• The Korean Journal of Applied Statistics
• /
• 제33권4호
• /
• pp.439-449
• /
• 2020

Meta-analysis provides a way of integrating several independent studies of interest. Since small studies with statistically significant results are more likely to be published, publication bias, which is a special case of selection bias, often occurs in meta analysis. Conditional likelihood and weighted estimating equation have been proposed to deal with publication bias, but they require to specify a correct selection probability model. In contrast, the pairwise pseudolikelihood approach can correct publication bias without fully specifying the correct selection probability model, but its performance in meta-analysis was not investigated. In this paper, we perform a numerical study about whether the pairwise pseudolikelihood approach is effective for solving publication bias arising from typical meta-analysis settings.

• Korean Journal of Acupuncture
• /
• 제35권2호
• /
• pp.70-81
• /
• 2018

Objectives : To investigate the association of trial registration status with presence of reporting bias including publication bias and outcome reporting bias in recently published randomized controlled trials (RCTs) of acupuncture. Methods : A PubMed search for RCTs of acupuncture published from March 2016 to February 2017 was conducted. Primary outcomes were identified and the direction of the results was judged as positive (i.e., statistically significant) or negative. The trial registration was identified by manually screening the trial registration number in the main text of the published article and classified into 1) prospective registration; 2) retrospective registration based on the registration date or; 3) no registration. Results : Of the 125 included RCTs, only 40 studies (32.0%) prospectively registered the study protocols. Among 65 RCTs that adequately reported the primary outcome, unregistered trials were more likely to report positive results than the registered ones (p=0.013). Of the 40 prospectively registered studies, 19 trials (47.5%) had the discrepancies between the registered and published primary outcomes and furthermore, 40% of them reported the positive findings. Conclusions : Unregistered trials were more likely to report positive results and the discrepancies between the registered and published primary outcomes were detected in about a half of the prospectively registered studies, 42.1% of which tended to report positive findings. Journal editors and researchers in this field should be alerted to various reporting biases.

• The Korean Journal of Applied Statistics
• /
• 제22권1호
• /
• pp.107-114
• /
• 2009

The present study was carried out to summarize the effect of ginseng in the experimental diabetic rats by meta-analysis related studies. The association measure to test effect of ginseng was the mean difference(MD) between group of rats induced streptozotocin(STZ) and group of rats induced STZ treated with ginseng about the considered effect factors. The level of FI, glucose and TG were significantly reduced(< 0.01), and the level of glycogen was significantly increased by treatment with ginseng (< 0.01) After checking the indication of publication bias for the combined MDs by using the funnel plots, the anti-diabetic effects of ginseng is clearly presented in FI, glucose, TG and glycogen (< 0.05).

• Journal of Preventive Medicine and Public Health
• /
• 제43권2호
• /
• pp.185-192
• /
• 2010

Objectives: We conducted a meta-analysis to investigate the relationship between low external doses of ionizing radiation exposure and the risk of cancer mortality among nuclear power plant workers. Methods: We searched MEDLINE using key words related to low dose and cancer risk. The selected articles were restricted to those written in English from 1990 to January 2009. We excluded those studies with no fit to the selection criteria and we included the cited references in published articles to minimize publication bias. Through this process, a total of 11 epidemiologic studies were finally included. A publication bias was tested for using Egger's test. The homogeneity test was performed before the integration of each of the standardized mortality ratios (SMRs) and the result proved that the studies were heterogeneous. Results: We found significant decreased deaths from all cancers (SMR = 0.75, 95% CI = 0.62 - 0.90), all cancers excluding leukemia, solid cancer, mouth and pharynx, esophagus, stomach, rectum, liver and gallbladder, pancreas, lung, prostate, lymphopoietic and hematopoitic cancer. The findings of this meta-analysis were similar with those of the 15 Country Collaborative Study conducted by the International Agency for Research on Cancer. A publication bias was found only for liver and gallbladder cancer (p = 0.015). Heterogeneity was observed for all cancers, all cancers excluding leukemia, solid cancer, esophagus, colon and lung cancer. Conclusions: Our findings of low mortality for stomach, rectum, liver and gallbladder cancers may explained by the health worker effect. Yet further studies are needed to clarify the low SMR of cancers, for which there is no useful screening tool, in nuclear power plant workers.

