Page 53 - EJMO-9-2

Eurasian Journal of
Medicine and Oncology TCM regulates gut microbiota in pediatric MP pneumonia

inflammatory factor IL-6 levels, and changes in intestinal Ultimately, 880 cases were included in the statistical
microbiota counts. analysis, comprising 455 in the experimental group
and 457 in the control group. No statistically significant
2.4. Risk of bias (ROB) in individual studies differences were observed when comparing the baseline

The Cochrane Collaboration’s ROB 2.0 tool was employed characteristics of both groups (Table 1).
to evaluate the ROB in the literature that we reviewed. The 3.3. Results of literature quality evaluation
evaluation covered six domains: (i) randomization bias;
(ii) bias due to deviations from intended interventions; The risk assessment of the included literature is depicted
(iii) outcome measurement bias; (iv) bias caused by missing in Figure 2. Regarding bias in the randomization process,
outcome data; (v) bias in the reporting of selected results; all 11 studies employed the random number table method
and (vi) overall bias. Based on these criteria, studies were for random assignment and were rated as having a low
classified as having a low, moderate, or high ROB. 17 ROB. However, none of the studies reported a concealed
allocation sequence and were therefore being rated as having
2.5. Statistical analysis an unknown risk. Seven studies 18,21-25,28 had a single author,
The statistical software Stata 12 was used to analyze the suggesting a higher likelihood that the trial intervention
included studies. Given the study aimed to assess the was administered by the same individual who evaluated
effectiveness of TCM in treating MPP in pediatric patients, the outcomes. This could potentially introduce bias due
a random-effects model was applied when the study to the evaluator’s awareness of the trial intervention, and
count was adequate. A fixed-effects model was used only these studies were rated as having a high ROB. In terms
18,19,21-28
when the sample size was small (k < 5). Heterogeneity of missing data, ten studies reported no cases of
was assessed using the I test, with I values interpreted as withdrawal, and the data were complete. The study by Li
2
2
follows: <25% indicating low heterogeneity, 25% ≤ I < 50% and Bai20 reported 10 dropouts in the experimental group
2
indicating moderate heterogeneity, and ≥50% indicating and eight in the control group.
high heterogeneity. Sensitivity analysis was carried out 3.4. Meta-analysis
to evaluate the stability of results in the presence of
heterogeneity. The risk ratio (RR) was used as the effect 3.4.1. Clinical effectiveness
size measure for categorical outcomes, and the weighted A total of 11 RCTs, comprising 880 patients, were included
mean difference (MD) was used for continuous outcomes, in the analysis of overall clinical effectiveness.18-28 The
both accompanied by 95% confidence intervals (CIs). results of the heterogeneity test indicated significant
Differences were considered statistically significant at heterogeneity (χ = 27.93, p = 0.002, I = 64.2%; Figure 3A).
2
2
p < 0.05. In addition, when the number of studies exceeded Consequently, a random-effects model was used for the
10, a funnel plot was generated to evaluate publication bias, meta-analysis. The analysis revealed a pooled RR of 1.19,
where symmetry within the funnel plot indicated no bias. 95% CI (1.13, 1.26), p < 0.001, indicating that the cure rate
Egger’s test was also performed for a quantitative analysis for children with MPP treated with TCM was 19% higher
of publication bias. than that of the control group, and this difference was
statistically significant.
3. Results
3.4.2. Quantitative changes in Bifidobacterium counts
3.1. Database search
Regarding the change in Bifidobacteria count, four studies
A total of 99 papers were initially screened, and 44 were were included in this analysis.20,25,26,28 The heterogeneity
identified as duplicates. Thirty-seven papers were excluded test showed no significant heterogeneity (p = 0.119,
for non-compliance with the inclusion criteria: non- I = 48.7%), so a fixed-effects model was applied (Figure 3B).
2
RCT studies, animal experimental studies, theoretical The analysis revealed a significant difference in the
mechanism of action studies, mismatched interventions, change in Bifidobacterium counts between the two groups
outcome indicators, and inclusion/exclusion criteria. (MD = 1.404, 95% CI: 1.180, 1.628, Z = 12.30, p < 0.05),
Ultimately, 11 papers 18-28 were selected for comprehensive with the TCM group showing a significant increase in
analysis. The literature screening process is depicted in Bifidobacteria count compared to the control group.
Figure 1.
3.4.3. Serum inflammatory factor IL-6 levels
3.2. Study characteristics
Regarding serum IL-6 levels, four studies18,21-23 reported
A total of 930 patients were initially considered, with on this parameter. The analysis showed significant
465 in the treatment group and 465 in the control group. heterogeneity (p = 0.00, I = 93.3%), so a random-effects
2

Volume 9 Issue 2 (2025) 45 doi: 10.36922/ejmo.8602