Global Stroke Corner
Johnston KC, Bruno A, Pauls Q, et al. Intensive vs Standard Treatment of Hyperglycemia and Functional Outcome in Patients With Acute Ischemic Stroke: The SHINE Randomized Clinical Trial. JAMA 2019;322(4):326-35.
Hyperglycemia during acute ischemic stroke is common and is associated with worse outcomes. The efficacy of intensive treatment of hyperglycemia in this setting remains unknown. Objectives: To determine the efficacy of intensive treatment of hyperglycemia during acute ischemic stroke. Design, Setting, and Participants: The Stroke Hyperglycemia Insulin Network Effort (SHINE) randomized clinical trial included adult patients with hyperglycemia (glucose concentration of >110 mg/dL if had diabetes or >/=150 mg/dL if did not have diabetes) and acute ischemic stroke who were enrolled within 12 hours from stroke onset at 63 US sites between April 2012 and August 2018; follow-up ended in November 2018. The trial included 1151 patients who met eligibility criteria. Interventions: Patients were randomized to receive continuous intravenous insulin using a computerized decision support tool (target blood glucose concentration of 80-130 mg/dL [4.4-7.2 mmol/L]; intensive treatment group: n = 581) or insulin on a sliding scale that was administered subcutaneously (target blood glucose concentration of 80-179 mg/dL [4.4-9.9 mmol/L]; standard treatment group: n = 570) for up to 72 hours. Main Outcomes and Measures: The primary efficacy outcome was the proportion of patients with a favorable outcome based on the 90-day modified Rankin Scale score (a global stroke disability scale ranging from 0 [no symptoms or completely recovered] to 6 [death]) that was adjusted for baseline stroke severity. Results: Among 1151 patients who were randomized (mean age, 66 years [SD, 13.1 years]; 529 [46%] women, 920 [80%] with diabetes), 1118 (97%) completed the trial. Enrollment was stopped for futility based on prespecified interim analysis criteria. During treatment, the mean blood glucose level was 118 mg/dL (6.6 mmol/L) in the intensive treatment group and 179 mg/dL (9.9 mmol/L) in the standard treatment group. A favorable outcome occurred in 119 of 581 patients (20.5%) in the intensive treatment group and in 123 of 570 patients (21.6%) in the standard treatment group (adjusted relative risk, 0.97 [95% CI, 0.87 to 1.08], P = .55; unadjusted risk difference, -0.83% [95% CI, -5.72% to 4.06%]). Treatment was stopped early for hypoglycemia or other adverse events in 65 of 581 patients (11.2%) in the intensive treatment group and in 18 of 570 patients (3.2%) in the standard treatment group. Severe hypoglycemia occurred only among patients in the intensive treatment group (15/581 [2.6%]; risk difference, 2.58% [95% CI, 1.29% to 3.87%]). Conclusions and Relevance: Among patients with acute ischemic stroke and hyperglycemia, treatment with intensive vs standard glucose control for up to 72 hours did not result in a significant difference in favorable functional outcome at 90 days. These findings do not support using intensive glucose control in this setting.