• Asian-Australasian Journal of Animal Sciences
• /
• 제31권11호
• /
• pp.1781-1794
• /
• 2018

Objective: This meta-analysis was conducted to evaluate the overall effect of direct-fed microbial (DFM) or probiotic supplementation on the log concentrations of culturable gut microbiota in broiler chickens. Methods: Relevant studies were collected from PubMed, SCOPUS, Poultry Science Journal, and Google Scholar. The studies included controlled trials using DFM supplementation in broiler chickens and reporting log concentrations of the culturable gut microbiota. The overall effect of DFM supplementation was determined using standardized mean difference (SMD) with a random-effects model. Subgroups were analyzed to identify pre-specified characteristics possibly associated with the heterogeneity of the results. Risk of bias and publication bias were assessed. Results: Eighteen taxa of the culturable gut microbiota were identified from 42 studies. The overall effect of DFM supplementation on the log concentrations of all 18 taxa did not differ significantly from the controls (SMD = -0.06, 95% confidence interval [-0.16, 0.04], p = 0.228, $I^2=85%$, n = 699 comparisons), but the 18 taxa could be further classified into three categories by the direction of the effect size: taxa whose log concentrations did not differ significantly from the controls (category 1), taxa whose log concentrations increased significantly with DFM supplementation (category 2), and taxa whose log concentrations decreased significantly with DFM supplementation (category 3). Category 1 comprised nine taxa, including total bacterial counts. Category 2 comprised four taxa: Bacillus, Bifidobacterium, Clostridium butyricum, and Lactobacillus. Category 3 comprised five taxa: Clostridium perfringens, coliforms, Escherichia coli, Enterococcus, and Salmonella. Some characteristics identified by the subgroup analysis were associated with result heterogeneity. Most studies, however, were present with unclear risk of bias. Publication bias was also identified. Conclusion: DFM supplementation increased the concentrations of some beneficial bacteria (e.g. Bifidobacterium and Lactobacillus) and decreased those of some detrimental bacteria (e.g. Clostridium perfringens and Salmonella) in the guts of broiler chickens.

• Journal of The Korean Society of Integrative Medicine
• /
• 제10권3호
• /
• pp.37-52
• /
• 2022

Purpose : The purpose of this meta-analysis was to evaluate the effects of electrical stimulation on patients with non-specific low back pain. Methods : Domestic databases were gathered from studies that conducted clinical trials associated with electrical stimulation and its impact on pain of non-specific low back patients. A total of 681 studies were identified, with 12 studies satisfying the inclusion data. The studies consisted of patient, intervention, comparison, outcome, and study design (PICO-SD). The search outcomes were items associated with low back pain. Cochrane risk of bias 2 (RoB 2) was used to evaluate the quality of 12 randomized controlled trials. Effect sizes (Hedges's g) in this study were computed as the corrected standard mean difference (SMD). A random-effect model was used to analyze the effect size because of the high heterogeneity among the studies. Egger's regression and 'trim-and-fill' tests were carried out to analyze the publication bias. Cumulative meta-analysis and sensitivity analysis were conducted to analyze the effect according to the sample size and the consistency of the effect size. Results : The following factors had a large overall effect size (Hedges's g=1.28, 95 % CI=.20~2.36) involving electrical stimulation on non-specific low back pain. The subgroup analysis all showed a statistical difference in the types of study design, electrical stimulation, and assessment tool. No statistically significant difference was found in the meta-regression analysis. Publican bias was found in the data. Conclusion : The findings in this study indicate that electrical stimulation interventions have a positive effect on patients with non-specific low back pain. However, due to the low quality of studies and publication bias, the results of our study should be interpreted cautiously.

