Figure 45 shows the numbers of ART cycles performed, live-birth deliveries, and infants born using ART from 1996 through 2004. The number of ART cycles performed in the United States has almost doubled, from 64,681 cycles in 1996 to 127,977 in 2004. The number of live-birth deliveries in 2004 (36,760) was about two and a half times higher than in 1996 (14,507). The number of infants born who were conceived using ART also increased steadily between 1996 and 2004. In 2004, 49,458 infants were born, which was more than double the 20,840 born in 1996. Because in some cases more than one infant is born during a live-birth delivery (e.g., twins), the total number of infants born is greater than the number of live-birth deliveries.

Figure 46 presents live birth rates for the four primary types of ART cycles. Live birth rates are presented per transfer rather than per cycle because that is the only way to directly compare cycles using fresh embryos with those using frozen embryos.

From 1996 through 2004, the live birth rates for fresh–nondonor cycles increased 21%, from 28% in 1996 to 34% in 2004. Over the same time period, live birth rates increased 47% for frozen–nondonor cycles, 30% for fresh–donor cycles, and 66% for frozen–donor cycles.

Singleton live births are an important measure of success because they have a much lower risk than multiple-infant births for adverse infant health outcomes, including prematurity, low birth weight, disability, and death. Figure 47 presents singleton live birth rates for the four primary types of ART cycles. Singleton live birth rates are presented per transfer rather than per cycle because that is the only way to directly compare cycles using fresh embryos with those using frozen embryos.

From 1996 through 2004, the singleton live birth rates for fresh–nondonor cycles increased 34%, from 17% in 1996 to 23% in 2004. Over the same time period, singleton live birth rates increased 72% for frozen–nondonor cycles, 33% for fresh–donor cycles, and 47% for frozen– donor cycles.

Figure 48 presents live birth rates per transfer, by woman’s age, for ART cycles using fresh nondonor eggs or embryos.

From 1996 through 2004, the live birth rates for women younger than 35 increased 27%, from 34% in 1996 to 43% in 2004. Over the same time period, live birth rates increased 22% for women 35–37, 17% for women 38–40, 29% for women 41–42, and 13% for women older than 42.

Singleton live births are an important measure of success because they have a much lower risk than multiple-infant births for adverse infant health outcomes, including prematurity, low birth weight, disability, and death. Figure 49 presents singleton live birth rates per transfer, by woman’s age, for ART cycles using fresh nondonor eggs or embryos.

From 1996 through 2004, the singleton live birth rate for women younger than 35 increased about 41%, from 19% in 1996 to 27% in 2004. Over the same time period, live birth rates increased 33% for women 35–37, 23% for women 38–40, 29% for women 41–42, and 22% for women older than 42.

Figure 50 presents the trends for number of embryos transferred in fresh–nondonor cycles that progressed to the embryo transfer stage. From 1996 to 2004, cycles that involved the transfer of one embryo increased slightly, from 6% to 8%; cycles that involved the transfer of two embryos increased dramatically, from 10% in 1996 to 39% in 2004. Cycles that involved the transfer of three embryos increased from 23% in 1996 to 32% in 2004, and cycles that involved the transfer of four or more embryos decreased from 62% in 1996 to 21% in 2004.

As shown in Figure 50, the number of embryos transferred in fresh–nondonor cycles has decreased during the past 9 years. Figure 51 shows the change over time in the number of embryos transferred for ART procedures in which the woman was younger than 35 and the couple chose to set aside some embryos for future cycles rather than transfer all available embryos at one time. Previous research suggests that the number of embryos available for an ART cycle is important in predicting success. Younger women also tend to have higher success rates (see Figure 31).

Overall, the number of embryos transferred decreased among couples who chose to transfer fewer embryos than were available. In 1996, almost two-thirds (64%) of ART cycles involved the transfer of four or more embryos; 33%, three embryos; and only 3%, two embryos. By 1998, the percentage of cycles in which four or more embryos were transferred had decreased to 33%; half of all ART cycles involved the transfer of three embryos, and 17% of cycles, two embryos. By 2004, four or more embryos were transferred in only 5% of cycles, three in 25% of cycles, and two in more than two-thirds (70%) of cycles.

Figure 52 presents live birth rates per transfer, by number of embryos transferred, for ART cycles using fresh nondonor eggs or embryos from 1996 through 2004. In general, the live birth rate was higher when two or more embryos were transferred. From 1996 through 2004, the live birth rate almost tripled, from 14% to 40%, for ART cycles that involved the transfer of two embryos. The live birth rates also increased for ART cycles that involved the transfer of either one or three embryos; however, live birth rates decreased 9%, from 32% to 29%, for ART cycles that involved the transfer of four or more embryos.

The relationship between number of embryos transferred and success rates is complicated by several factors, such as the woman’s age and embryo quality. Trends over time may reflect changes in these factors.

Figure 53 shows changes over time in the relationship between live birth rates and number of embryos transferred for ART procedures in which the woman was younger than 35 and the couple chose to set aside some embryos for future cycles rather than transfer all available embryos at one time. Previous research suggests that the number of embryos available for an ART cycle is an important predictor of success. Younger women also tend to have higher success rates (see Figure 31).

For this group, live birth rates increased over time for transfers involving any number of embryos. The increase in success rates was largest when two embryos were transferred. In 1996, the chance for a live birth using ART was highest (41%) when three embryos were transferred; however, in 2004, the chance for a live birth using ART was highest (53%) when two embryos were transferred.

Figure 53: Live Births per Transfer in Women Who Were Younger Than 35 and Set Aside Extra Embryos for Future Use, by Number of Embryos Transferred, 1996–2004.

Multiple-infant births are associated with greater problems for both mothers and infants, including higher rates of caesarean section, prematurity, low birth weight, and infant disability or death. Figure 54 shows the multiple-infant birth rates for the four primary types of ART procedures.

From 1996 through 2004, the percentage of live-birth deliveries that were multiple-infant births decreased 15%, from 38% in 1996 to 33%, for fresh–nondonor cycles. Over the same time period, the percentage of live-birth deliveries that were multiple-infant births decreased 10% for frozen–nondonor cycles and 4% for fresh–donor cycles. In all years except 1997, the multiple-infant birth rates remained stable for frozen–donor cycles.

Have multiple-infant birth rates for cycles using fresh nondonor eggs or embryos changed for all ART patients or only for those in particular age groups?

Figure 55 shows that the multiple-infant birth rates decreased slightly between 1996 and 2004 for women in all age groups. In 1996, 43% of live-birth deliveries to women younger than 35 were multiple-infant births, compared with 36% in 2004. Among women older than 42, the multiple-infant birth rates decreased from 14% in 1996 to 8% in 2004.

Have the percentages of singletons, twins, and triplets or more changed for ART cycles using fresh nondonor eggs or embryos?

Figure 56 presents the trends in live birth rates and percentage of multiple-infant births for ART cycles using fresh nondonor eggs or embryos. Overall, live birth rates per transfer increased from 28% in 1996 to 35% in 2003 and decreased slightly, to 34%, in 2004. From 1996 to 2004, the percentage of singleton live births increased from 62% to 67%; the percentage of twin births remained stable, ranging from 30% to 32%; and the percentage of triplet-or-more births decreased from 7% in 1996 to 3% in 2004.

It is important to note that twins, albeit to a lesser extent than triplets or more, are still at substantially greater risk for illness and death than singletons. These risks include low birth weight, preterm birth, and neurological impairments such as cerebral palsy. Both the twin and triplet-or-more birth rates remain significantly higher for ART births than for births resulting from natural conception.