Figure 43 shows the numbers of ART cycles performed, live-birth deliveries, and infants born using ART from 1996 through 2003. The number of ART cycles performed in the United States has almost doubled, from 64,681 cycles in 1996 to 122,872 in 2003. The number of live-birth deliveries in 2003 (35,785) 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 2003. In 2003, 48,756 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 44 presents live birth rates for the four primary types of ART procedures. 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. Trends in live birth rates were considered in two ways. First, we assessed whether there was a change in the live birth rate over the previous year (that is, we compared the 2003 live birth rates with the 2002 live birth rates). We also assessed the total change in live birth rates from 1996 (the first full year of data collection) through 2003.

Between 2002 and 2003, the live birth rates were comparable for all types of ART procedures. From 1996 through 2003, live birth rates increased 24% for fresh–nondonor cycles, 62% for frozen–nondonor cycles, 31% for fresh–donor cycles, and 45% for frozen–donor cycles.

Figure 44: Live Births per Transfer, by Type of ART Procedure, 1996, 2002, and 2003.

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 45 presents singleton live birth rates for the four primary types of ART procedures. 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. Trends in singleton live birth rates were considered in two ways. First, we assessed whether there was a change in the singleton live birth rate over the previous year (that is, we compared the 2003 singleton live birth rates with the 2002 singleton live birth rates). We also assessed the total change in singleton live birth rates from 1996 (the first full year of data collection) through 2003.

Between 2002 and 2003, the singleton live birth rates were comparable for all types of ART procedures. From 1996 through 2003, the singleton live birth rates increased 33% for fresh–nondonor cycles, 66% for frozen–nondonor cycles, 34% for fresh–donor cycles, and 46% for frozen–donor cycles.

Figure 46 presents live birth rates per transfer, by woman’s age, for ART cycles using fresh nondonor eggs or embryos. Trends in live birth rates were considered in two ways. First, we assessed whether there was a change in the live birth rate over the previous year (that is, we compared the 2003 live birth rates with the 2002 live birth rates). We also assessed the total change in live birth rates from 1996 (the first full year of data collection) through 2003.

Between 2002 and 2003, the live birth rates were comparable for women in all age groups. The increases in live birth rates from 1996 through 2003 were 29% for women younger than 35, 26% for women 35–37, 21% for women 38–40, 31% for women 41–42, and 9% 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 47 presents singleton live birth rates per transfer, by woman’s age, for ART cycles using fresh nondonor eggs or embryos. Trends in singleton live birth rates were considered in two ways. First, we assessed whether there was a change in the singleton live birth rate over the previous year (that is, we compared the 2003 singleton live birth rates with the 2002 singleton live birth rates). We also assessed the total change in singleton live birth rates from 1996 (the first full year of data collection) through 2003.

Between 2002 and 2003, the singleton live birth rates were comparable for women in all age groups. From 1996 through 2003, the singleton live birth rate for women younger than 35 increased about 40%, from about 19% in 1996 to almost 27% in 2003. Likewise, over the same time period, live birth rates increased 36% for women 35–37, 25% for women 38–40, 31% for women 41–42, and 15% for women older than 42.

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 48 shows multiple-infant birth rates for the four primary types of ART procedures. Trends in multiple-infant birth rates were considered in two ways. First, we assessed whether there was a change in these rates over the previous year (that is, we compared the 2003 rates with the 2002 rates). We also assessed the total change in multiple-infant birth rates from 1996 (the first full year of data collection) through 2003.

Between 2002 and 2003, the multiple-infant birth rates decreased 5% for both fresh–donor and frozen–donor cycles and about 3% for fresh–nondonor cycles. During the same time period, multiple-infant birth rates increased 3% for frozen–nondonor cycles. The multiple-infant birth rates from 1996 to 2003 decreased 11% for fresh–nondonor cycles, 6% for frozen–nondonor cycles, and 3% for fresh–donor cycles. The multiple-infant birth rates from 1996 through 2003 were similar for frozen–donor cycles.

Figure 49 compares twin and triplet-or-more birth rates for ART cycles using fresh nondonor eggs or embryos in 1996 (the first full year of data collection), 2002, and 2003. Twins made up the vast majority of multiple-infant births in each of these years. Since 1996, the triplet-or-more birth rate for fresh–nondonor cycles has decreased, but there has been no change in the twin birth rate.

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.