Level of evidence supporting the Chinese cardiovascular disease clinical practice guidelines and its evolution in the past two decades.

BACKGROUND

Clinical practice is guided by guidelines in the era of evidence-based medicine to improve healthcare. The consistency between the strength of recommendations and the underlying quality of evidence in clinical guidelines and its evolution dynamically reflects the status of medical practice in important aspects. This study aimed to evaluate the levels of evidence (LOEs) supporting different classes of recommendations (CORs) in Chinese cardiovascular disease (CVD) guidelines between 2003 and 2021, and changes over time.

METHODS

Clinical guideline documents on cardiovascular topics issued by leading professional organizations were retrieved in the Databases of SinoMed and Wanfang Med Online from inception to June 2021. All guidelines were screened through abstract and full-text reading, and included if satisfying the pre-specified criteria. 79 Chinese guideline documents on 12 sub-topics including a total of 5195 recommendations and the designated CORs/LOEs, were abstracted. The number of recommendations of Class Ⅰ, Class Ⅱ, Class Ⅲ, LOE A, LOE B, and LOE C were identified for each guideline document. The proportion of CORs, LOEs, and COR-LOE combinations in guidelines and the changes among those with ≥2 versions.

FINDINGS

A total of 79 guidelines were included in the analysis. When examining the status of current guidelines, among the 3325 recommendations derived from 59 documents during 2011-2021, 735 recommendations (22.1%) were classified as LOE A, 1280 (38.5%) as LOE B, and 1310 (39.4%) as LOE C. 596 recommendations (17.9%) were characterized as Class Ⅰ-LOE A, accounting for the majority of LOE A recommendations but only one-third of Class I recommendations. Evidence levels varied greatly across different sub-topics and individual guidelines. There are 9 guidelines on 5 sub-topics having ≥2 versions. When analyzing the changes over time, although an increase was observed in the total number of recommendations, the proportion of recommendations designated as Class Ⅰ-LOE A did not significantly improve (19.1% [current] vs 19.0% [prior],  = 0.97).

INTERPRETATION

In current Chinese CVD guidelines, the high level of evidence lacks, and its alignment with strong recommendations is deficient. Although it shows moderate improvements in certain major topics (e.g., coronary artery disease, interventional therapy, surgery) in the past two decades, the overall proportion of Class I-LOE A recommendations remains small, suggesting that conduction, and particularly translation, of high-quality studies like RCTs addressing CVDs-related questions are still essential and demanded, especially for areas with less attention.

FUNDING

This study was supported by Beijing Nova Program from Beijing Municipal Science & Technology Commission (Z211100002121063, Z201100006820002); Fundamental Research Funds for the Central Universities (3332022023); National Key R&D Program of China (2020YFC2004705); CAMS Innovation Fund for Medical Sciences (2021-I2M-5-003); National Natural Science Foundation of China (81825003, 91957123, 82270376).