Kitagawa K, Yamamoto Y, Arima H, et al. Effect of Standard vs Intensive Blood Pressure Control on the Risk of Recurrent Stroke: A Randomized Clinical Trial and Meta-analysis. JAMA Neurol 2019 [published Online First: 2019/07/30]
The Systolic Blood Pressure Intervention Trial (SPRINT) demonstrated that a systolic blood pressure (BP) target less than 120 mm Hg was superior to less than 140 mm Hg for preventing vascular events. This trial excluded patients with prior stroke; therefore, the ideal BP target for secondary stroke prevention remains unknown. Objective: To assess whether intensive BP control would achieve fewer recurrent strokes vs standard BP control. Design, Setting, and Participants: Randomized clinical trial (RCT) of standard vs intensive BP control in an intent-to-treat population of patients who had a history of stroke. Patients were enrolled between October 20, 2010, and December 7, 2016. For an updated meta-analysis, PubMed and the Cochrane Central Library database were searched through September 30, 2018, using the Medical Subject Headings and relevant search terms for cerebrovascular disease and for intensive BP lowering. This was a multicenter trial that included 140 hospitals in Japan; 1514 patients who had a history of stroke within the previous 3 years were approached, but 234 refused to give informed consent. Interventions: In total, 1280 patients were randomized 1:1 to BP control to less than 140/90 mm Hg (standard treatment) (n = 640) or to less than 120/80 mm Hg (intensive treatment) (n = 640). However, 17 patients never received intervention; therefore, 1263 patients assigned to standard treatment (n = 630) or intensive treatment (n = 633) were analyzed. Main Outcomes and Measures: The primary outcome was stroke recurrence. Results: The trial was stopped early. Among 1263 analyzed patients (mean [SD] age, 67.2 [8.8] years; 69.4% male), 1257 of 1263 (99.5%) completed a mean (SD) of 3.9 (1.5) years of follow-up. The mean BP at baseline was 145.4/83.6 mm Hg. Throughout the overall follow-up period, the mean BP was 133.2/77.7 (95% CI, 132.5-133.8/77.1-78.4) mm Hg in the standard group and 126.7/77.4 (95% CI, 125.9-127.2/73.8-75.0) mm Hg in the intensive group. Ninety-one first recurrent strokes occurred. Nonsignificant rate reductions were seen for recurrent stroke in the intensive group compared with the standard group (hazard ratio [HR], 0.73; 95% CI, 0.49-1.11; P = .15). When this finding was pooled in 3 previous relevant RCTs in a meta-analysis, the risk ratio favored intensive BP control (relative risk, 0.78; 95% CI, 0.64-0.96; P = .02; absolute risk difference, -1.5%; 95% CI, -2.6% to -0.4%; number needed to treat, 67; 95% CI, 39-250). Conclusions and Relevance: Intensive BP lowering tended to reduce stroke recurrence. The updated meta-analysis supports a target BP less than 130/80 mm Hg in secondary stroke prevention.
Helwig SA, Ragoschke-Schumm A, Schwindling L, et al. Prehospital Stroke Management Optimized by Use of Clinical Scoring vs Mobile Stroke Unit for Triage of Patients With Stroke: A Randomized Clinical Trial. JAMA Neurol 2019
Transferring patients with large-vessel occlusion (LVO) or intracranial hemorrhage (ICH) to hospitals not providing interventional treatment options is an unresolved medical problem. Objective: To determine how optimized prehospital management (OPM) based on use of the Los Angeles Motor Scale (LAMS) compares with management in a Mobile Stroke Unit (MSU) in accurately triaging patients to the appropriate hospital with (comprehensive stroke center [CSC]) or without (primary stroke center [PSC]) interventional treatment. Design, Setting, and Participants: In this randomized multicenter trial with 3-month follow-up, patients were assigned week-wise to one of the pathways between June 15, 2015, and November 15, 2017, in 2 regions of Saarland, Germany; 708 of 824 suspected stroke patients did not meet inclusion criteria, resulting in a study population of 116 adult patients. Interventions: Patients received either OPM based on a standard operating procedure that included the use of the LAMS (cut point >/=4) or management in an MSU (an ambulance with vascular imaging, point-of-care laboratory, and telecommunication capabilities). Main Outcomes and Measures: The primary end point was the proportion of patients accurately triaged to either CSCs (LVO, ICH) or PSCs (others). Results: A predefined interim analysis was performed after 116 patients of the planned 232 patients had been enrolled. Of these, 53 were included in the OPM group (67.9% women; mean [SD] age, 74  years) and 63 in the MSU group (57.1% women; mean [SD] age, 75  years). The primary end point, an accurate triage decision, was reached for 37 of 53 patients (69.8%) in the OPM group and for 63 of 63 patients (100%) in the MSU group (difference, 30.2%; 95% CI, 17.8%-42.5%; P < .001). Whereas 7 of 17 OPM patients (41.2%) with LVO or ICH required secondary transfers from a PSC to a CSC, none of the 11 MSU patients (0%) required such transfers (difference, 41.2%; 95% CI, 17.8%-64.6%; P = .02). The LAMS at a cut point of 4 or higher led to an accurate diagnosis of LVO or ICH for 13 of 17 patients (76.5%; 6 triaged to a CSC) and of LVO selectively for 7 of 9 patients (77.8%; 2 triaged to a CSC). Stroke management metrics were better in the MSU group, although patient outcomes were not significantly different. Conclusions and Relevance: Whereas prehospital management optimized by LAMS allows accurate triage decisions for approximately 70% of patients, MSU-based management enables accurate triage decisions for 100%. Depending on the specific health care environment considered, both approaches are potentially valuable in triaging stroke patients.
