Lipid‐Lowering Trials Are Not Representative of Patients Managed in Clinical Practice: A Systematic Review and Meta‐Analysis of Exclusion Criteria
Martina Aeschbacher‐Germann,Nathalie Kaiser,Alexandre Speierer,Manuel R. Blum,Douglas C. Bauer,Cinzia Del Giovane,Drahomir Aujesky,Baris Gencer,Nicolas Rodondi,Elisavet Moutzouri
DOI: https://doi.org/10.1161/JAHA.122.026551
2022-12-26
Abstract:Background Randomized clinical trials (RCTs) might not be representative of the real‐world population because of unreasonable exclusion criteria. We sought to determine which groups of patients are excluded from RCTs that included lipid‐lowering therapy. Methods and Results We retrieved all trials from the Cholesterol Treatment Trialists Collaboration and systematically searched for large (≥1000 participants) lipid‐lowering therapy RCTs, defined as statins, ezetimibe, and PCSK9 inhibitors. We predefined groups: older adults (>70 or >75 years), women, non‐Whites, chronic kidney failure, heart failure, immunosuppression, cancer, dementia, treated thyroid disease, chronic obstructive pulmonary disease, mental illness, atrial fibrillation, multimorbidity (≥2 chronic diseases), and polypharmacy. We counted the number of RCTs excluding patients of the predefined groups and meta‐analyzed the prevalence of included patients to obtain pooled estimates with a random‐effects model. We included 42 RCTs (298 605 patients). Eighty‐one percent of trials excluded patients with severe and 76% those with moderate kidney failure. Seventy‐one percent of trials excluded groups of women, 64% excluded patients with moderate to severe heart failure, 64% those with immunosuppressant conditions, 48% those with cancer, 29% those with dementia, and 29% of trials excluded older adults. The pooled prevalence for patients >70 years of age was 25% (95% CI, 0%–49%), 11% (3%–18%) for >75 years of age, and 51% (38%–63%) for multimorbidity. Conclusions The majority of lipid‐lowering therapy trials excluded patients with common diseases, such as moderate‐to‐severe kidney disease or heart failure or with immunosuppression. Underrepresenting certain populations, including women and older adults, might lead to limited transportability of study results and uncertainty on possible side‐effects and efficacy in these groups. Future trials should promote diversity in the recruitment strategies and improve equity in cardiovascular research. Registration URL: ClinicalTrials.gov; Unique Identifier: CRD42021253909. Nonstandard Abbreviations and Acronyms CTTC Cholesterol Treatment Trialists Collaboration LLT lipid‐lowering therapy Clinical Perspective What Is New? Patients with conditions such as severe renal dysfunction (creatinine clearance <30 mL/min), severe congestive heart failure, immunosuppressant conditions, dementia, and mental illness were often excluded in large lipid‐lowering therapy trials, while more than two‐thirds of all studies excluded groups of women. One‐third of the trials excluded patients >75 years of age, with a pooled prevalence of elderly patients (>75 years of age) of only 11%, while no trial, even recent trials, listed multimorbidity or polypharmacy as exclusion or inclusion criterion, with a pooled prevalence of multimorbid patients (defined as ≥2 chronic diseases) of 51% (95% CI, 38%–63%), suggesting that we need more trials that specifically target these populations. What Are the Clinical Implications? Most lipid‐lowering trials excluded large groups of patients managed in clinical practice, which could possibly lead to uncertainty in treating these groups, while exclusion of groups of women may contribute to limited knowledge of the efficacy and safety of lipid‐lowering therapy in women compared with men. Mortality from cardiovascular disease (CVD) is high, causing 8.9 million deaths worldwide by 2019, 1 ≈45% of all deaths in Europe and 16% globally. 2 Primary and secondary CVD prevention measures, such as lifestyle changes or treatment with lipid‐lowering medication, can reduce CVD mortality. Guidelines widely recommend statins as secondary prevention. 3 However, their benefits and harms are less clear for primary prevention, because the net benefit of statins on absolute risk reduction depends on the individual's baseline risk of CVD, 3 , 4 , 5 and their use is more controversial. 6 Evidence of the effects of statins on CVD risk might have been drawn from trials with selected patient groups that do not represent real‐world demographics, as previously observed in RCTs in oth -Abstract Truncated-
English Else