Reference Ranges for Left Modified Myocardial Performance Index: Systematic Review and Meta-Analysis

The modified myocardial performance index (mod-MPI) is a noninvasive Doppler-derived metric used to evaluate fetal cardiac function. However, the reference ranges for mod-MPI in normal fetuses are not clearly defined, which limits the use of this technology in fetuses with potential cardiac compromise. Thus, we aimed to perform a systematic review and meta-analysis of published mod-MPI reference ranges across gestation.

The published literature was systematically searched, and all published articles in any language that provided values for the left ventricular mod-MPI obtained in low-risk, singleton fetuses were considered eligible for further review. All retrieved titles and abstracts were independently reviewed by two researchers. Mean and standard deviation by gestational week was extracted or calculated from published data. DerSimonian-Laird random-effects models were used to estimate pooled means and 95% confidence intervals (CIs).

The search resulted in 618 unique citations, of which 583 did not meet inclusion criteria, leaving 35 abstracts selected for full-text review. Review of the references of these 35 articles identified another 5 studies of interest. Of the 40 articles reviewed, six met inclusion criteria. There was significant heterogeneity seen in the mod-MPI results reported. Mod-MPI increased as pregnancy progressed in all studies. The pooled mean mod-MPI at 11 weeks' gestation was 0.400 (95% CI 0.374-0.426) and increased to 0.585 (95% CI 0.533-0.637) at 41 weeks' gestation. The increase was linear in 5 of 6 studies, while in 1 study, the mod-MPI was stable until 27 weeks' gestation, and then increased throughout the third trimester. Despite all having trends increasing over pregnancy, there was no study in which all the weekly means fell within the pooled 95% CI.

While mod-MPI does increase over gestation, the true "reference ranges" for fetuses remain elusive. Future efforts to further optimize calculation of time intervals possibly via automation are desperately needed to allow for reproducibility of this potentially very useful tool to assess fetal cardiac function.