Representation of women, older patients, ethnic, and racial minorities in trials of atrial fibrillation

BACKGROUND

Representation trends of women, older adults, and ethnic/racial minorities in randomized controlled trials (RCTs) of atrial fibrillation (AF) are uncertain.

METHODS

We systematically reviewed 134 AF related RCTs (phase II and III) encompassing 149,162 participants using Medline and ClinicalTrials.gov through April 2019 to determine representation trends of women, older patients (≥75 years), and ethnic/racial minorities. Weighted data on the prevalence of AF from epidemiological studies were used to compare the representation of the studied groups of interest in AF RCTs to their expected burden of the disease.

RESULTS

Only 18.7% of the RCTs reported proportion of older patients, and 12.7% RCTs reported ethnic/racial minorities. The proportions of women in RCTs versus general population were 35.2% and 35.1%, of Hispanics were 11.9% and 5.2%, of Blacks were 1.2% and 5.7%, of American Indian/Alaskans were 0.2% and 0.2%, of Asians were 14.2% and 2.4%, of native Hawaiian/Pacific Islanders were 0.05% and 0.1% and of non-Whites were 19.5% and 22.5%, respectively. The weighted mean age (SD) across the trials was 65.3 (3.2) years which was less than the corresponding weighted mean age of 71.1 (4.5) years in the comparative epidemiological data.

CONCLUSION

The reporting of older patients and ethnic/racial minorities was poor in RCTs of AF. The representation of women and American Indian/Alaskan natives matched their expected population share of disease burden. Hispanics and Asians were over-represented and Blacks, native Hawaiian/Pacific Islanders and non-Whites were under-represented in RCTs of AF.

Abstract 239: Contemporary Trends in Outcomes and Resource Utilization in Sudden Cardiac Arrest With Chronic Liver Disease

Background: Chronic Liver disease (CLD) with sudden cardiac arrest (SCA) is associated higher mortality and morbidity due to pro inflammatory state in CLD.

Methods: Patients with CLD with SCA were selected from national inpatient sample from 2005-15 using relevant International Classification of Diseases, Ninth Revision (ICD-9) codes. Pearson's chi-squared test and independent samples t-test was used for statistical testing.

Results: 69,113 weighted encounters were identified. Mortality was 64% (44,232 of 69,113). Mean age was 56.4(13.8). 42.9% (29,650 of 69,113) were women. Mean length of stay was 6.2 days. Mean cost was just $119,801. Of the surviving patients, 48.9% (12,145 of 24,881) were discharged home, 10.7% (2,660 of 24,881) to short term hospital, 38.7% (9,624 of 24,881) to long term care hospital. 0.4% (275 of 69,113) had implantable cardiac defibrillator, 8.2% (5,667 of 69,113) underwent left heart catheterization, 3% (2,097 of 69,113) underwent Percutaneous Coronary Intervention, 1.8% (1,212 of 69,113) had Percutaneous ventricular assist device placement.

Conclusion: Recent trends have shown decreased mortality and decreased length of stay in chronic liver disease with SCA, however, there has been an increased cost of hospital stay and resource utilization. Underlying reasons for this need further investigation.

Abstract 28: Health Care Cost Burden of Infective Endocarditis in United States: A Nationwide Study

Introduction: Aging population and recent opioid epidemic have contributed to increase incidence and cost of Infective endocarditis (IE). We aimed to assess health care burden of IE using nationally representative derived data that is designed for inpatient health care cost and utilization.

Methods: Using the National inpatient sample (NIS), we examined 591,871 hospitalizations with a diagnosis of IE from 2002 to 2016. Risk Adjusted rates were calculated using univariate generalized logistic model for analysis of co variance. Trends were assessed with Linear regression and Pearson’s Chi square when appropriate.

