Background
A controversial and unresolved question in reproductive medicine is the utility of preimplantation genetic testing for aneuploidy as an adjunct to in vitro fertilization. Infertility is prevalent, but its treatment is notoriously expensive and typically not covered by insurance. Therefore, cost-effectiveness is critical to consider in this context.
Objective
This study aimed to analyze the cost-effectiveness of preimplantation genetic testing for aneuploidy for the treatment of infertility in the United States.
Study Design
As reported to the Society for Assisted Reproductive Technology Clinic Outcomes Reporting System, a national data registry, in vitro fertilization cycles occurring between 2014 and 2016 in the United States were analyzed. A probabilistic decision tree was developed using empirical outputs to simulate the events and outcomes associated with in vitro fertilization with and without preimplantation genetic testing for aneuploidy. The treatment strategies were (1) in vitro fertilization with intended preimplantation genetic testing for aneuploidy and (2) in vitro fertilization with transfers of untested embryos. Patients progressed through the treatment model until they achieved a live birth or 12 months after ovarian stimulation. Clinical costs related to both treatment strategies were extracted from the literature and considered from both the patient and payer perspectives. Outcome metrics included incremental cost (measured in 2018 US dollars), live birth outcomes, incremental cost-effectiveness ratio, and incremental cost per live birth between treatment strategies.
Results
The study population included 114,157 first fresh in vitro fertilization stimulations and 44,508 linked frozen embryo transfer cycles. Of the fresh stimulations, 16.2% intended preimplantation genetic testing for aneuploidy and 83.8% did not. In patients younger than 35 years old, preimplantation genetic testing for aneuploidy was associated with worse clinical outcomes and higher costs. At age 35 years and older, preimplantation genetic testing for aneuploidy led to more cumulative births but was associated with higher costs from both perspectives. From a patient perspective, the incremental cost per live birth favored the no preimplantation genetic testing for aneuploidy strategy from the <35 years age group to the 38 years age group and beginning at age 39 years favored preimplantation genetic testing for aneuploidy. From a payer perspective, the incremental cost per live birth favored preimplantation genetic testing for aneuploidy regardless of patient age.
Conclusion
The cost-effectiveness of preimplantation genetic testing for aneuploidy is dependent on patient age and perspective. From an economic perspective, routine preimplantation genetic testing for aneuploidy should not be universally adopted; however, it may be cost-effective in certain scenarios.
Why was this study conducted?
Is preimplantation genetic testing for aneuploidy (PGT-A) a cost-effective adjunct to in vitro fertilization (IVF) for the treatment of infertility in the United States?
Key findings
In this economic evaluation of 153,665 national IVF cycles, the cost-effectiveness of PGT-A largely depends on patient age and payer perspective and should not be universally adopted. In patients younger than 35 years old, PGT-A led to higher costs without better outcomes. From a societal perspective, PGT-A can lower the incremental costs per live birth by reducing multifetal gestations.
What does this add to what is known?
The findings from this study further underscore that financial incentives between infertile patients and payers are often misaligned.
Introduction
The utilization of preimplantation genetic testing for aneuploidy (PGT-A) remains one of the most controversial and intensely contested areas in reproductive medicine. Previously, morphology-based grading has dictated the selection of embryos for transfer into the uterus in in vitro fertilization (IVF). However, trophectoderm biopsy of blastocyst-stage embryos combined with PGT-A has gained popularity, as studies have demonstrated that PGT-A may decrease miscarriage rates, increase implantation rates, , and shorten treatment time to pregnancy. However, the value of universal adoption of PGT-A for all IVF patients is controversial. A Practice Committee of the American Society for Reproductive Medicine’s review of PGT-A in 2018 identified only 3 randomized controlled trials. All trials had relatively small sample sizes, enrolled patients with favorable prognoses, randomized at blastocyst (rather than at cycle start), and occurred at high-volume referral centers with expertise in trophectoderm biopsy, thus limiting the generalizability of their results. , , Subsequently, an additional multicentered, large, randomized controlled trial demonstrated a higher ongoing pregnancy rate per transfer in older patients (ages 35–40) randomized to PGT-A; however, there was no difference seen in women aged <35 years, and no significant benefit was seen in intention-to-treat analysis.
An increasingly important issue is the cost-effectiveness of PGT-A. Of note, 1 study found that when applied to patients with recurrent pregnancy loss, IVF with PGT-A was 100-fold more expensive than spontaneously conceived pregnancies. Furthermore, 2 previous cost-effectiveness studies found that PGT-A is cost-effective in women over 37 years old with infertility when at least 1 blastocyst is available to biopsy. , This is problematic as the quality and quantity of the subsequent embryos are not known at the start of treatment. A fourth study found that PGT-A is cost-effective in patients with at least 1 embryo available, although the magnitude of cost savings varied by age, and in a subset of young patients with supernumerary embryos, PGT-A was more costly than IVF alone.
