The heterogeneity statistic provides a measure of how dissimilar/inconsistent the individual study effects are in your pooled analysis. If the 95% confidence intervals of the individual study effects do not overlap, this is considered a dissimilarity (so I2 is in fact also a function of the precision of the point estimates). I2>60% could be regarded as an indicator of substantial heterogeneity.

While you can always mathematically perform pooling in a meta-analysis, the question remains whether the final outcome makes sense. While your pooled effect appears to be significantly different from 0, I2 = 100% seems to indicate that the individual effects differ a lot on a case-by-case basis due to unexplained causes. Generally, in case of substantial heterogeneity, we would downgrade the quality of pooled evidence as the studies exhibit large degrees of inconsistency. But I2 should not be interpreted as a stand-alone. One should examine the nature of inconsistency. See the following article for more details:

Article GRADE guidelines: 7. Rating the quality of evidence - Inconsistency

Regardless, it is important to explain the underlying reasons for these large variations, especially if you want to inform policy decisions based on your analysis. Subgroup analyses and meta-regression could be used to explain parts of the variation, which may drive down the value for I2. But be mindful that post-hoc meta-regressions are prone to p-value hacking, and therefore perceived less valuable than meta-regressions based on a priori hypotheses.

• Mean estimated concentration of radon in drinking water is 25.01 Bq/L.
• There is high heterogeneity (I2 = 100%), which means the effect sizes vary widely from one study to another.
• There is residual heterogeneity (I2 = 62%, p = 0.01), which means that there are still some unknown factors contributing to the variability in the effect sizes between the studies.
• The results suggest that there is a significant association between exposure to radon in drinking water and the risk of cancer.