Results: Mean age of admission was 59.2[SD;18.8]. Females represented 40.9% (242,024 of 591871) of admissions. The proportion of patients with history of drug use increased from 11.5% (3,361 of 29,226) in 2002 to 23.1% (12,385 of 53,615) in 2016. We noticed an overall downtrend in mortality, from 16.4% (4,450 of 27,134) in 2002 to 8.8% (5,370 of 61022) in 2016. There was also a downtrend in the mean length of stay (16.2 days in 2002 to 14.2 days in 2016, p<0.01) but increase in the adjusted for inflation average cost ($100,643 in 2002 to $168,734 in 2016, p<0.01). There was also an increase in hospital-based procedures like valvular surgery.

Conclusion: In-hospital mortality from IE has been on a decline but total hospitalization and average cost of stay in IE has increased. The reasons behind this warrant further investigation.

Abstract 208: Trends of Comorbidities in Clinical Trials of Lipid Lowering Therapies: A Systematic Review

Background: Prevalence of coexisting comorbidities in clinical trials of lipid lowering therapies remains uncertain.

Methods: We selected 61 trials ( n = 485,678) with follow-up of ≥3 months from databases of MEDLINE and ClinicalTrials.gov through December 2018. Trends in the prevalence of comorbid conditions across study period were tested using simple linear regression.

Results: There was a downtrend in enrollment of patients with coronary artery disease (CAD) (57.1% [18,968 of 33,230] to 30.9% [59,078 of 190,930]) from 1993-1998 period to 2014-2018 and peripheral artery disease (PAD) (11.6% [9062 of 78,249] to 4.2% [7996 of 190,930]) from 1999-2003 period to 2014-2018; however, there was a significant increase in enrollment of hypertensives from 23.8% (7897 of 33,230) to 59.2% (112,978 of 190930) from 1993-1998 period to 2014-2018. Enrollment of cerebrovascular accident (CVA) improved from 2.1% (701 of 33,230) to 7.2% (13,836 of 190,930) during the same periods. Enrollment of patients with chronic kidney disease (CKD), chronic obstructive pulmonary disease (COPD) and cancer remained consistently low. The overall recruitment of patients with smoking (20.6% [55004 of 267631]) and heart failure (HF) (7.6% [20395 of 267631]) was higher among industry sponsored trials and PAD (10.2% [2279 of 22340) among university funded trials. The reporting on cancer, COPD and CKD was provided only by industry sponsored trials.

Conclusion: Overall reporting on comorbidity was consistently low across trials of lipid lowering therapy.

Abstract 103: A Meta Analysis of Antibiotic Envelope Use to Prevent Cardiac Implantable Electronic Device Infections

Introduction: Cardiac Implantable Electronic Device (CIED) infections are a major source of morbidity, mortality and increased healthcare cost. There are several well established risk factors associated with increased infection. An antibiotic coated envelope has been developed as an infection prevention intervention.

Hypothesis: Whether antibiotic coated envelope is an effective preventive strategy against Cardiac Implantable Electronic Device Infections.

Methods: 6 trials comparing use of antibiotic envelope to control, including 1 randomized trial and 5 cohort studies were selected using PubMed and Embase data bases through May 2019. The efficacy end point was prevention of CIED Infection. Outcomes were combined using random effects model and estimated by odds ratio with 95% confidence intervals (CI).

Results: In our analysis of total 12,024 patients undergoing CIED implantation, 5,844 patients received antibiotic envelope while 6,180 patients were included in the control group. Antibiotic envelope was superior to control in reducing the risk of device infection (OR 0.40 CI [0.17-0.95], p=0.04).

Conclusions: The use of antibiotic envelope in Cardiac Implantable Electronic Devices (CIED) is associated with reduced incidence of CIED infection.

Participation of Women and Older Participants in Randomized Clinical Trials of Lipid-Lowering Therapies: A Systematic Review

IMPORTANCE

Randomized clinical trials (RCTs) of lipid-lowering therapies form the evidence base for national and international guidelines. However, concerns exist that women and older patients are underrepresented in RCTs.

OBJECTIVE

To determine the trends of representation of women and older patients (≥65 years) in RCTs of lipid-lowering therapies from 1990 to 2018.