No previous cost-effectiveness study has utilized national data or assessed costs from the start of IVF treatment, which is arguably the best time for patients to make decisions. In addition, previous studies have focused on the costs related to the IVF cycle and PGT-A testing platform alone , and did not incorporate societal costs associated with multifetal gestations. This study aimed to use actual national data as reported to the Society for Assisted Reproductive Technology Clinic Outcome Reporting System (SART CORS) to determine the cost-effectiveness of PGT-A at cycle start for treating infertility in the United States from the perspectives of the patient and payer.
Materials and Methods
Data collection from the Society for Assisted Reproductive Technology Clinic Outcome Reporting System
The study population included first fresh IVF cycles in women over 18 years of age occurring between January 2014 and December 2015, which were reported to the SART CORS, and any linked frozen embryo transfer (FET) cycles that occurred within 12 months of ovarian stimulation between January 2014 and December 2016. The SART CORS contains data from over 90% of IVF clinics in the United States. Data were collected and verified by the SART.
Cycles involving donor oocytes, gamete or zygote intrafallopian transfer, oocyte or embryo banking for fertility preservation, previously frozen oocytes, preimplantation genetic testing for monogenic disorders, or structural rearrangements were excluded.
Cycles were categorized by the intent to perform PGT-A. Within the group intending PGT-A, only cycles that led to a freeze-all strategy during the stimulation were included (any fresh transfer during an intended PGT-A cycle was excluded, as this is not standard of care and these cycles could not be clearly classified as 1 strategy). Cycles were analyzed by age group (<35 years, every year between 35 and 42 years, and >42 years).
Building the model
A probabilistic decision tree was created using the TreeAge Pro Healthcare 2020 R1.2 (TreeAge Software, Williamstown, MA). The model involved 2 distinct strategies that a patient could choose for her first IVF cycle: IVF with intended PGT-A and IVF alone with transfers of untested embryos ( Figure 1 ). For this model, specific treatment stages and nodes that were critical and would incur costs were chosen. Patients progressed through the model until they achieved a live birth, completed their embryo transfer cycles, or 12 months after stimulation. Each arm had a specific cost and age-dependent probability of progression.
For the IVF with PGT-A arm, patients progressed through a stimulated cycle (egg retrieval, insemination with conventional IVF, or intracytoplasmic sperm injection [ICSI]) and had biopsies of qualifying embryos. From there, no patient had a fresh embryo transfer in the PGT-A arm. Each patient underwent 0, 1, 2, 3, 4, 5, or 6 subsequent FET cycles. For the no PGT-A arm, patients progressed through a stimulated cycle. From that point, patients either had a fresh embryo transfer, FET, or no embryo transfer. Patients with a fresh embryo transfer who consequently were not pregnant or had a miscarriage could then, if applicable, progress with an additional 0 to 6 subsequent FET cycles.
For both arms, each transfer had 5 distinct possible outcomes: not pregnant (including biochemical pregnancies), clinical miscarriage, singleton live birth, twin live birth, or higher-order multiple live birth. Any pregnancy without a live birth, such as a pregnancy termination or an ectopic pregnancy, was included as a clinical miscarriage ( Figure 1 ).
The study was approved by the Partners Healthcare Institutional Review Board and the SART Research Committee.
Clinical costs
Clinical costs related to both treatment arms were estimated and extracted from the literature where possible and converted to 2018 US dollars ($) using the overall Consumer Price Index. Certain treatment stages incurred costs that were age dependent and empirically calculated from the SART CORS data between treatment arms. For instance, the number of gonadotropins used for ovarian stimulation was age dependent, ICSI was utilized at a higher rate than conventional insemination in PGT-A cycles vs no PGT-A cycles, and the rates and subsequent costs of miscarriage were age and treatment dependent and adjusted accordingly. Cost data are outlined in Supplemental Table 3 .
Costs were considered from 2 perspectives—the patient and payer. From the patient’s perspective, patients were responsible for all IVF-related costs, including medications, IVF cycle, and PGT-A–related procedures where applicable. The payer perspective incorporated the aforementioned IVF costs and obstetrical costs related to prenatal care, miscarriage management, and birth.
