What the latest evidence says about ten common IVF add-ons

Researchers combined the results of 85 randomized trials to evaluate ten of the most commonly used IVF add-ons and found that endometrial scratching improved the chance of having a baby, while most other add-ons showed little or no benefit or require more high-quality research.

IVF add-ons are tests, treatments, or procedures used alongside standard IVF in an attempt to improve the chance of pregnancy or having a baby. Examples include embryo glue, endometrial scratching, PGT-A, intralipids, and acupuncture. While many add-ons are widely offered, it’s not clear how well they work.

Lensen et al. (2026) performed a meta-analysis of 85 randomized trials evaluating ten of the most commonly used IVF add-ons. Unlike many previous meta-analyses, they excluded studies with issues that could make the results less reliable, like not registering the trial before it started.

---

Study details

• Study design: Systematic review and meta-analysis of 85 randomized controlled trials (RCTs) evaluating 10 common IVF add-ons: PGT-A, endometrial scratching, EmbryoGlue, ERA, corticosteroids, PICSI, acupuncture, intralipids, intrauterine PRP, and ovarian PRP.
• Primary outcome: Live birth

The authors carefully evaluated whether each study was trustworthy before including it. They found 157 potentially eligible randomized trials, but excluded 72 (46%) because they weren’t registered before the study began or had other concerns that could make their results less reliable, such as retractions or questionable study methods. An example of why this matters is highlighted in the acupuncture section below.

The authors also rated the quality of the evidence for each outcome, indicating how confident they were that the results reflect what really happens. You can read more about quality of evidence and why it matters in the associated glossary term.

Endometrial scratching showed a higher chance of live birth

Endometrial scratching involves making a small injury to the lining of the uterus before embryo transfer in an attempt to improve implantation. Exactly how it works isn’t clear, but it may trigger a temporary inflammatory response that recruits immune cells and other factors that promote implantation..

Endometrial scratching was the only add-on associated with a higher chance of live birth. This finding was based on 19 randomized trials involving more than 6,300 participants, making it one of the largest bodies of evidence included in the review.

• Live birth: 31.6% vs 28.8%
  • There was a 20% higher odds of live birth with endometrial scratching than without it (odds ratio [95% CI]: 1.20 [1.02–1.41], 19 studies, 6,319 participants, I2= 39%, quality of evidence: moderate).
  • An odds ratio is not the same as a percentage increase in the outcome itself. For example, if the live birth rate were 40.0% without endometrial scratching, an odds ratio of 1.20 corresponds to an estimated live birth rate of about 44.4% with endometrial scratching. Use the Odds Ratio (OR) calculator at the end of this post to help interpret this and other odds ratios reported in the study.
• Clinical pregnancy: 16% higher odds with endometrial scratching (OR 1.16 [1.01–1.32], 22 studies, 7223 participants, I2= 31%, quality of evidence: moderate)
• Miscarriage: No significant difference (OR 1.11 [0.90–1.37], 21 studies, 7123 participants, I2= 0%, quality of evidence: moderate)

EmbryoGlue improved clinical pregnancy but not live birth

EmbryoGlue was associated with a higher chance of clinical pregnancy, but not live birth.

• Live birth: No significant difference (OR 1.12 [0.91–1.37], 7 studies, 3775 participants, I2= 47%, quality of evidence: low)
• Clinical pregnancy: 25% higher odds with EmbryoGlue (OR 1.25 [1.02–1.53], 11 studies, 3989 participants, I2= 23%, quality of evidence: low)
• Miscarriage: No significant difference (OR 0.95 [0.73–1.23], 8 studies, 3205 participants, I2= 0%, quality of evidence: low)

PICSI may reduce miscarriage

PICSI is a sperm selection technique used in the embryology lab to help select mature sperm before ICSI.

PICSI didn’t improve live birth or clinical pregnancy, but it was associated with lower odds of miscarriage. Most of this evidence came from one large randomized trial.

• Live birth: No significant difference (OR 1.12 [0.98–1.29], 1 study, 2772 participants, quality of evidence: low)
• Clinical pregnancy: No significant difference (OR 0.97 [0.84–1.12], 2 studies, 3327 participants, I2= 0%, quality of evidence: moderate)
• Miscarriage: 42% lower odds with PICSI (OR 0.58 [0.42–0.80], 2 studies, 3327 participants, I2= 0%, quality of evidence: moderate)

Acupuncture for embryo transfer showed no clear benefit

Acupuncture did not improve live birth. Although there was a higher chance of clinical pregnancy, there was also a higher chance of miscarriage.

