Meta-analyses on specific disease patterns

Meta-analyses require a minimal number of included clinical trials to arrive at meaningful conclusions. For this reason, we restrict the following overview to meta-analyses which included a minimum of three clinical trials. Most Cochrane reviews were not able to identify more than two trials for many of the indications investigated. Meta-analyses investigating a subgroup of other published meta-analyses were excluded from the following overview, as well as global meta-analyses analysing specific indications only in sub-group analyses.

Gaertner et al. (2021) analysed 22 publications reporting on 28 different comparisons (18 comparisons to placebo, 9 to an active control, and in one case of no treatment). Meta-analysis of the placebo-controlled trials yielded an overall effect size of Hedge’s g = 0.18 (p = 0.059), and the analysis of active comparator trials resulted in an effect size of g = 0.26. This was mainly due to the nonrandomized studies, as the effect size in the randomized trials converged to zero. The authors concluded that “homeopathic Arnica has a small effect size over and against placebo in preventing excessive hematoma and other sequelae of surgeries. The effect is comparable to that of anti-inflammatory substances.”

Rotella et al. (2020) performed a systematic review and meta-analysis of randomized controlled trials in psychiatric disorders. Nine studies met all inclusion criteria. Identified studies concerned the following disorders: major depressive disorder (n = 4), generalized anxiety disorder (n = 1), attention-deficit / hyperactivity disorder (n = 2), and premenstrual syndrome / dysphoric disorder (n = 2). Meta-analysis performed on a minimum of three trials could be conducted on major depressive disorder and yielded a standardized mean difference of –0.51 (p=0.28). The authors concluded that “available data on homeopathy in psychiatric disorders are insufficient to support their use in clinical practice.”

Boehm et al. (2014) published a systematic review and meta-analysis of clinical trials on homeopathy for fibromyalgia. Meta-analyses of three RCTs revealed effects of homeopathy on tender point count (standardized mean difference (SMD) = −0.42; p = 0.03) and pain intensity (SMD = −0.54; p = 0.02), compared to placebo. The authors conclude that “the results of the studies (…) define a sufficient basis for discussing the possible benefits of homeopathy for patients suffering from fibromyalgia syndrome although any conclusions based on the results of this review have to be regarded as preliminary.”

Raak et al. (2012) published a systematic review and meta-analysis including four studies on combined application of potentized Hypericum perforatum and Arnica montana in dentistry. The authors state that the overall odds ratio of 0.24 favours homeopathic treatment, but just misses statistical significance (p=0.06). 

Jacobs et al. (2003) published a meta-analysis of three randomized controlled trials on individualized homeopathic treatment of childhood diarrhoea. The authors state that “combined analysis shows a duration of diarrhea of 3.3 days in the homeopathy group compared with 4.1 in the placebo group (p=0.008)” and conclude that “the results from these studies confirm that individualized homeopathic treatment decreases the duration of acute childhood diarrhea”.

Taylor et al. (2000) performed a meta-analysis of four randomized controlled trials on homeopathic immunotherapy for allergic complaints (hay fever, asthma, perennial rhinitis). The meta-analysis performed after four weeks of treatment showed a mean symptom reduction on a visual analogue scale of 28% (10.9 mm) for homoeopathy compared with 3% (1.1 mm) for placebo (p = 0.0007). The authors concluded that “the objective results reinforce earlier evidence that homoeopathic dilutions differ from placebo.”

Barnes et al. (1997) performed a systematic review and meta-analysis of controlled clinical trials investigating the effects of a homeopathic treatment of postoperative ileus. Meta-analysis yielded a statistically significant (p < 0.05) weighted mean difference (WMD) in favour of homeopathy (compared with placebo) on the time to first flatus. Exclusion of methodologically weak trials did not substantially alter the results. The authors concluded that “our analyses suggest that homeopathic treatment administered immediately after abdominal surgery may reduce the time to first flatus when compared with placebo administration”.

Lüdtke et al. (1997) performed a meta-analysis of seven placebo-controlled clinical trials on the effects of Galphimia glauca on patients with hay fever. The overall rate of improved eye-symptoms was about 1.25 times higher in the verum than in the placebo group. Across the single studies the results were highly comparable except for one study run in 1985. The authors concluded that “a significant superiority of Galphimia glauca over placebo is demonstrated. Estimates of verum success rates are comparable with those of conventional antihistamins, but no side effects occurred.” Since not all single studies were analyzed with an intention to treat analysis, the results might be biased to a certain extent. A worst-case control calculation still yielded an overall rate of improved eye-symptoms being 1.23 times higher in the verum than in the placebo group.

Summarizing, 5 out of the 8 meta-analyses on specific disease patterns with a minimum of three included trials yielded statistically significant effects of the investigated homeopathic interventions against placebo (p<0.05), two yielded a trend (p=0.06) and one did not give significant results (p=0.26). In some meta-analyses it was stated that additional trials with higher methodological quality would be needed to draw more robust conclusions. Based on the available evidence, however, it seems that homeopathic preparations induce specific remedy effects superior to placebo when applied in a clinically meaningful way.