Abstract TMP92: Pioglitazone for Secondary Stroke Prevention in Pre-diabetes: Implications for Real-world Practice

Pioglitazone reduces insulin resistance and cardiovascular risk among patients with type 2 diabetes. The Insulin Resistance Intervention after Stroke (IRIS) trial randomized 3876 non-diabetic participants with recent stroke/TIA and insulin resistance based on the Homeostasis Model of Insulin Resistance (HOMA-IR) score to pioglitazone (PIO) vs. placebo (PLA). Over 4.8 years, the relative risk reduction (RRR) with PIO was 27% for stroke/myocardial infarction, 25% for stroke, 29% for acute coronary syndrome (ACS), and 52% for new-onset diabetes. PIO could therefore be considered a new therapy for secondary stroke prevention before diabetes onset. However, it is not in widespread use, not only because of increased fracture risk, but perhaps also because HOMA-IR is not commonly assessed in clinical practice. In contrast, the presence of prediabetes is often identified in this population. We therefore analyzed the intention-to-treat (ITT) and the on-treatment effect of PIO (defined as adherence of > 80%) in IRIS participants with prediabetes by the American Diabetes Association criteria: HbA1C 5.7%-6.4%, or fasting glucose 100-125 mg/dL (n=2885; PIO 1456, PLA 1429). By ITT, for pioglitazone the hazard ratio and 95% confidence interval in Cox regression for stroke/MI was 0.70 (0.56-0.88, p=0.002), for stroke 0.72 (0.56-0.93, p=0.013), for ACS 0.72 (0.52-1.00, p = 0.052), for stroke/MI/hospitalized heart failure (HHF) 0.78 (0.63-0.96, p=0.02), and for new-onset diabetes HR 0.46 (0.34-0.61, p <0.0001). Among participants with good adherence, pioglitazone participants (n=644) were more likely to be men (76.2% vs 70.6%, p=0.02) and more likely to be smokers (15.9% vs. 11.7%, p=0.02) compared to PLA participants (n=810). In this subgroup, the HR for stroke/MI was 0.57 (0.39-0.84, p=0.004), for stroke 0.64 (0.42-0.99, p=0.04), for ACS 0.47 (0.26-0.85, p=0.01), for stroke/MI/HHF 0.61 (0.42-0.88, p=0.008), and for new-onset diabetes HR 0.18 (0.10-0.33, p<0.0001)

Conclusion: Pioglitazone should be considered efficacious for secondary prevention in stroke/TIA participants with prediabetes, simply assessed by fasting plasma glucose or A1C.

Abstract W P369: International Variability in Stroke Preventive Care

Introduction: Research protocols for stroke prevention trials commonly specify goals for preventive care. We report variability in goal achievement between 3 countries participating in an on-going stroke secondary prevention trial.

Methods: The Insulin Resistance Intervention after Stroke (IRIS) trial is testing pioglitazone, compared with placebo, for prevention of stroke and myocardial infarction among non-diabetic patients with a recent ischemic stroke/TIA. Preventive care is provided by personal physicians, although achievement of prevention goals is monitored and reported to participants and their physicians annually. Goals are from the American Heart Association guidelines: blood pressure (BP) <140/90 mmHg, statin therapy, and anti-platelet or anticoagulation depending on clinical indications. At baseline and year 1, we compared the proportions of participants meeting these goals in the largest enrolling countries: Canada (CA), United Kingdom (UK) and United States (US).

Results: Participant characteristics were similar across countries, except the proportions of women and blacks were lower in the UK and CA than the US, and self-reported hypertension was more common in US than CA or UK. At baseline, achievement of BP goal was lower in the UK (53%) compared with the US (66%) and CA (75%) (Chi 2 p<0.0001). Statin therapy was used more commonly in the UK (91%) and CA (88%) compared with the US (80%) (Chi 2 p<0.0001). Differences persisted at year 1. At baseline, use of antithrombotic therapy was high (99%) in all countries. However, at year 1, use fell in US (96%) compared to CA and UK (p=0.02).

