Effects of interventions
We calculated Hedges’s g value and 95% confidence interval (CI) for all outcomes. The overall effects were weighted by the inverse of the variance on the effects of steroids on VN in five studies. As a result, Hedges’s g was 0.172 (95% CI 0.048 to 0.295, p = 0.006), which indicates that the steroid treatment had a small but significant therapeutic effect on VN. The heterogeneity for the overall effect size was moderate (I2 = 59.61) (Figure 3).
In this study, the effects of the intervention on the study group and the control groups were evaluated in the short-term (1 month), mid-term (3 & 6 months), and long-term (12 months) depending on the follow-up period. Hedges’s g was 0.106 (95% CI -0.142 to 0.353, p = 0.403, I2 = 40.97) and -0.059 (95% CI -0.252 to 0.135, p=0.552, I2 = 38.49) in the short-term and mid-term. There was no statistically significant difference in the effect of steroids in these two groups. However, Hedges’s g was 0.496 (95% CI 0.285 to 0.708, p < 0.001, I2 = 69.76) in the long-term. We found that the steroid treatment had a moderate statistically significant therapeutic effect on long-term follow-up of VN (Figure 4(A)).
To evaluate the possible effect of treatments other than steroids, we divided the comparison design as follows: 1) ST versus NST and 2) combined treatment of ST and others versus other treatments alone. When comparing the outcomes of ST and NST, Hedges’s g was 0.299 (95% CI 0.107 to 0.490, p = 0.002, I2 = 69.01). This finding indicates that the steroid treatment had a statistically significant but small therapeutic effect on VN. When comparing the effect of combined treatment of ST and others versus other treatments alone, Hedges’s g was 0.082 (95% CI -0.080 to 0.243, p = 0.322, I2 = 42.79). This result reflects that there is no statistically significant difference in the therapeutic effect of steroid (Figure 4(B)).
To analyze the possible effect of the outcome variables on the therapeutic effect of steroids on VN, the comparison was made in terms of complete caloric recovery, improvement of CP, and DHI score. When the therapeutic effect was measured in terms of complete caloric recovery and CP improvement, Hedges’s g was 0.364 (95% CI 0.181 to 0.547, p < 0.001, I2 = 11.44) and 0.592 (95% CI 0.315 to 0.869, p < 0.001, I2 = 65.91), respectively. The results demonstrated that steroids had a significant (medium sized) therapeutic effect on VN, assessed by complete caloric recovery and CP improvement. In contrast, with regard to the DHI score, Hedges’s g was -0.323 (95% CI -0.533 to -0.113, p < 0.001, I2 = 0). This result suggests that steroids had a negative effect (of small effect size) on VN when assessed by DHI score (Figure 4(C)).