Statistical analysis
Means and standard deviation were generated for continuous variables and frequencies and proportions for categorical variables. Statistical significance was assessed using Wilcoxon rank-sum tests for continuous variables and chi-square tests for categorical variables. A significance level was set at an α of 0.01.
Sensitivity analyses
Discrete 1-way sensitivity analyses were performed to estimate how potential modifications to specific cost parameters changed the overall expected costs. As much of the IVF cycle costs are similar between treatment strategies (eg, the costs of gonadotropins and ovarian stimulation or monitoring), the sensitivity analyses reflected strategies that providers offering PGT-A could adjust to impact affordability.
Additional model scenarios accounted for variable practices in miscarriage management. As the SART does not collect data surrounding miscarriage management (expectant, medical, or surgical with a dilation and curettage), 2 model scenarios were run on the basis of the maximal bounds of miscarriage management cost (expectant, $0; cost of surgical management, $1383).
Outcomes
The primary outcome of the analysis was the incremental cost between the 2 treatment strategies from cycle start to completion of treatment. Secondary cost-effectiveness outcomes included the incremental cost-effectiveness ratio (ICER), defined as the difference in cost divided by the difference in live birth rates (LBRs) between the treatment strategies and the incremental cost per live birth.
Results
Society for Assisted Reproductive Technology outcomes
The study population included 114,157 first fresh IVF stimulations and 44,508 linked FET cycles. Supplemental Table 1 shows the demographic characteristics of patients undergoing fresh cycle starts. Of fresh stimulations, 18,449 (16.2%) intended PGT-A and 95,708 (83.8%) did not, consistent with previous studies. Patients intending PGT-A had higher gravidity and parity and significantly higher measures of ovarian reserve and lower body mass index (BMI) (all P values of <.001). A total of 2856 stimulated cycle starts were excluded from our study population as they included a fresh transfer of an embryo with biopsy of some or all remaining embryos.
Supplemental Table 2 shows cycle characteristics of the fresh cycles. Overall, cycles of patients undergoing PGT-A were less likely to result in cancelation before oocyte retrieval (1.9% vs 9.9%; P <.001). PGT-A cycles resulted in more oocytes retrieved (15.3 vs 13.4; P <.001) and more pronuclear embryos created (9.5 vs 7.9; P <.001). These trends persisted overall and across each age cohort. Of those cycles intending PGT-A with at least 1 pronuclear (2PN) embryo created, 6.4% had no biopsy performed. The average number of biopsied embryos was 4.8 and ranged from 5.8 in the <35 age group to 2.7 in the over 42 age group.
Table 1 shows the characteristics of transfers performed in 2014–2016, resulting from stimulated cycles initiated in 2014–2015. The total number of transfers (either fresh or frozen) was similar across age groups between the PGT-A and no PGT-A cohorts, although there was a statistically significant difference in the overall group only (1.3 in the no PGT-A group vs 1.2; P <.001). PGT-A cycles resulted in fewer embryos being transferred per transfer (1.2 vs 1.7; P <.001) and a higher rate of single embryo transfer (81.7% vs 38.5%; P <.001).
| Characteristic | Patient age at time of stimulation start (y) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| <35 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | >42 | Total | |
| No transfer, n (% per cycle start) | |||||||||||
