRS3-4) is similarly effective, as prior meta-analysis (n=14) included cases of mRS2 among those classified as pre-stroke disability. This study aimed to meta-analyse the outcomes of thrombectomy for LVO in those with moderate disability (mRS3-4). Methods We searched Pubmed, Medline and Clinicaltrials.gov using search terms "stroke", "large vessel occlusion", "disability", "prestroke disability", or "pre-stroke disability" on 4/1/2026. From 384 studies, we included those with >90 patients or if there were extractable data on cases with mRS3-4. Analyses were conducted in R using the meta and metafor packages. Results Of 27 studies, two duplicates were excluded. In the remaining 25 studies, 5084/60199 patients with a pre-stroke disability (10.6%, 95% CI 7.8-13.7, I2=99.3%) received thrombectomy. Mean age was 81.0±2.8years. The majority were female 64.9% (95% CI 61.8-67.9%, I2=97.2%). Return to baseline occurred in 27.3% (95% CI 23.2-31.5%, I2=98.9%) and symptomatic intracerebral hemorrhage occurred in 6.5% (95% CI 5.5-7.7%, I2=94.8%). Mortality was 39.5% (95% CI 34.0-45.1%, I2=99.4%). Conclusions The proportion of patients with pre-stroke disability treated by thrombectomy in LVO registries remains low while a modest proportion of the patients returned to baseline following thrombectomy. Surprisingly a large proportion of patients in registries were female, reversing trends seen previously. The high heterogeneity (I2) suggests a large variation in results, and further work in the form of a trial is needed. Conflict of interest Jason Vuong: nothing to disclose, Shaloo Singhal: nothing to disclose, Michael Valente: nothing to disclose, Angelos Sharobeam: nothing to disclose, Lubna Shakhatreh: nothing to disclose, Andy Lim: nothing to disclose, Peter Mitchell: nothing to disclose, Bernard Yan: nothing to disclose, Henry Ma: nothing to disclose, Thanh Phan: nothing to disclose.

Background Dementia prevalence is increasing in Australia. It is unknown whether there are hotspots for dementia in metropolitan or non-metropolitan areas. This knowledge is important for healthcare planning.Objective This paper will examine where hotspots for self-reported dementia in Australia are, and whether they are adequately serviced by multidisciplinary memory clinics. Methods We used self-reported dementia data from the 2021 Australian Census at the local government area (LGA) level. LGAs represent public administrative regions within Australian states and territories. Standardized prevalence ratios (SPR) were calculated for each LGA by dividing the number of self-reported cases by the expected number of cases. Spatial relationships were investigated with Bayesian spatial regression using integrated nested Laplace approximations. Memory clinics were located using Australian Dementia Network and government websites. Results Self-reported dementia prevalence was lower in metropolitan areas (72.3 per 10,000) compared to non-metropolitan areas (79.9 per 10,000). There are 108 multidisciplinary memory clinics in Australia, 83 of which are metropolitan. Hotspots for self-reported dementia occurred in non-metropolitan east coast of New South Wales (SPR: 2.13), southeast Queensland (SPR: 1.72), and northwest of greater city Adelaide (SPR: 2.55). LGAs in major cities had lower SPRs (Melbourne: 0.40; Sydney: 0.45), apart from Western Adelaide (SPR: 1.88). Conclusions Hotspots for self-reported dementia were mainly in non-metropolitan Australia, whereas memory clinic services were mostly in metropolitan areas, raising issues of equity and access to services.

Effective multidisciplinary collaboration underpins integrated care for people living with Huntington’s disease (HD). Traditional evaluations focus on patient outcomes or self-reported collaboration, which may not capture the underlying structure of team interactions. Social network analysis (SNA) provides an objective method to map collaboration structures and quantify changes associated with service redesign. This study aimed to evaluate structural changes in multidisciplinary team collaboration following service redesign in a specialist HD clinic using encounter-derived SNA. We conducted a retrospective before–after cohort study comparing pre-redesign (2014–2015, Before) and post-redesign (2018–2019, After) periods, including all clinic attendees. Provider participation was extracted from electronic health records and aggregated at the patient-day level. Co-occurrence networks were constructed using Pearson correlation coefficients of provider co-occurrence patterns across shared patient-days, generating weighted undirected networks of disciplinary collaboration. Network metrics, including community structure, modularity (Q, a measure of network division into communities), and local (degree, betweenness) and global (PageRank) centrality measures were used to characterise network structure. Secondary analyses examined patient-level outcomes, including survival, functional decline, non-elective hospital admissions, non-elective residential care placement, emergency services and specialist palliative care involvement, and advance care plan completion. A total of 685 patients contributed 9797 encounters (3926 Before; 5871 After). Co-occurrence patterns shifted from more concentrated dyadic co-participation toward more interconnected allied health involvement across patient-days. Network communities decreased from nine (Q = 0.603) to seven (Q = 0.262), reflecting reduced fragmentation and increased cross-disciplinary integration. Allied health providers showed the largest increases in degree centrality, while PageRank changes were heterogeneous across disciplines. The neurologist remained structurally central across both periods, while the clinical liaison showed reduced relative centrality. Overall, the network demonstrated increased multidisciplinary integration with sustained medical centrality. Short-term patient outcomes were largely unchanged. Service redesign in a specialist HD clinic was associated with reduced network fragmentation and reorganisation of multidisciplinary collaboration, particularly among allied health providers, alongside minimal short-term changes in patient outcomes. Encounter-derived SNA provides an objective method to assess structural changes in team collaboration, complementing traditional outcome measures and supporting evaluation of multidisciplinary care processes in complex chronic healthcare settings.