Cramer SC, Dodakian L, Le V, et al. Efficacy of Home-Based Telerehabilitation vs In-Clinic Therapy for Adults After Stroke: A Randomized Clinical Trial. JAMA Neurol 2019 [published Online First: 2019/06/25]
Many patients receive suboptimal rehabilitation therapy doses after stroke owing to limited access to therapists and difficulty with transportation, and their knowledge about stroke is often limited. Telehealth can potentially address these issues. Objectives: To determine whether treatment targeting arm movement delivered via a home-based telerehabilitation (TR) system has comparable efficacy with dose-matched, intensity-matched therapy delivered in a traditional in-clinic (IC) setting, and to examine whether this system has comparable efficacy for providing stroke education. Design, Setting, and Participants: In this randomized, assessor-blinded, noninferiority trial across 11 US sites, 124 patients who had experienced stroke 4 to 36 weeks prior and had arm motor deficits (Fugl-Meyer [FM] score, 22-56 of 66) were enrolled between September 18, 2015, and December 28, 2017, to receive telerehabilitation therapy in the home (TR group) or therapy at an outpatient rehabilitation therapy clinic (IC group). Primary efficacy analysis used the intent-to-treat population. Interventions: Participants received 36 sessions (70 minutes each) of arm motor therapy plus stroke education, with therapy intensity, duration, and frequency matched across groups. Main Outcomes and Measures: Change in FM score from baseline to 4 weeks after end of therapy and change in stroke knowledge from baseline to end of therapy. Results: A total of 124 participants (34 women and 90 men) had a mean (SD) age of 61 (14) years, a mean (SD) baseline FM score of 43 (8) points, and were enrolled a mean (SD) of 18.7 (8.9) weeks after experiencing a stroke. Among those treated, patients in the IC group were adherent to 33.6 of the 36 therapy sessions (93.3%) and patients in the TR group were adherent to 35.4 of the 36 assigned therapy sessions (98.3%). Patients in the IC group had a mean (SD) FM score change of 8.36 (7.04) points from baseline to 30 days after therapy (P < .001), while those in the TR group had a mean (SD) change of 7.86 (6.68) points (P < .001). The covariate-adjusted mean FM score change was 0.06 (95% CI, -2.14 to 2.26) points higher in the TR group (P = .96). The noninferiority margin was 2.47 and fell outside the 95% CI, indicating that TR is not inferior to IC therapy. Motor gains remained significant when patients enrolled early (<90 days) or late (>/=90 days) after stroke were examined separately. Conclusions and Relevance: Activity-based training produced substantial gains in arm motor function regardless of whether it was provided via home-based telerehabilitation or traditional in-clinic rehabilitation. The findings of this study suggest that telerehabilitation has the potential to substantially increase access to rehabilitation therapy on a large scale.
Palmer R, Dimairo M, Cooper C, et al. Self-managed, computerised speech and language therapy for patients with chronic aphasia post-stroke compared with usual care or attention control (Big CACTUS): a multicentre, single-blinded, randomised controlled trial. Lancet Neurol 2019;18(9):821-33.