• Journal of Korean Academy of Nursing
• /
• 제41권6호
• /
• pp.821-833
• /
• 2011

Purpose: The purpose of this study was to evaluate the effectiveness of foot reflexology on fatigue, sleep and pain. Methods: A systematic review and meta-analysis were conducted. Electronic database and manual searches were conducted on all published studies reporting the effects of foot reflexology on fatigue, sleep, and pain. Forty four studies were eligible including 15 studies associated with fatigue, 18 with sleep, and 11 with pain. The effects of foot reflexology were analyzed using Comprehensive Meta-Analysis Version 2.0. The homogeneity and the fail-safe N were calculated. Moreover, a funnel plot was used to assess publication bias. Results: The effects on fatigue, sleep, and pain were not homogeneous and ranged from 0.63 to 5.29, 0.01 to 3.22, and 0.43 to 2.67, respectively. The weighted averages for fatigue, sleep, and pain were 1.43, 1.19, and 1.35, respectively. No publication bias was detected as evaluated by fail-safe N. Foot reflexology had a larger effect on fatigue and sleep and a smaller effect on pain. Conclusion: This meta-analysis indicates that foot reflexology is a useful nursing intervention to relieve fatigue and to promote sleep. Further studies are needed to evaluate the effects of foot reflexology on outcome variables other than fatigue, sleep and pain.

• Asian Pacific Journal of Cancer Prevention
• /
• 제14권3호
• /
• pp.1697-1701
• /
• 2013

Background: Studies of associations between genetic polymorphism of glutathione S-transferase M1 (GSTM1) and glutathione S-transferase T1 (GSTT1) with risk of nasopharyngeal cancer (NPC) have generated conflicting results. Thus, a meta-analysis was performed to clarify the effects of GSTM1 and GSTT1 polymorphisms on the risk of developing NPC. Materials and Methods: A literature search in two electronic databases namely PubMed and EMBASE up to December 2012 was conducted and eligible papers were finally selected based on the inclusion and exclusion criteria. The pooled odds ratio (OR) and presence of heterogeneity and publication bias in those studies were evaluated. Results: A total of 9 studies concerning nasopharyngeal cancer were evaluated. Analyses of all relevant studies showed increased NPC risk to be significantly associated with the null genotypes of GSTMI (OR=1.43, 95%CI 1.24-1.66) and GSTT1 (OR=1.28, 95%CI=1.09-1.51). In addition, evidence of publication bias was detected among the studies on GSTM1 polymorphism. Conclusions: This meta-analysis demonstrated the GSTM1 and GSTT1 null genotypes are associated with an increased risk of NPC.

• Korean Journal of Adult Nursing
• /
• 제29권2호
• /
• pp.177-189
• /
• 2017

Purpose: A meta-analysis was conducted to identify the effect of cognitive improvement programs for the elderly with mild cognitive impairment (MCI) in Korea. Methods: Five databases, as well as relevant reference lists, of studies published from 2000 to 2016, were searched. Fourteen studies were identified. Quality assessments of included studies were conducted using the Scottish Intercollegiate Guidelines Network checklist. An R program was used to analyze effect sizes and to identify possible sources of heterogeneity among studies. The potential for publication bias was investigated using a funnel plot, Egger's regression test and sensitivity analysis. Results: The total effect size was large (Standardized Mean Difference [SMD]=1.44, 95% CI: 1.11~1.77), with cognition based intervention (SMD= 1.77, 95% CI: 1.26~2.29) and exercise intervention (SMD=1.13, 95% CI: 0.82~1.44). Statistically significant moderators were identified intervention type by meta-ANOVA analyses. Finally, no significant evidence of publication bias was found. Conclusion: There is clear evidence that cognitive improvement programs can greatly enhance cognition in elderly with MCI. Future research should examine the effects of non-pharmacological interventions targeting elderly populations with mild-to-severe cognitive impairment in order to develop and enhance the effectiveness of cognitive improvement programs in Korea.

• Journal of the Korea Society of Computer and Information
• /
• 제23권3호
• /
• pp.71-77
• /
• 2018

The purpose of this study was to investigate the effectiveness of elastic band resistance training for muscle strength among community-dwelling older adults. The systematic review and meta-analysis was conducted by following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA). Data were pooled using fixed effect models. Sit to stand, arm curl, and grip strength were analyzed for main effects. Heterogeneity between studies was assessed using the I2 statistics and publication bias was evaluated by funnel plots. Twelves studies were included representing 611 participants. Elastic band resistance training was effective for lower (d=3.89, 95% CI: 3.03, 4.75) and upper extremity muscle strength (d=4.08, 95% CI: 2.94, 5.23). Heterogeneity was moderate and no significant publication bias was detected. Based on these findings, there is clear evidence that elastic band resistance training has significant positive effects on muscle strength among community-dwelling older adults. Further study will be needed to perform subgroup analysis using number of sessions and exercise intensity as predictors.