Abstract Background and aims While national epidemiological data provide estimates of overall stroke incidence, stroke risk may vary substantially across regions, and the drivers of geographic variation in Taiwan remain unclear. Methods Nationwide stroke incidence data (ischemic stroke, intracerebral hemorrhage, and subarachnoid hemorrhage) from 349 administrative districts in Taiwan were analyzed using the National Health Insurance Research Database for 2021. To account for unstable coefficients from administrative districts with small populations, we merge adjacent small districts such that all districts have a population >10,000. Standardized stroke ratios (SSRs) were calculated by dividing the observed and expected cases for the region. SSR >1.5 was defined as hotspot. Bayesian spatial regression models using the integrated nested Laplace approximation (INLA) were adopted, with covariates including the proportion of the population aged ≥ 65 years, sex, hypertension, and diabetes. Results Significant spatial clustering was observed (Moran’s I = 0.38, p < 0.001), and spatial models substantially outperformed non-spatial models. Among them, the intrinsic conditional autoregressive (ICAR) Model demonstrated the best fit and an interactive visualization of the results is available at: https://albert-l-phan.github.io/Taiwan_Stroke/. Among covariates, hypertension and diabetes mellitus were strongly associated with higher stroke burden (coefficients 4.72 and 4.05, respectively with P values < 0.05). After accounting for spatial dependence and chronic disease prevalence, the proportion of the population aged ≥ 65 years and sex were no longer statistically significant. Conclusions The stroke incidence in Taiwan shows marked geographic clustering, providing a quantitative basis for allocating public health expenditure and healthcare resources to support more equitable stroke care planning. Conflict of interest All authors: Nothing to disclose

Introduction: Landmark clinical trials have examined the safety and efficacy of tenecteplase, demonstrating non-inferiority when compared to intravenous alteplase in the < 4.5hr time window. Sub-analysis of tenecteplase performance in large and medium sized occlusions have been performed. However, the efficacy of tenecteplase in patients with more distal occlusions, such as those in the distal branches of the middle cerebral artery (MCA), in the anterior (ACA), or posterior cerebral artery (PCA) territories, is unknown. Methods: We performed a pre-planned imaging analysis of TASTE, a multicenter randomized trial of tenecteplase and alteplase in patients presenting within 4.5hrs of symptom onset with perfusion imaging-confirmed stroke and evidence of target mismatch (penumbra/core ratio >1.8 and an absolute difference >15 mL) on CT perfusion. Occlusion site was categorized into: ICA + proximal M1, distal M1, proximal M2, distal M2, M3 and beyond, ACA, and PCA. Patients enrolled into TASTE did not receive endovascular therapy. Primary outcome was mRS 0-1 at 90 days. The treatment effect of tenecteplase and alteplase was compared, stratifying by occlusion site, and adjusting for age, baseline NIHSS and pre-morbid mRS in modified Poisson regression models. Results: Of the 601 patients included in the primary analysis, 295 (49%) received tenecteplase and 306 (51%) received alteplase. Tenecteplase was associated with a higher odds of mRS 0-1 with distal (M3 and beyond) occlusions (tenecteplase: 62/81 [76%] vs. alteplase: 59/93 [63%], aRR: 1.23 [95% CI: 1.04-1.46]). Rates of 90-day mRS 0-1 were comparable for proximal M2 (tenecteplase: 27/48 [56%] vs. alteplase: 29/50 [58%], aRR: 1.02 [95% CI: 0.74-1.41]) and distal M2 occlusions (tenecteplase: 30/46 [65%] vs. alteplase: 28/48 [58%], aRR: 1.14 [95% CI: 0.84-1.54]). No difference in tenecteplase treatment effect was observed for anterior (tenecteplase: 12//21 [57%] vs. alteplase: 5/13 [38%], aRR: 1.45 [95% CI: 0.70-2.99]) or posterior cerebral artery (tenecteplase: 21/46 [46%] vs. alteplase: 22/46 [47%], aRR: 0.92 [95% CI: 0.62-1.36]) occlusions. Conclusion: Patients with more distal MCA occlusions (M3 and beyond) who are treated with tenecteplase are more likely to achieve mRS 0-1 at 90 days than those treated with alteplase. Differences in treatment effect with more proximal MCA and posterior cerebral artery occlusions were not observed.