Post-stroke aphasia might improve over many years with speech and language therapy; however speech and language therapy is often less readily available beyond a few months after stroke. We assessed self-managed computerised speech and language therapy (CSLT) as a means of providing more therapy than patients can access through usual care alone. METHODS: In this pragmatic, superiority, three-arm, individually randomised, single-blind, parallel group trial, patients were recruited from 21 speech and language therapy departments in the UK. Participants were aged 18 years or older and had been diagnosed with aphasia post-stroke at least 4 months before randomisation; they were excluded if they had another premorbid speech and language disorder caused by a neurological deficit other than stroke, required treatment in a language other than English, or if they were currently using computer-based word-finding speech therapy. Participants were randomly assigned (1:1:1) to either 6 months of usual care (usual care group), daily self-managed CSLT plus usual care (CSLT group), or attention control plus usual care (attention control group) with the use of computer-generated stratified blocked randomisation (randomly ordered blocks of sizes three and six, stratified by site and severity of word finding at baseline based on CAT Naming Objects test scores). Only the outcome assessors and trial statistician were masked to the treatment allocation. The speech and language therapists who were doing the outcome assessments were different from those informing participants about which group they were assigned to and from those delivering all interventions. The statistician responsible for generating the randomisation schedule was separate from those doing the analysis. Co-primary outcomes were the change in ability to retrieve personally relevant words in a picture naming test (with 10% mean difference in change considered a priori as clinically meaningful) and the change in functional communication ability measured by masked ratings of video-recorded conversations, with the use of Therapy Outcome Measures (TOMs), between baseline and 6 months after randomisation (with a standardised mean difference in change of 0.45 considered a priori as clinically meaningful). Primary analysis was based on the modified intention-to-treat (mITT) population, which included randomly assigned patients who gave informed consent and excluded those without 6-month outcome measures. Safety analysis included all participants. This trial has been completed and was registered with the ISRCTN, number ISRCTN68798818. FINDINGS: From Oct 20, 2014, to Aug 18, 2016, 818 patients were assessed for eligibility, of which 278 (34%) participants were randomly assigned (101 [36%] to the usual care group; 97 [35%] to the CSLT group; 80 [29%] to the attention control group). 86 patients in the usual care group, 83 in the CSLT group, and 71 in the attention control group contributed to the mITT. Mean word finding improvements were 1.1% (SD 11.2) in the usual care group, 16.4% (15.3) in the CSLT group, and 2.4% (8.8) in the attention control group. Word finding improvement was 16.2% (95% CI 12.7 to 19.6; p<0.0001) higher in the CSLT group than in the usual care group and was 14.4% (10.8 to 18.1) higher than in the attention control group. Mean changes in TOMs were 0.05 (SD 0.59) in the usual care group (n=84), 0.04 (0.58) in the CSLT group (n=81), and 0.10 (0.61) in the attention control group (n=68); the mean difference in change between the CSLT and usual care groups was -0.03 (-0.21 to 0.14; p=0.709) and between the CSLT and attention control groups was -0.01 (-0.20 to 0.18). The incidence of serious adverse events per year were rare with 0.23 events in the usual care group, 0.11 in the CSLT group, and 0.16 in the attention control group. 40 (89%) of 45 serious adverse events were unrelated to trial activity and the remaining five (11%) of 45 serious adverse events were classified as unlikely to be related to trial activity. INTERPRETATION: CSLT plus usual care resulted in a clinically significant improvement in personally relevant word finding but did not result in an improvement in conversation. Future studies should explore ways to generalise new vocabulary to conversation for patients with chronic aphasia post-stroke.
Toyoda K, Uchiyama S, Yamaguchi T, et al. Dual antiplatelet therapy using cilostazol for secondary prevention in patients with high-risk ischaemic stroke in Japan: a multicentre, open-label, randomised controlled trial. Lancet Neurol 2019;18(6):539-48.