Conclusions: Secondary preventive care for stroke varied among 3 countries for BP, statin therapy and antithrombotic therapies despite the IRIS protocol specifying uniform goals. These findings may be the result of disagreement among practitioners in the 3 countries for secondary prevention goals, variability in care delivery, or variability in research implementation. Understanding and resolving variability may lead to more efficient research and improved care for patients.

Boosting enrolment in clinical trials: validation of a regional network model

BACKGROUND

Clinical trials of stroke therapy have been hampered by slow rates of enrolment.

PURPOSE

Our purpose is to validate a previously developed model for accelerating enrolment in clinical trials by replicating it at new locations. The model employs coordinators who travel from a host institution to enrol participants from a network of participating hospitals. Active surveillance assures identification of all eligible patients.

METHODS

Among 70 U.S. investigators participating in National Institutes of Health-funded trial of stroke prevention, five investigators were invited to develop local identification and outreach networks (LIONs). Each LION comprised a LION coordinating centre servicing multiple hospitals. Hospitals provided names of patients with stroke or transient ischaemic attack to researchers at the LION coordinating centre who initiated contact; patients were offered home visits for consent and randomization. Outcomes were feasibility, enrolment, data quality, and cost.

RESULTS

Five LIONs varied in size from two to eight hospitals. All 24 hospitals we approached agreed to participate. The average monthly rate of enrolment at the research sites increased from 1.4 participants to 3.5 after expanding from a single institution model to the LION format (mean change = 2.1, range 0.9-3.7). Monthly performance improved over time. Data quality was similar for LIONs and non-LION sites, except for drug adherence which was lower at LIONs. The average cost to randomize and follow one participant during the study interval was 2.4 times the cost under the per-patient, cost-reimbursement strategy at non-LION sites. The cost ratio declined from 3.4 in year one to 1.8 in year two.

LIMITATIONS

The LION strategy requires unprecedented collaboration and trust among institutions. Applicability beyond stroke requires confirmation.

CONCLUSION

LIONs are a practical, reproducible method to increase enrolment in trial research. Twelve months were required for the average site to reach its potential. The per-participant cost at LIONs was higher than conventional sites but declined over time.

Internal medicine clerkship directors' perceptions about student interest in internal medicine careers

BACKGROUND

Experienced medical student educators may have insight into the reasons for declining interest in internal medicine (IM) careers, particularly general IM.

OBJECTIVE

To identify factors that, according to IM clerkship directors, influence students' decisions for specialty training in IM.

DESIGN

Cross-sectional national survey.

PARTICIPANTS

One hundred ten institutional members of Clerkship Directors in IM.

MEASUREMENTS

Frequency counts and percentages were reported for descriptive features of clerkships, residency match results, and clerkship directors' perceptions of factors influencing IM career choice at participating schools. Perceptions were rated on a five-point scale (1 = very much pushes students away from IM careers; 5 = very much attracts students toward IM careers).

RESULTS

Survey response rate was 83/110 (76%); 80 answered IM career-choice questions. Clerkship directors identified three educational items attracting students to IM careers: quality of IM faculty (mean score 4.3, SD = 0.56) and IM rotation (4.1, SD = 0.67), and experiences with IM residents (3.9, SD = 0.94). Items felt most strongly to push students away from IM careers were current practice environment for internists (mean score 2.1, SD = 0.94), income (2.1, SD = 1.08), medical school debt (2.3, SD = 0.89), and work hours in IM (2.4, SD = 1.05). Factor analysis indicated three factors explaining students' career choices: value/prestige of IM, clerkship experience, and exposure to internists.

CONCLUSIONS

IM clerkship directors believe that IM clerkship experiences attract students toward IM, whereas the income and lifestyle for practicing internists dissuade them. These results suggest that interventions to enhance the practice environment for IM could increase student interest in the field.