Endovascular thrombectomy for acute large-vessel occlusion in patients with active cancer remains a difficult clinical decision. Multiple cohort studies and meta-analyses indicate that, when standard imaging and clinical criteria are applied, reperfusion success is high and symptomatic intracranial hemorrhage is broadly comparable with non-cancer populations, arguing against categorical exclusion. At the same time, registries show lower 90-day functional independence and higher 90-day mortality in active-cancer cohorts—effects likely driven by malignancy-related systemic factors and pre-existing functional compromise rather than procedural harm. This narrative review synthesizes efficacy and safety signals, highlights decision-grade outcomes that are seldom reported—early neurological change, performance status around 90 days, and whether systemic anticancer therapy is started or resumed. It sets out a clinical–ethical–economic framework to support selection, consent, and aftercare, including a pragmatic pathway for under-resourced settings based on non-contrast computed tomography (CT) or single-phase CT angiography where advanced perfusion imaging or magnetic resonance imaging are unavailable. We outline an Asia–Pacific collaborative program designed to identify subgroups that may benefit from reperfusion and initiation of active cancer care.

Introduction: The predictive value of CT markers of intracerebral haemorrhage (ICH) expansion is time-dependent, but data in the ultra-early period (<2 h from onset) are limited. We aimed to describe the frequency of these CT markers, their association with haematoma volume, haematoma expansion (HE) and functional outcome at 90-days. We also investigated the effect of tranexamic acid on HE in the presence of these markers. Patients and methods: We performed a pooled analysis of individual patient data from the STOP-AUST and STOP-MSU placebo-controlled randomised trials of tranexamic acid, including ICH patients scanned within 2 h of symptom onset. Logistic regression was used to assess the association between CT markers and HE or 90-days functional outcomes (poor outcome defined as mRS3-6). Results: Among 246 patients, the swirl sign (74.3%) was the most frequent CT marker and the blend sign least frequent (7.3%). All markers were associated with increased baseline haematoma volume, and excluding the black hole sign, all were more common in patients with 24-h HE. The blend and spot signs were associated with 24-h HE and heterogenous density, swirl sign, hypodensity and island sign were associated with poor 90-day function outcomes in univariate logistic regression. However, the area under the receiver-operating-characteristic curve was similar for all markers and indicated low discriminative ability (Chi-squared test p = 0.81). A potential benefit of tranexamic acid in HE reduction was observed in patients with the spot sign (interaction p = 0.01) Conclusions: The discriminative utility of CT markers of HE in the early timeframe appears insufficient. There may be an effect of tranexamic acid in spot sign positive patients <2 h from onset.

Background: Identifying patients likely to have significant hematoma expansion (HE) has been a challenge in clinical trials of intracerebral hemorrhage (ICH). Non-contrast CT (NCCT) markers of HE have been described. Time from symptom onset to CT affects the predictive value of these markers, with limited data in the ultra-early time period (<3h from onset). We aimed to describe the frequency of NCCT markers in the ultra-early period of ICH, and their association with HE. Methods: We performed a pooled analysis of individual patient data from the STOP-AUST and STOP-MSU randomized controlled trials, which tested the effect of tranexamic acid on hematoma growth. We included ICH patients scanned within 2 hours of symptom onset. The presence of NCCT density and shape markers was assessed by two authors. HE was defined as an increase in hematoma volume of >33% or >6ml from baseline. Regression analyses were adjusted for treatment group and baseline hematoma volume. Results: There were 246 patients included in this analysis (median age 67 years, 38.8% female, median time from onset to imaging 75 min [IQR 59-88 min], 50.4% tranexamic acid), of whom 105 (42.7%) had HE on 24-hour imaging. Inter-rater agreement was excellent for all NCCT markers (kappa score >0.8). Most patients (85.7%) had ≥1 marker of HE. The most frequent marker was the swirl sign (74.3%) and the least frequent was the blend sign (7.3%) (Table 1). HE occurred more often in patients with any marker at baseline (45.5% vs 25.6%, p=0.03). The blend sign (15.2% vs 1.4%, p<0.01), island sign (38.1% vs 24.8%, p=0.01) and satellite sign (44.8% vs 34.8%, p=0.04) were more common in patients with HE than those without. The presence of any marker was not independently associated with HE (aOR1.63 [95%CI 0.70-3.80], p=0.26). The blend sign was the only sign independently associated with HE (aOR 11.58 [95%CI 2.57-52.19], p=0.01). On diagnostic performance for HE, specificity was highest for the blend sign (0.99, 95% CI 0.95-1.00), with limited sensitivity of 0.15 (95%CI 0.09-0.24) (Table 1). The area under the receiver operator curve showed similar values for all markers suggesting low discriminative ability (Delong test p=0.81). Conclusions: NCCT markers are commonly present in the ultra-early period of ICH and associated with subsequent hematoma growth. Despite association with HE, the discriminative utility in the early timeframe appears insufficient, and the search for an optimal target should continue.