Although dual antiplatelet therapy with aspirin and clopidogrel reduces early recurrence of ischaemic stroke, with long-term use this type of therapy is no longer effective and the risk of bleeding increases. Given that cilostazol prevents stroke recurrence without increasing the incidence of serious bleeding compared with aspirin, we aimed to establish whether dual antiplatelet therapy involving cilostazol is safe and appropriate for long-term use. METHODS: In a multicentre, open-label, randomised controlled trial across 292 hospitals in Japan, patients with high-risk non-cardioembolic ischaemic stroke identified on MRI were randomly assigned to two groups in a 1:1 ratio to receive monotherapy with either oral aspirin (81 or 100 mg, once per day) or clopidogrel (50 or 75 mg, once per day) alone, or a combination of cilostazol (100 mg, twice per day) with aspirin or clopidogrel. Randomisation was done centrally (using block randomisation with a block size of six per each participating hospital) through a web-based registration system and was done by EPS Corporation. The patients were required to have at least 50% stenosis of a major intracranial or extracranial artery or two or more of the vascular risk factors. Trial medication was continued for half a year or longer, for a maximum of 3.5 years. The primary efficacy outcome was the rate of first recurrence of symptomatic ischaemic stroke. Safety outcomes were severe or life-threatening bleeding; any adverse events; serious adverse events; and any bleeding events. Efficacy analyses were done in the intention-to-treat population and safety analyses were done in the as-treated population. This trial was registered with ClinicalTrials.gov (number NCT01995370) and UMIN Clinical Trials Registry (number 000012180). FINDINGS: Participants were recruited from Dec 13, 2013, to March 31, 2017. 932 patients assigned to the dual therapy group and 947 patients assigned to the monotherapy group were included in the intention-to-treat analysis. The trial was stopped after the enrolment of 1884 patients of an anticipated 4000 patients because of the delay in recruitment. Ischaemic stroke recurred in 29 (3%) of 932 patients (annualised rate 2.2%) on dual therapy including cilostazol and 64 (7%) of 947 patients (annualised rate 4.5%) on monotherapy during a median 1.4 years follow-up (hazard ratio [HR] 0.49, 95% CI 0.31-0.76, p=0.0010). Severe or life-threatening bleeding occurred in eight patients (annualised rate 0.6%) on dual therapy and 13 patients (annualised rate 0.9%) on monotherapy (HR 0.66, 95% CI 0.27-1.60, p=0.35). Occurrence of any type of adverse event was similar between the groups (255 [28%] of 910 patients in the dual therapy group vs 219 [24%] of 921 patients in the monotherapy group); as was occurrence of serious adverse events (87 [10%] vs 142 [15%]) and bleeding events (38 [4%] vs 33 [4%]). Gastrointestinal bleeding, which affected nine (<1%) of 910 patients in the monotherapy group and nine (<1%) of 921 patients in the dual therapy group, was the most common type of bleeding. INTERPRETATION: The combination of cilostazol with aspirin or clopidogrel had a reduced incidence of ischaemic stroke recurrence and a similar risk of severe or life-threatening bleeding compared with treatment with aspirin or clopidogrel alone in patients at high risk for recurrent ischaemic stroke.
Yasuda S, Kaikita K, Akao M, et al. Antithrombotic Therapy for Atrial Fibrillation with Stable Coronary Disease. N Engl J Med 2019;381(12):1103-13.
There are limited data from randomized trials evaluating the use of antithrombotic therapy in patients with atrial fibrillation and stable coronary artery disease. METHODS: In a multicenter, open-label trial conducted in Japan, we randomly assigned 2236 patients with atrial fibrillation who had undergone percutaneous coronary intervention (PCI) or coronary-artery bypass grafting (CABG) more than 1 year earlier or who had angiographically confirmed coronary artery disease not requiring revascularization to receive monotherapy with rivaroxaban (a non-vitamin K antagonist oral anticoagulant) or combination therapy with rivaroxaban plus a single antiplatelet agent. The primary efficacy end point was a composite of stroke, systemic embolism, myocardial infarction, unstable angina requiring revascularization, or death from any cause; this end point was analyzed for noninferiority with a noninferiority margin of 1.46. The primary safety end point was major bleeding, according to the criteria of the International Society on Thrombosis and Hemostasis; this end point was analyzed for superiority. RESULTS: The trial was stopped early because of increased mortality in the combination-therapy group. Rivaroxaban monotherapy was noninferior to combination therapy for the primary efficacy end point, with event rates of 4.14% and 5.75% per patient-year, respectively (hazard ratio, 0.72; 95% confidence interval [CI], 0.55 to 0.95; P<0.001 for noninferiority). Rivaroxaban monotherapy was superior to combination therapy for the primary safety end point, with event rates of 1.62% and 2.76% per patient-year, respectively (hazard ratio, 0.59; 95% CI, 0.39 to 0.89; P = 0.01 for superiority). CONCLUSIONS: As antithrombotic therapy, rivaroxaban monotherapy was noninferior to combination therapy for efficacy and superior for safety in patients with atrial fibrillation and stable coronary artery disease.