• Live birth: No significant difference (OR 1.04 [0.80–1.34], 10 studies, 3735 participants, I2= 63%, quality of evidence: very low)
• Clinical pregnancy: 35% higher odds (OR 1.35 [1.07–1.70], 21 studies, 4947 participants, I2= 65%, quality of evidence: very low)
• Miscarriage: 35% higher odds (OR 1.35 [1.04–1.76], 10 studies, 3803 participants, I2= 0%, quality of evidence: very low)

The authors found 91 potentially eligible randomized trials, but excluded 69 (76%) because they weren’t registered before the study began or had other concerns that could make their results less reliable. This included 13 small studies from the same research group that didn’t respond to questions about randomization and if the same participants were used across studies.

This is a good example of why study quality matters. One group may have reported the same participants across multiple studies, making acupuncture appear more effective than it really is. Even randomized trials can have serious problems that make their results less reliable. That’s why it’s important for meta-analyses to carefully evaluate which studies are trustworthy before combining their results.

The miscarriage effect is surprising, but the authors rated the quality of evidence as very low, so it’s not clear whether acupuncture really has these effects based on the research they included.

Corticosteroids showed no clear benefit

There was no convincing evidence that corticosteroids improved outcomes.

• Live birth: No significant difference (OR 0.95 [0.70–1.27], 2 studies, 814 participants, I2= 0%, quality of evidence: moderate)
• Clinical pregnancy: No significant difference (OR 1.11 [0.91–1.36], 11 studies, 1931 participants, I2= 0%, quality of evidence: moderate)
• Miscarriage: No significant difference (OR 1.24 [0.85–1.82], 3 studies, 1057 participants, I2= 0%, quality of evidence: low)

PRP showed no clear benefit

Platelet-rich plasma (PRP) uses a concentrated sample of a patient’s own blood platelets, which is injected into the uterus or ovaries in an attempt to improve IVF outcomes. When injected into the uterus, the goal is to improve embryo transfer outcomes. When injected into the ovaries, the goal is to improve the number and quality of eggs.

There was no convincing evidence that either type of PRP improved outcomes.

Intrauterine PRP

• Live birth: No significant difference (OR 1.46 [0.63–3.37], 2 studies, 138 participants, I2= 0%, quality of evidence: very low)
• Clinical pregnancy: No significant difference (OR 1.60 [0.57–4.47], 4 studies, 232 participants, I2= 50%, quality of evidence: very low)
• Miscarriage: No significant difference (OR 1.21 [0.31–4.76], 3 studies, 168 participants, I2= 0%, quality of evidence: very low)

Intraovarian PRP

• Live birth: No significant difference (OR 0.40 [0.13–1.21], 1 study, 60 participants, quality of evidence: very low)
• Clinical pregnancy: No significant difference (OR 0.53 [0.12–2.28], 2 studies, 143 participants, I2= 76%, quality of evidence: very low)
• Miscarriage: No significant difference (OR 0.17 [0.02–1.58], 1 study, 60 participants, quality of evidence: very low)

Intralipids may improve live birth, but the evidence is very uncertain

Intralipids are intravenous fat emulsions sometimes used during IVF to try to modify the immune system and improve implantation.

Only one small study that evaluated intralipids was deemed trustworthy enough for inclusion, making it difficult to draw firm conclusions.

• Live birth: 231% higher odds (OR 3.31 [1.24–8.80], 1 study, 105 participants, quality of evidence: very low)
• Clinical pregnancy: 231% higher odds (OR 3.31 [1.24–8.80], 1 study, 105 participants, quality of evidence: very low)
• Miscarriage: Not available

For both live birth and clinical pregnancy, the 95% confidence interval was very wide (1.24 to 8.80). This means the true effect could range from a modest improvement (OR 1.24, or 24% higher odds) to a very large one (OR 8.80, or 780% higher odds). Because the possible range is so large, it’s hard to know how much benefit intralipids actually provide. This uncertainty is mainly because the results are based on a single small study of just 105 participants. More high-quality studies are needed before firm conclusions can be drawn.

PGT-A reduced miscarriage but showed no benefit for live birth

Unlike the other add-ons in this review, PGT-A doesn’t improve embryo quality or the uterine environment. Instead, it helps select which embryo to transfer. Because of this, the first-transfer results below mainly tell us whether using PGT-A to select embryos leads to better outcomes than selecting embryos based on conventional embryo grading alone.

• Live birth: No significant difference (OR 1.16 [0.91–1.47], 5 studies, 2578 participants, I2= 42%, quality of evidence: moderate)
• Clinical pregnancy: No significant difference (OR 1.29 [0.82–2.02], 6 studies, 2690 participants, I2= 82%, quality of evidence: moderate)
• Miscarriage: 38% lower odds with PGT-A (OR 0.62 [0.40–0.97], 6 studies, 2690 participants, I2= 40%, quality of evidence: low)

These results suggest that, for the first embryo transfer, using PGT-A to select embryos didn’t increase the chance of live birth or clinical pregnancy compared with conventional embryo grading. However, it was associated with a lower chance of miscarriage.

