Background: In 2014 we developed an evidence-based patient decision aid (PDA) for whether patients should take Tongxinluo capsule (a Chinese patent drug) or isosorbide dinitrate for angina, and validated it in 54 patients. Despite this the PDA was considered to be of 'moderate' quality according to IPDASi v4.0 (International Patient Decision Aids Standard instrument), the quality of clinical evidence was 'very low' as assessed by GRADE v3.2.2.
Objectives: Drawing on our experience of generating evidence from primary research in traditional medicine (TM), we aim to propose solutions to the current challenges of evidence production.
Methods: The PDA development team reviewed the process of evidence production and feedback from the validation study, identified problems encountered, and brainstormed possible solutions.
Results: We found that empirical evidence in TM, such as clinical experience supporting the use of an intervention, cannot be objectively graded or adequately used together with research evidence. Secondly, the conduct and reporting of clinical research was too low in quality to generate convincing evidence. One research project estimated that 7% of randomized controlled trials (RCTs) published in Chinese journals are real RCTs. Thirdly, it is debatable whether we should provide 'very low' quality evidence to patients, as they have other factors to consider before making a choice.
Conclusions: Possible solutions include: 1. Reinforce the implementation of CONSORT for TM and STRICTA (Standards for Reporting Interventions in Controlled Trials of Acupuncture), and develop standards for the conduct and reporting of TM clinical research; 2. Implement trial registration and results submission in TM, with a publically accessible database; 3. Formulate regulations for applicants of approved new drugs to report original data of clinical research used to support its approval to market; 4. Produce evidence based on utility as clinical decision-making involves multiple choices. Up-to-date methods such as network meta-analysis are recommended to compare multiple interventions on outcomes such as efficacy, safety, economics, acceptability, and time costs.