Roshandel G, Khoshnia M, Poustchi H, et al. Effectiveness of polypill for primary and secondary prevention of cardiovascular diseases (PolyIran): a pragmatic, cluster-randomised trial. Lancet 2019;394(10199):672-83. doi: 10.1016/S0140-6736(19)31791-X [published Online First: 2019/08/27]
A fixed-dose combination therapy (polypill strategy) has been proposed as an approach to reduce the burden of cardiovascular disease, especially in low-income and middle-income countries (LMICs). The PolyIran study aimed to assess the effectiveness and safety of a four-component polypill including aspirin, atorvastatin, hydrochlorothiazide, and either enalapril or valsartan for primary and secondary prevention of cardiovascular disease. METHODS: The PolyIran study was a two-group, pragmatic, cluster-randomised trial nested within the Golestan Cohort Study (GCS), a cohort study with 50 045 participants aged 40-75 years from the Golestan province in Iran. Clusters (villages) were randomly allocated (1:1) to either a package of non-pharmacological preventive interventions alone (minimal care group) or together with a once-daily polypill tablet (polypill group). Randomisation was stratified by three districts (Gonbad, Aq-Qala, and Kalaleh), with the village as the unit of randomisation. We used a balanced randomisation algorithm, considering block sizes of 20 and balancing for cluster size or natural log of the cluster size (depending on the skewness within strata). Randomisation was done at a fixed point in time (Jan 18, 2011) by statisticians at the University of Birmingham (Birmingham, UK), independent of the local study team. The non-pharmacological preventive interventions (including educational training about healthy lifestyle-eg, healthy diet with low salt, sugar, and fat content, exercise, weight control, and abstinence from smoking and opium) were delivered by the PolyIran field visit team at months 3 and 6, and then every 6 months thereafter. Two formulations of polypill tablet were used in this study. Participants were first prescribed polypill one (hydrochlorothiazide 12.5 mg, aspirin 81 mg, atorvastatin 20 mg, and enalapril 5 mg). Participants who developed cough during follow-up were switched by a trained study physician to polypill two, which included valsartan 40 mg instead of enalapril 5 mg. Participants were followed up for 60 months. The primary outcome-occurrence of major cardiovascular events (including hospitalisation for acute coronary syndrome, fatal myocardial infarction, sudden death, heart failure, coronary artery revascularisation procedures, and non-fatal and fatal stroke)-was centrally assessed by the GCS follow-up team, who were masked to allocation status. We did intention-to-treat analyses by including all participants who met eligibility criteria in the two study groups. The trial was registered with ClinicalTrials.gov, number NCT01271985. FINDINGS: Between Feb 22, 2011, and April 15, 2013, we enrolled 6838 individuals into the study-3417 (in 116 clusters) in the minimal care group and 3421 (in 120 clusters) in the polypill group. 1761 (51.5%) of 3421 participants in the polypill group were women, as were 1679 (49.1%) of 3417 participants in the minimal care group. Median adherence to polypill tablets was 80.5% (IQR 48.5-92.2). During follow-up, 301 (8.8%) of 3417 participants in the minimal care group had major cardiovascular events compared with 202 (5.9%) of 3421 participants in the polypill group (adjusted hazard ratio [HR] 0.66, 95% CI 0.55-0.80). We found no statistically significant interaction with the presence (HR 0.61, 95% CI 0.49-0.75) or absence of pre-existing cardiovascular disease (0.80; 0.51-1.12; pinteraction=0.19). When restricted to participants in the polypill group with high adherence, the reduction in the risk of major cardiovascular events was even greater compared with the minimal care group (adjusted HR 0.43, 95% CI 0.33-0.55). The frequency of adverse events was similar between the two study groups. 21 intracranial haemorrhages were reported during the 5 years of follow-up-ten participants in the polypill group and 11 participants in the minimal care group. There were 13 physician-confirmed diagnoses of upper gastrointestinal bleeding in the polypill group and nine in the minimal care group. INTERPRETATION: Use of polypill was effective in preventing major cardiovascular events. Medication adherence was high and adverse event numbers were low. The polypill strategy could be considered as an additional effective component in controlling cardiovascular diseases, especially in LMICs.
Munoz D, Uzoije P, Reynolds C, et al. Polypill for Cardiovascular Disease Prevention in an Underserved Population. N Engl J Med 2019;381(12):1114-23.