DATA SOURCES

The electronic databases of MEDLINE and ClinicalTrials.gov were searched from January 1990 through December 2018.

STUDY SELECTION

RCTs of lipid-lowering therapies with sample sizes of at least 1000 patients and follow-up periods of at least 1 year were included.

DATA EXTRACTION AND SYNTHESIS

Two independent investigators abstracted the data on a standard data collection form.

MAIN OUTCOMES AND MEASURES

Patterns of representation of women and older adults were examined overall in lipid-lowering RCTs and according to RCT-level specific characteristics. The participation-to-prevalence ratio (PPR) metric was used to estimate the representation of women compared with their share of disease burden.

RESULTS

A total of 60 RCTs with 485 409 participants were included. The median (interquartile range) number of participants per trial was 5264 (1062-27 564). Overall, representation of women was 28.5% (95% CI, 24.4%-32.4%). There was an increase in the enrollment of women from the period 1990 to 1994 (19.5%; 95% CI, 18.4%-20.5%) to the period 2015 to 2018 (33.6%; 95% CI, 33.4%-33.8%) (P for trend = .01). Among common limiting factors were inclusion of only postmenopausal women or surgically sterile women (28.3%; 95% CI, 18.5%-40.7%) or exclusion of pregnant (23.3%; 95% CI, 14.4%-35.4%) and lactating (16.6%; 95% CI, 9.3%-28.1%) women. Women were underrepresented compared with their disease burden in lipid RCTs of diabetes (PPR, 0.74), heart failure (PPR, 0.27), stable coronary heart disease (PPR, 0.48), and acute coronary syndrome (PPR, 0.51). Only 23 RCTs with 263 628 participants reported the proportion of older participants. Overall representation of older participants was 46.7% (95% CI, 46.5%-46.9%), which numerically increased from 31.6% (95% CI, 30.8%-32.3%) in the period 1995 to 1998 to 46.2% (95% CI, 46.0%-46.5%) in the period 2015 to 2018 (P for trend = .43). A total of 53.0% (95% CI, 41.8%-65.3%) and 36.6% (95% CI, 25.6% to 49.3%) trials reported outcomes according to sex and older participants, respectively, which did not improve over time.

CONCLUSIONS AND RELEVANCE

In this systematic review of RCTs of lipid-lowering therapies, the enrollment of women and older participants increased over time, but women and older participants remained consistently underrepresented. This limits the evidence base for efficacy and safety in these subgroups.

Abstract 207: Cardiovascular Outcomes of Sodium-glucose Co-transporter 2 Inhibitors: A Meta-analysis

Background: Concerns exist that mortality benefit might vary across different drugs (canagliflozin, empagliflozin, dapagliflozin) belonging to sodium-glucose cotransporter-2 (SGLT2) inhibitors.

Methods: Twenty-two randomized controlled trials (n=59,018) having at least 500 participants and follow-up ≥1 year were selected from MEDLINE and EMBASE databases through October 2019. Meta-analysis was performed using fixed effect model (for I 2 < 25% {25 out of 100}) or random effects model.

Results: SGLT2 inhibitors decreased risk of all-cause mortality (Hazard Ratio (HR) 0.85, [95% {95 out of 100} confidence interval, 0.79-0.91]). This benefit was consistent for canagliflozin (HR, 0.86, [0.76-0.97]), empagliflozin (HR, 0.70, [0.79-0.91]) and dapagliflozin (HR, 0.90, [0.82-0.99]) (P-interaction = 0.06). SGLT2 inhibitors reduced the risk of cardiovascular mortality (HR, 0.82 [0.74-0.90]), myocardial infarction (MI) (HR, 0.88 [0.80-0.97]), major adverse cardiovascular events (HR, 0.84 [0.78-0.91]), and heart failure hospitalizations (HR, 0.68 [0.62-0.75]).

Conclusion: Canagliflozin, empagliflozin, dapagliflozin were associated with significant reduction in all-cause mortality. SGLT2 inhibitor was also associated with cardiovascular benefits in type 2 diabetes mellitus.