| No PGT-A | 6486 (13.1) a | 1297 (17.9) a | 1313 (19.9) a | 1361 (23.6) a | 1377 (25.0) a | 1517 (29.1) a | 1456 (32.4) a | 1442 (38.2) a | 1196 (40.5) a | 2327 (51.0) a | 19,772 (20.7) a |
| PGT-A | 1543 (25.3) | 435 (30.4) | 452 (31.9) | 539 (35.1) | 582 (39.0) | 728 (45.4) | 836 (53.1) | 781 (61.8) | 682 (73.9) | 958 (86.5) | 7536 (40.9) |
| No transfer, n (% per cycles with at least 1 2PN embryo) | |||||||||||
| No PGT-A | 2598 (5.7) a | 496 (7.7) a | 536 (9.2) a | 519 (10.5) a | 546 (11.7) a | 602 (14.0) a | 577 (16.0) a | 588 (20.1) a | 513 (22.6) a | 1027 (31.5) a | 8002 (9.5) a |
| PGT-A | 1324 (21.7) | 395 (27.6) | 407 (28.7) | 479 (31.2) | 520 (34.9) | 662 (41.3) | 776 (49.2) | 733 (58.0) | 637 (69.0) | 877 (79.2) | 6810 (36.9) |
| No transfer, n (% per cycles with biopsy performed) | 1032 (18.5) | 323 (24.5) | 317 (24.8) | 400 (28.6) | 445 (32.8) | 566 (39.2) | 672 (47.6) | 632 (56.7) | 547 (69.4) | 745 (83.2) | 5679 (34.2) |
| Patients proceeding to transfer, n (%) b | |||||||||||
| No PGT-A | 43,086 (86.9) a | 5946 (82.1) a | 5303 (80.2) a | 4419 (76.5) a | 4124 (75.0) a | 3690 (70.9) a | 3037 (67.6) a | 2335 (61.8) a | 1757 (59.5) a | 2239 (49.0) a | 75,936 (79.3) a |
| PGT-A | 4745 (74.7) | 1028 (69.6) | 1002 (68.1) | 1027 (64.9) | 942 (61.0) | 900 (54.6) | 759 (47.0) | 500 (38.2) | 246 (26.1) | 153 (13.5) | 11,302 (59.2) |
| Total number of transfers, mean (SD) c | |||||||||||
| No PGT-A | 1.3 (0.6) | 1.3 (0.6) | 1.3 (0.6) | 1.3 (0.5) | 1.2 (0.5) | 1.2 (0.5) | 1.2 (0.4) | 1.1 (0.4) | 1.1 (0.3) | 1.1 (0.3) | 1.3 (0.6) a |
| PGT-A | 1.3 (0.6) | 1.3 (0.5) | 1.3 (0.5) | 1.2 (0.5) | 1.2 (0.5) | 1.2 (0.4) | 1.1 (0.4) | 1.1 (0.3) | 1.1 (0.3) | 1.1 (0.3) | 1.2 (0.5) |
| Embryos transferred, (mean, SD) | |||||||||||
| No PGT-A | 1.6 (0.5) a | 1.7 (0.6) a | 1.7 (0.6) a | 1.8 (0.6) a | 1.9 (0.7) a | 2 (0.7) a | 2.2 (0.8) a | 2.4 (1) a | 2.5 (1.1) a | 2.5 (1.3) a | 1.7 (0.7) a |
| PGT-A | 1.2 (0.4) | 1.2 (0.4) | 1.2 (0.4) | 1.2 (0.4) | 1.2 (0.4) | 1.1 (0.4) | 1.1 (0.3) | 1.1 (0.3) | 1.1 (0.3) | 1.1 (0.4) | 1.2 (0.4) |
| Single embryo transfer (SET), n (%) d | |||||||||||
| No PGT-A | 25,325 (45.1) a | 2928 (38.5) a | 2324 (34.6) a | 1723 (31.2) a | 1250 (24.7) a | 1062 (24.1) a | 732 (20.8) a | 525 (20.1) a | 372 (19.5) a | 643 (27.2) a | 36,884 (38.5) a |
| PGT-A | 4609 (78.8) | 1000 (79.6) | 996 (82.3) | 989 (81.8) | 905 (82.1) | 868 (85.7) | 749 (89.6) | 485 (89.8) | 235 (89.4) | 142 (88.8) | 10,978 (81.7) |
| Multiple embryo transfer, n (%) d | |||||||||||
| No PGT-A | 30,813 (54.9) a | 4676 (61.5) a | 4398 (65.4) a | 3802 (68.8) a | 3805 (75.3) a | 3348 (75.9) a | 2784 (79.2) a | 2082 (79.9) a | 1537 (80.5) a | 1721 (72.8) a | 58,966 (61.5) a |
| PGT-A | 1238 (21.2) | 256 (20.4) | 214 (17.7) | 220 (18.2) | 197 (17.9) | 145 (14.3) | 87 (10.4) | 55 (10.2) | 28 (10.6) | 18 (11.3) | 2458 (18.3) |
| Stage of embryos transferred, n (%) e | |||||||||||
| Cleavage | 12,857 (22.9) | 2284 (30.0) | 2217 (33.0) | 1957 (35.4) | 1995 (39.5) | 1868 (42.3) | 1740 (49.5) | 1443 (55.4) | 1134 (59.4) | 1597 (67.6) | 29,092 (30.3) |
| Blastocyst | 43,299 (77.1) | 5322 (70.0) | 4509 (67.0) | 3569 (64.6) | 3060 (60.5) | 2543 (57.7) | 1776 (50.5) | 1164 (44.6) | 775 (40.6) | 767 (32.4) | 66,784 (69.7) |
c Of all patients who had at least 1 transfer—includes fresh and frozen
Cycle outcomes are shown in Table 2 . The cumulative LBRs per initiated stimulated cycle were equivalent in the <35- and 35-year-old cohorts (both having a P value of >.01) and started to diverge at age 36, with significantly higher cumulative LBRs in the PGT-A cohort throughout all older age groups. The LBR per embryo transfer was significantly higher in the PGT-A cohort across all ages. Similarly, PGT-A cycles resulted in a significantly higher singleton LBR and lower miscarriage rate across all ages than no PGT-A cycles.
