These data demonstrate significant differences in ICSI utilization between individual ART-mandated states as well as compared to non-mandated state aggregate data. Covered infertility benefits in ART-mandated states vary dramatically (Table 2). Multiple factors may influence the decision of whether to use ICSI or conventional IVF. Clinic-specific policies, discretion of treating physicians, and/or embryology laboratory personnel preference to use ICSI in patients with low oocyte yields, perceived poor oocyte quality or unexplained infertility are factors that can influence the use of ICSI.
The two ART-mandated states with the highest utilization of ICSI were HI and IL. Remarkably, a large portion of the ART clinics in Hawaii (80%) and in Illinois (48%) reported >75% ICSI usage, in contrast to no ART clinics in MA. HI reported dramatically higher male factor infertility rates (65.8%) than any other ART-mandated and non-mandated state. Furthermore, CT and MA showed a similar trend of lower ICSI use associated with more favorable clinical outcomes than reported in HI and IL. However, significantly greater male factor rates in HI than in other ART-mandated states and non-mandated states in aggregate may justify greater utilization of ICSI. This observation raises speculation of whether increased concern for potential poor, failed fertilization, unexplained infertility, or the perception of increased competition in specific geographic locations may have contributed to greater utilization of ICSI.
Interestingly, further analysis of the states with mandated ART coverage identified two states (IL and MA) with similar ART mandate structure and demographics but dramatically different ICSI utilization profiles. Both state mandates have a similar timeline for infertility diagnosis, covered cost of diagnostic tests, laboratory procedures that include ICSI, plan-dependent cost of medication and embryo cryopreservation (Table 2). Both states have heterogeneous, racially diverse populations within major metropolitan centers (Chicago and Boston) that are the homes of multiple academic teaching medical centers as well as many IVF private practices. However, IL had remarkably higher ICSI rates compared with ICSI rates of MA. Yet, in contrast, male factor diagnosis rates did not differ significantly between these two states, nor did PGT or singleton live birth rates. This wide difference in ICSI rates between IL and MA may in part be due to a fewer number of clinics and thus a greater annual ART cycle volume per IVF clinic in MA than in IL. It is speculated that lower and more selective use of ICSI utilization in MA could contribute to greater uniformity and less variability in employing ICSI by MA clinics compared to those clinics in IL.
Furthermore, several unique factors may explain why HI demonstrated the highest rate of ICSI among the ART-mandated states. Such factors contributing to increased ICSI utilization may include the limited number of ART clinics with significantly lower annual clinical volume, island-specific demographic distribution and a limited number of laboratory directors. Thus, we believe trends of ICSI use in HI may not be as representative nor as generalizable to ICSI and outcome rates of states in the continental U.S.
ICSI use varied significantly among the ART-mandated states while demonstrating no differences in live birth rates. This analysis of individual ART-mandated states suggests that the prevalence of male factor and the presence of a state insurance mandate are not the only factors influencing ICSI use. It is suggested that other possible non-clinical factors, such as the number of ART clinics in a given geographic area, clinic-specific policies, and/or patient/physician preferences, may impact the rate of ICSI utilization in a given state and will require further examination.
There are several strengths and limitations to this study. The primary strength is the use of the CDC dataset that incorporates >98% of ART cycles performed in the U. S. The improvement in the reporting of the 2018 data set compared to previous years included outcomes specific for each age group. Live birth rates were reported per transfer and specific for fresh non-donor embryos resulting from ICSI use.
Limitations include the fact that states with single clinics (AR, RI) were excluded as the variance calculation was possible only for states with two or more clinics. Hence, we could not assess the impact of every state’s ART mandate. Provided by the CDC, male factor rates are “per clinic” and are not age group-specific nor do they include details regarding the specific types of male factor diagnosis. An additional limitation of this study is that semen parameters were not collected nor available from the CDC dataset to help better understand the origin of the greater rates of male factor in HI.