This paper mainly introduces the design, advantages, disadvantages and its application of single case of randomized controlled trials (N-of-1 trials), and also introduces the commonly used data analysis methods of N-of-1 trials including nonparametric test and parameter test methods (t-test, paired t-test, analysis of variance), mixed-effects model, and meta-analysis. N-of-1 trials are suitable for individualized treatment, and could be expected to be widely used in the research of modern medicine and traditional Chinese medicine.
The CONSORT extension for reporting N-of-1 trials (CENT 2015) is designed to guide N-of-1 and series N-of-1 reporting. This study introduced the terminology (period, block or pair, sequence, washout period, and run-in period), the scope, the checklist and the diagram of CENT 2015 and demonstrated the complete guide frame for N-of-1, and thus to provide reference for relevant studies and improve the reporting quality of N-of-1 in China.
ObjectiveA simulation study was used to generate the multivariate normal distribution data with a residual effect based on series of N-of-1 trials. The statistical performance of paired t-test, mixed effect model and Bayesian mixed effect model were compared.MethodsThree-cycles N-of-1 trials were set, and the participants were randomly assigned to 2 different treatments in each cycle. The simulation study included the following procedures: producing six-dimensional normal distribution data, randomly allocating intervention methods and patients, adding residual effects, constructing and evaluating 3 models, and setting the parameters. The sample sizes were set as 3, 5, 8 and 10, and the correlation coefficients among different times were set as 0.0, 0.5 and 0.8. Different proportions of residual effects for the 2 groups were set. Type I error, power, mean error (ME), and mean square error (MSE) were used to compare the 3 models.ResultsWhen there was no residual effect in the 2 groups, type I errors of 3 models were approximately 0.05, and their MEs were approximately 0. Paired t-test had the highest power and the lowest MSE. When the residual effect existed in the 2 groups, the type I error of paired t-test increased, and its estimated value deviated from the true value (ME≠0). Type I errors of the mixed effect model and Bayesian mixed-effect model were approximately 0.05, and they had the same power. The estimated values of the two models were close to the true value (ME was approximately 0).ConclusionsWhen there is no residual effect (0% vs. 0%), paired t-test is suitable for data analysis of N-of-1 trials. When there is a residual effect, the mixed effect model and Bayesian mixed-effect model are suitable for data analysis of N-of-1 trials.
N-of-1 trials are prospective clinical randomized cross-over controlled trials with multiple rounds of trial phase alternation designed with regard to a single patient. N-of-1 trials can provide clinical decision-makers with high-level evidence of the comparison of effect of intervention measures. Recently, an international team composed of many scholars published a SPIRIT extension for N-of-1 trials list (SPENT 2019) on the BMJ, with the purposes of clarifying the content design and improving the integrity and transparency of N-of-1 trial protocols. This article showed a detailed interpretation of the 14 main extension sub-items of the SPENT 2019 list with specific cases, aiming to further standardize the publication of domestic N-of-1 trials.
Objective To improve the sensitivity and broaden the applicability of N-of-1 trials in traditional Chinese medicine (TCM), the clinical application and methodology of single-case experimental designs (N-of-1trials, multiple-baseline designs; MBDs) were expounded, compared, and discussed. Methods This paper introduced the current utility of N-of-1 trials in TCM research, introduced MBDs, and compared the methodologies of N-of-1 trials, MBDs and crossover design. Finally, two design schemes to improve the sensitivity and applicability of N-of-1 trials were illustrated. Results N-of-1 trials conformed to the TCM concept of treatment based on syndrome differentiation; however, due to the complex composition of TCM, the results were easily affected by carryover effect. In MBDs, the intervention was introduced in a staggered way, no washout period was needed, and the required sample size was small. MBDs were generally used to preliminarily indicate the effect of intervention; however, the statistical analysis was relatively complicated, and there were few MBDs used in clinical trials of TCM at present. Compared with crossover trials, single-case experimental designs had advantages and disadvantages. N-of-1 trials might best reflect the individualized treatment of TCM and a suitable statistical model (e.g., hierarchical Bayesian statistical method) was expected to improve the sensitivity and applicability of N-of-1 trials in TCM. Combining clinical trial designs (e.g., the combination of N-of-1 trials and MBDs) would complement the limitations of N-of-1 trials, and expand the scope of conditions applicable for study. Conclusion N-of-1 trials have both advantages and disadvantages in TCM research. Improved statistical models or combined study designs will improve the sensitivity and broaden the applicability of N-of-1 trials in TCM.
ObjectiveA series of single-case randomized controlled trials (N-of-1 trials), with placebo Chinese herbs used as a control, were conducted to observe the efficacy of the syndrome differentiation treatment formula in the stable phase of bronchiectasis by using a modified mixed-effects model (MEM) to detect the "carryover effects" of Chinese herbs, and to explore the establishment of an N-of-1 trial method that reflects the characteristics of syndrome differentiation treatment in traditional Chinese medicine (TCM). MethodsA single-center clinical trial was conducted in which a single case was studied in a multiple crossover, randomized controlled, and blinded manner. There were three rounds of the trial, each with two observation periods (treatment period and control period) of 4 weeks each. In the treatment period, an individualized formula based on syndrome differentiation was given, and in the control period, a placebo formula was administered. The primary indicator was the patients’ self-rated 7-point symptom Likert scale score, and other indicators included chronic obstructive pulmonary disease assessment test (CAT) score, 24 h sputum volume, TCM syndrome score, and safety index. Paired t test was used to analyze single case data and MEM designed for "carryover effects" was used to analyze group data. ResultsA total of 21 subjects were formally enrolled, and 15 (75%) completed three rounds of N-of-1 trials. Three of the cases showed statistically significant differences in overall symptom Likert scale score. At the group level, the MEM designed for "carryover effects" found statistically significant residual effects on three indicators (overall symptom score, respiratory symptom score, and CAT score). After excluding the "carryover effects", the model analyzed the statistically significant differences between the intervention effects of the two formulas on the overall symptom score, respiratory symptom score, CAT score and TCM syndrome score. The sensitivity of the MEM was higher than that of the meta-analysis when residual effects existed in the N-of-1 trials. ConclusionThe N-of-1 trials of Chinese herbs designed in this study can well demonstrate the characteristics of TCM syndrome differentiation and treatment. The modified MEM can detect the residual effects of TCM and improve the sensitivity of data statistics. However, due to the inherent nature of N-of-1 trials, the sensitivity of this study method at the individual level is low and more cases and diseases need to be studied for further improvement.