| Characteristic | Patient age at time of stimulation start (y) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| <35 | 35 | 36 | 37 | 38 | 39 | 40 | 41 | 42 | >42 | Total | |
| Cumulative live births, n (%) a | |||||||||||
| No PGT-A | 27,623 (55.7) | 3288 (45.4) | 2800 (42.3) b | 2142 (37.1) c | 1727 (31.4) c | 1313 (25.2) c | 861 (19.2) c | 478 (12.7) c | 283 (9.6) c | 132 (2.9) c | 40,647 (42.5) c |
| PGT-A | 3296 (54.1) | 676 (47.2) | 664 (46.9) | 673 (43.8) | 594 (39.8) | 553 (34.5) | 456 (28.9) | 289 (22.9) | 127 (13.8) | 78 (7.0) | 7406 (40.1) |
| Live birth rate, (% per transfer) d | |||||||||||
| No PGT-A | 49.2 c | 43.2 c | 41.6 c | 38.8 c | 34.2 c | 29.8 c | 24.5 c | 18.3 c | 14.8 c | 5.6 c | 42.4 c |
| PGT-A | 56.4 | 53.7 | 54.8 | 55.6 | 53.9 | 54.5 | 54.5 | 53.5 | 48.3 | 48.8 | 55.1 |
| Singleton live births, n (%) a | |||||||||||
| No PGT-A | 21,131 (42.6) c | 2538 (35.0) c | 2210 (33.4) c | 1661 (28.7) c | 1357 (24.7) c | 1034 (19.9) c | 714 (15.9) c | 404 (10.7) c | 245 (8.3) c | 121 (2.7) c | 31,415 (32.8) c |
| PGT-A | 2877 (47.2) | 584 (40.8) | 588 (41.5) | 594 (38.7) | 532 (35.7) | 500 (31.2) | 427 (27.1) | 275 (21.8) | 115 (12.5) | 76 (6.9) | 6568 (35.6) |
| Twin live births, n (%) a | |||||||||||
| No PGT-A | 6342 (12.8) c | 740 (10.2) c | 577 (8.7) c | 465 (8.0) c | 352 (6.4) c | 269 (5.2) b | 141 (3.1) c | 70 (1.9) | 37 (1.3) | 10 (0.2) | 9003 (9.4) c |
| PGT-A | 411 (6.7) | 89 (6.2) | 75 (5.3) | 78 (5.1) | 61 (4.1) | 53 (3.3) | 28 (1.8) | 13 (1.0) | 12 (1.3) | 2 (0.2) | 822 (4.5) |
| Higher-order multiple live births, n (%) a | |||||||||||
| No PGT-A | 150 (0.3) | 10 (0.1) | 13 (0.2) | 16 (0.3) | 18 (0.3) | 10 (0.2) | 6 (0.1) | 4 (0.1) | 1 (0.0) | 1 (0.0) | 229 (0.2) |
| PGT-A | 8 (0.1) | 3 (0.2) | 1 (0.1) | 1 (0.1) | 1 (0.1) | 0 (0.00) | 1 (0.1) | 1 (0.1) | 0 (0.0) | 0 (0.0) | 16 (0.1) c |
| Miscarriage, n (%) | |||||||||||
| No PGT-A a | 4062 (8.2) c | 660 (9.1) c | 637 (9.6) c | 536 (9.3) c | 555 (10.1) c | 472 (9.1) c | 407 (9.1) c | 309 (8.2) c | 229 (7.8) c | 174 (3.8) c | 8041 (8.4) c |
| PGT-A a | 309 (5.1) | 71 (5.0) | 60 (4.2) | 76 (5.0) | 73 (4.9) | 59 (3.7) | 56 (3.6) | 49 (3.9) | 21 (2.3) | 16 (1.4) | 790 (4.3) |
| No PGT-A d | 4062 (7.2) c | 660 (8.7) c | 637 (9.5) c | 536 (9.7) c | 555 (11.0) c | 472 (10.7) | 407 (11.6) c | 309 (11.9) | 229 (12.0) | 174 (7.4) | 8041 (8.4) c |
| PGT-A d | 309 (5.3) | 71 (5.6) | 60 (5.0) | 76 (6.3) | 73 (6.6) | 59 (5.8) | 56 (6.7) | 49 (9.1) | 21 (8.0) | 16 (10.0) | 790 (5.9) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
