The relationship of publication language, study population, risk of bias, and treatment effects in acupuncture related systematic reviews: a meta-epidemiologic study
Bibliographic record
Abstract
Abstract Background There are debates in acupuncture related systematic reviews and meta-analyses on whether searching Chinese databases to get more Chinese-language studies may increase the risk of bias and overestimate the effect size, and whether the treatment effects of acupuncture differ between Chinese and non-Chinese populations. Methods In this meta-epidemiological study, we searched the Cochrane library from its inception until December 2021, and identified systematic reviews and meta-analyses with acupuncture as one of the interventions. Paired reviewers independently screened the reviews and extracted the information. We repeated the meta-analysis of the selected outcomes to separately pool the results of Chinese- and non-Chinese-language acupuncture studies and presented the pooled estimates as odds ratios (OR) with 95% confidence interval (CI). We calculated the Ratio of ORs (ROR) by dividing the OR of the Chinese-language trials by the OR of the non-Chinese-language trials, and the ROR by dividing the OR of trials addressing Chinese population by the OR of trials addressing non-Chinese population. We explored whether the impact of a high risk of bias on the effect size differed between studies published in Chinese- and in non-Chinese-language, and whether the treatment effects of acupuncture differed between Chinese and non-Chinese population. Results We identified 84 Cochrane acupuncture reviews involving 33 Cochrane groups, of which 31 reviews (37%) searched Chinese databases. Searching versus not searching Chinese databases significantly increased the contribution of Chinese-language literature both to the total number of included trials (54% vs. 15%) and the sample size (40% vs. 15%). When compared with non-Chinese-language trials, Chinese-language trials were associated with a larger effect size (pooled ROR 0.51, 95% CI 0.29 to 0.91). We also observed a higher risk of bias in Chinese-language trials in blinding of participants and personnel (97% vs. 51%) and blinding of outcome assessment (93% vs. 47%). The higher risk of bias was associated with a larger effect estimate in both Chinese-language (allocation concealment: high/unclear risk vs. low risk, ROR 0.43, 95% CI 0.21 to 0.87) and non-Chinese-language studies (blinding of participants and personnel: high/unclear risk vs. low risk, ROR 0.41, 95% CI 0.23 to 0.74). However, we found no evidence that the higher risk of bias would increase the effect size of acupuncture in Chinese-language studies more often than in non-Chinese-language studies (the confidence intervals of all ROR in the high-risk group included 1, Table 3). We further found acupuncture appeared to be more effective in Chinese than in non-Chinese population (Table 4). Conclusions The findings of this study suggest the higher risk of bias may lead to an overestimation of the treatment effects of acupuncture but would not increase the treatment effects in Chinese-language studies more often than in other language studies. The difference in treatment effects of acupuncture was probably associated with differences in population characteristics. Trial registration We registered our protocol on the Open Science Framework (OSF) ( https://doi.org/10.17605/OSF.IO/PZ6XR ).
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.192 | 0.415 |
| Meta-epidemiology (narrow) | 0.004 | 0.003 |
| Meta-epidemiology (broad) | 0.019 | 0.053 |
| Bibliometrics | 0.017 | 0.021 |
| Science and technology studies | 0.002 | 0.003 |
| Scholarly communication | 0.010 | 0.007 |
| Open science | 0.004 | 0.004 |
| Research integrity | 0.005 | 0.005 |
| Insufficient payload (model declined to judge) | 0.004 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".