These results are based on the first embryo transfer only. The cumulative results below look at outcomes across all embryo transfers from an IVF cycle, rather than only the first transfer. This is especially relevant for PGT-A because the cumulative results look at the overall chance of having a baby after all eligible embryos have been transferred.

• Cumulative live birth: No significant difference (OR 1.04 [0.65–1.68], 3 studies, 1,762 participants, I2= 71%, quality of evidence: very low)
• Cumulative clinical pregnancy: 43% lower odds (OR 0.57 [0.35–0.94], 2 studies, 1,662 participants, I2= 70%, quality of evidence: very low)
• Cumulative miscarriage: 43% lower odds (OR 0.57 [0.41–0.79], 2 studies, 1,662 participants, I2= 0%, quality of evidence: very low)

Based on the limited available data, PGT-A didn’t improve the overall chance of having a baby from an IVF cycle. Cumulative clinical pregnancy was lower, while cumulative miscarriage was also lower. One possible explanation is that PGT-A may exclude some embryos labeled as non-euploid that could still have led to a pregnancy. However, these results are based on only two or three studies, so more research is needed before firm conclusions can be drawn.

Conclusion

Overall, this review found that most commonly used IVF add-ons were not supported by strong evidence of improving live birth. Endometrial scratching was the only add-on associated with a higher live birth rate, while EmbryoGlue improved clinical pregnancy but not live birth. PICSI and PGT-A were associated with lower miscarriage rates.

Many of the other add-ons either showed no clear benefit or were supported by only a small number of studies. Even when results looked promising, like for intralipids, the evidence was too limited to draw firm conclusions.

This review also highlights an important point about medical research. Different meta-analyses can reach different conclusions depending on which studies they include. Rather than including every available trial, the authors used strict criteria to identify studies they considered trustworthy before combining the results. While this resulted in fewer studies being included, the goal was to base the overall conclusions on studies the authors considered the most trustworthy. Compared with a meta-analysis that includes all available studies, this approach is more conservative because it excludes studies with features that could make their results less reliable.

Overall, the quality of the evidence (ie. certainty of evidence) ranged from very low to moderate. This means that the authors considered that some results were well supported by the evidence, but many still require larger, high-quality randomized trials before we can be confident that they truly improve IVF outcomes.

Explore this study with a calculator

🔒 Subscriber resource: Enter different baseline rates to see what this study’s reported odds ratios mean. This tool is for educational purposes and should not be used to predict an individual’s outcome because many other factors can influence the result.

How it works: Choose one of the odds ratios reported in the study and enter a baseline rate to see how it changes.

• Example: This study found that endometrial scratching was associated with an odds ratio of 1.20 for live birth. If the baseline live birth rate is 40.0%, the calculator estimates an average live birth rate of 44.4% with endometrial scratching.

Statistically significant odds ratios reported in this study:

• Endometrial scratching
  • Live birth: OR 1.20
  • Clinical pregnancy: OR 1.16
• EmbryoGlue
  • Clinical pregnancy: OR 1.25
• PICSI
  • Miscarriage: OR 0.58
• Acupuncture
  • Clinical pregnancy: OR 1.35
  • Miscarriage: OR 1.35
• Intralipids
  • Live birth: OR 3.31
  • Clinical pregnancy: OR 3.31
• PGT-A (first transfer)
  • Miscarriage: OR 0.62
• PGT-A (cumulative)
  • Clinical pregnancy: OR 0.57
  • Miscarriage: OR 0.57

This study’s primary outcome was live birth. Other outcomes, such as clinical pregnancy and miscarriage, were secondary outcomes that provide additional information but are generally considered less definitive.

---

Want to read more about IVF add-ons?

Reference

Lensen S, Wilkinson J, Steeper M et al.Safety and effectiveness of ten common in-vitro fertilisation add-ons: a systematic review and meta-analysisThe Lancet Obstetrics, Gynaecology, & Women’s Health, 2026; 2, e624-e636

 

---

About Embryoman

Embryoman (Sean Lauber) is a former embryologist and the founder of Remembryo, an IVF research and fertility education website. After working in an IVF lab in the US, he returned to Canada and now focuses on making fertility research more accessible. He holds a Master’s in Immunology and launched Remembryo in 2018 to help patients and professionals make sense of IVF research. Sean shares weekly study updates on Facebook, Instagram, and Reddit regularly. He also answers questions on Reddit or in his private Facebook group.

---