Objectives: Knowledge of the trajectory of post-stroke depression is important to identify high-risk patients, develop precise management programs and enhance prognosis. We aimed to characterise the course of depressive symptoms within the first year post-stroke and to evaluate associations with time.Materials and Methods: Depressive symptoms were measured using the Montgomery-Åsberg Depression Rating Scale (MADRS) within the first week, and at 3- and 12-months post-stroke. Scores were dichotomised into symptoms ‘present’ (MADRS ≥ 7) or ‘absent’ (MADRS < 7). The course of depressive symptoms within individuals was mapped and categorised using a trajectory diagram. The association between time and the presence of depressive symptoms was investigated using random effects logistic regression. Logistic regression was also used to assess the likelihood of participants having depressive symptoms later, given their status at earlier time points.Results: Of 142 ischaemic stroke survivors included for analysis, almost half (47.9%) experienced a change in depressive symptom status over time. Depressive symptoms were common at each timepoint (35-43%), although an association between time and frequency of depressive symptoms was not evident. Stroke survivors with depressive symptoms at 3 months were more likely to have depressive symptoms at 12 months, compared to those without symptoms at 3 months.Conclusion: Our findings provide evidence for a dynamic trajectory of depressive symptoms in individuals in the first year post-stroke. The importance of repeated screening for depression is highlighted, though most necessary at 3 months post-stroke.

BACKGROUND:The impact of Emergency Medical Service (EMS) stroke performance indicators on patient outcomes is unclear. We aimed to evaluate adherence to EMS indicators (15 minutes response, 60 minutes transport to a stroke centre or both) and associations with treatment access and patient outcomes. METHOD:Retrospective cohort study using data from the Australian Stroke Clinical Registry and ambulance records from Victoria (2015-2017) for patients (≥18 years) with stroke transported by ambulance. Multivariable regression models were used to evaluate EMS adherence to performance indicators and outcomes of interest. RESULTS:Among 4206 patients (72 years; 42 % female), EMS indicators were less frequently met for patients in regional areas (aOR: 0.27; 95 % CI: 0.17, 0.43). Socio-economic disparities were observed, with patients in advantaged areas less likely to receive an ambulance response within 15 minutes. Adherence to EMS performance indicators was associated with shorter off-stretcher time and increased thrombolytic therapy use (aOR: 1.62; 95 % CI: 1.24, 2.11). EMS adherence to performance indicators did not significantly impact functional independence, or health-related quality of life. CONCLUSIONS:EMS adherence to performance indicators improved off-stretcher time and thrombolytic therapy use but was influenced by geographic and socio-economic status. Refining EMS performance targets and tailoring public and EMS education on stroke is needed.

BACKGROUND:The STANDFIRM (Shared Team Approach Between Nurses and Doctors for Improved Risk Factor Management; ANZCTR registration ACTRN12608000166370) trial was designed to test the effectiveness of chronic disease care management for modifying the Framingham risk score (FRS) among patients with stroke or transient ischemic attack. The primary outcome of change in FRS was not met. We determine baseline characteristics that predict reduction in FRS at 12 months and whether future FRS is predetermined at baseline. METHODS AND RESULTS:We used machine learning regression methods to evaluate 35 variables encompassing demographics, risk factors, psychological, social and education status, and laboratory tests. We determine the optimal machine learning and associated tuning parameters from the following: random forest, extreme gradient boosting, category boosting, support vector regression, multilayer perceptron neural network, and K-nearest neighbor. Training (n=404) and test (n=103) data were evenly matched for age, sex, baseline, and 12-month FRS. The optimal model for predicting FRS at 12 months was category boosting (R2=0.712; root mean square error, 7.32). The 5 variables with highest Shapley values for category boosting were baseline FRS (Shapley additive explanation [SHAP], 8.42 of total of 12.12), age (SHAP, 1.58), systolic blood pressure (SHAP, 0.23), male sex (SHAP, 1.05), and London Handicap (SHAP, 0.18). Machine learning methods were poor at determining change in FRS at 12 months (R2<0.22). CONCLUSIONS:Our findings suggest that change in FRS as an end point in secondary stroke trials may have limited value as it is largely determined at baseline. In this cohort, category boosting was the optimal method to predict future FRS but not change in FRS.