Persons with low socioeconomic status and nonwhite persons in the United States have high rates of cardiovascular disease. The use of combination pills (also called "polypills") containing low doses of medications with proven benefits for the prevention of cardiovascular disease may be beneficial in such persons. However, few data are available regarding the use of polypill therapy in underserved communities in the United States, in which adherence to guideline-based care is generally low. METHODS: We conducted a randomized, controlled trial involving adults without cardiovascular disease. Participants were assigned to the polypill group or the usual-care group at a federally qualified community health center in Alabama. Components of the polypill were atorvastatin (at a dose of 10 mg), amlodipine (2.5 mg), losartan (25 mg), and hydrochlorothiazide (12.5 mg). The two primary outcomes were the changes from baseline in systolic blood pressure and low-density lipoprotein (LDL) cholesterol level at 12 months. RESULTS: The trial enrolled 303 adults, of whom 96% were black. Three quarters of the participants had an annual income below $15,000. The mean estimated 10-year cardiovascular risk was 12.7%, the baseline blood pressure was 140/83 mm Hg, and the baseline LDL cholesterol level was 113 mg per deciliter. The monthly cost of the polypill was $26. At 12 months, adherence to the polypill regimen, as assessed on the basis of pill counts, was 86%. The mean systolic blood pressure decreased by 9 mm Hg in the polypill group, as compared with 2 mm Hg in the usual-care group (difference, -7 mm Hg; 95% confidence interval [CI], -12 to -2; P = 0.003). The mean LDL cholesterol level decreased by 15 mg per deciliter in the polypill group, as compared with 4 mg per deciliter in the usual-care group (difference, -11 mg per deciliter; 95% CI, -18 to -5; P<0.001). CONCLUSIONS: A polypill-based strategy led to greater reductions in systolic blood pressure and LDL cholesterol level than were observed with usual care in a socioeconomically vulnerable minority population.
Attia ZI, Noseworthy PA, Lopez-Jimenez F, et al. An artificial intelligence-enabled ECG algorithm for the identification of patients with atrial fibrillation during sinus rhythm: a retrospective analysis of outcome prediction. Lancet 2019;394(10201):861-67. doi: 10.1016/S0140-6736(19)31721-0 [published Online First: 2019/08/06]
Atrial fibrillation is frequently asymptomatic and thus underdetected but is associated with stroke, heart failure, and death. Existing screening methods require prolonged monitoring and are limited by cost and low yield. We aimed to develop a rapid, inexpensive, point-of-care means of identifying patients with atrial fibrillation using machine learning. METHODS: We developed an artificial intelligence (AI)-enabled electrocardiograph (ECG) using a convolutional neural network to detect the electrocardiographic signature of atrial fibrillation present during normal sinus rhythm using standard 10-second, 12-lead ECGs. We included all patients aged 18 years or older with at least one digital, normal sinus rhythm, standard 10-second, 12-lead ECG acquired in the supine position at the Mayo Clinic ECG laboratory between Dec 31, 1993, and July 21, 2017, with rhythm labels validated by trained personnel under cardiologist supervision. We classified patients with at least one ECG with a rhythm of atrial fibrillation or atrial flutter as positive for atrial fibrillation. We allocated ECGs to the training, internal validation, and testing datasets in a 7:1:2 ratio. We calculated the area under the curve (AUC) of the receiver operatoring characteristic curve for the internal validation dataset to select a probability threshold, which we applied to the testing dataset. We evaluated model performance on the testing dataset by calculating the AUC and the accuracy, sensitivity, specificity, and F1 score with two-sided 95% CIs. FINDINGS: We included 180 922 patients with 649 931 normal sinus rhythm ECGs for analysis: 454 789 ECGs recorded from 126 526 patients in the training dataset, 64 340 ECGs from 18 116 patients in the internal validation dataset, and 130 802 ECGs from 36 280 patients in the testing dataset. 3051 (8.4%) patients in the testing dataset had verified atrial fibrillation before the normal sinus rhythm ECG tested by the model. A single AI-enabled ECG identified atrial fibrillation with an AUC of 0.87 (95% CI 0.86-0.88), sensitivity of 79.0% (77.5-80.4), specificity of 79.5% (79.0-79.9), F1 score of 39.2% (38.1-40.3), and overall accuracy of 79.4% (79.0-79.9). Including all ECGs acquired during the first month of each patient's window of interest (ie, the study start date or 31 days before the first recorded atrial fibrillation ECG) increased the AUC to 0.90 (0.90-0.91), sensitivity to 82.3% (80.9-83.6), specificity to 83.4% (83.0-83.8), F1 score to 45.4% (44.2-46.5), and overall accuracy to 83.3% (83.0-83.7). INTERPRETATION: An AI-enabled ECG acquired during normal sinus rhythm permits identification at point of care of individuals with atrial fibrillation.