The short answer
It depends on what went wrong. ESHRE's 2025 stimulation guideline supports protocol changes mainly for lowering OHSS risk in women who over-respond 1. It finds little support for raising the dose in predicted low responders, adding a dual trigger, or freezing all embryos routinely 1,3,5. None of its recommendations targets embryo quality 1.
After a failed cycle, changing the protocol can feel like the obvious move: a different injection, a higher dose, a new trigger, a 'long' instead of a 'short'. Some of these changes have good reasons behind them in particular situations; others have little evidence behind them.
This page does not recommend a protocol or a dose; that depends on your history, tests and records. It explains the main options, what ESHRE's 2025 ovarian stimulation guideline says about each, and where the guideline finds that a change is not supported 1. It is organised around what happened in your last cycle, and ends with questions to take to your doctor.
Reading your report
- Antagonist protocol
- Stimulation starts with your period, and a GnRH antagonist injection is added a few days later to stop early ovulation. Often called a 'short protocol' in clinics.
- Long agonist protocol
- A GnRH agonist is started before stimulation, usually in the previous cycle, to switch off the pituitary first. Often called the 'long protocol'.
- Gonadotropins (FSH, hMG)
- The daily stimulation injections. Dose is usually written in international units (IU).
- Trigger
- The final injection that matures the eggs before collection: hCG, a GnRH agonist, or both together (a 'dual trigger').
- Freeze-all
- No fresh transfer: all suitable embryos are frozen and transferred in a later cycle.
- Low / normal / high responder
- ESHRE's terms for how strongly the ovaries respond to stimulation. It replaced 'poor responder' with 'low responder' to avoid suggesting a poor prognosis 1.
- Modified natural cycle
- IVF that relies mainly on the one follicle your body grows each month, with minimal medication.
- Double stimulation (DuoStim)
- Two rounds of stimulation and egg collection within one menstrual cycle, to collect more eggs when no fresh transfer is planned.
What did not cause this
"The protocol I chose made it fail"
Protocol choice is a medical decision, not yours. None of the evidence-based recommendations in ESHRE's 2025 guideline is supported by high-quality evidence 1, so specialists are often choosing between options without strong evidence either way.
"I should have pushed for a higher dose"
A 2024 Cochrane review found no clear justification for raising the standard dose in predicted poor or normal responders, and noted that higher doses add cost 3.
Does changing the protocol improve the next cycle?
Sometimes, but the evidence base is thin. ESHRE's 2025 guideline makes 121 recommendations, and none of its evidence-based recommendations is supported by high-quality evidence; most rest on low or very low-quality evidence 1. Its critical outcomes are live birth per started cycle and moderate or severe OHSS 1.
The guideline's own conclusion is that there are limited reasons to depart from conventional dosing 1. It also dropped the concept of 'mild stimulation' because the studies were too varied to compare 1. The useful question is therefore not 'what can we change?' but 'what went wrong, and is there evidence that a specific change fixes that?'
Agonist or antagonist: does switching help?
ESHRE recommends the antagonist protocol over agonist protocols for the general IVF population, because results are comparable and it is safer (strong recommendation, moderate-quality evidence) 1. A Cochrane review of 73 trials found no conclusive difference in live birth (odds ratio 1.02) and less OHSS with the antagonist: if the OHSS risk with an agonist protocol is 11%, the risk with an antagonist would be 6–9% 2.
The same review found that cycles were cancelled for poor response slightly more often with the antagonist protocol (odds ratio 1.32) 2. If an agonist protocol is used, ESHRE recommends the long version over the short or ultrashort ones (strong recommendation, low-quality evidence) 1. If all embryos will be frozen anyway, it says progestin tablets can be used to prevent early ovulation instead of GnRH drugs (conditional, very low-quality evidence) 1.
Condition-specific guidance points the same way. The PCOS guideline says the antagonist protocol cannot be recommended over the long agonist protocol for pregnancy rates, but recommends it because it allows an agonist trigger and freeze-all to reduce OHSS 6. See our PCOS and endometriosis pages.
Should the dose go up or down next time?
For predicted normal responders, ESHRE says neither a reduced nor an increased dose is probably recommended over a conventional dose (conditional, very low-quality evidence) 1. For predicted low responders, it does not recommend a gonadotropin dose higher than 300 IU (strong recommendation, very low-quality evidence) 1. For predicted high responders, it says a reduced dose is probably recommended to lower OHSS risk (conditional, very low-quality evidence) 1.
The 2024 Cochrane review on dose selection (26 studies, 8,520 women) did not find that tailoring the dose changed live birth in any group, though it could not rule out differences 3. It concluded that current evidence does not clearly justify raising the standard dose for predicted poor or normal responders, and that it is unclear whether a lower dose reduces OHSS in high responders 3.
Changing the dose partway through stimulation is probably not recommended, and ESHRE says this makes choosing the right starting dose for your characteristics more important 1.
What is a dual trigger, and does ESHRE support it?
A dual trigger adds a GnRH agonist to the usual hCG injection before egg collection. ESHRE says it is probably not recommended for predicted normal responders or for low responders (conditional, low-quality evidence) 1. For high responders, adding hCG to an agonist trigger is also probably not recommended (conditional, very low-quality evidence) 1.
An agonist trigger on its own is a different tool. ESHRE does not recommend it for the general IVF population when a fresh transfer is planned (strong, low-quality evidence), but recommends it for women at risk of OHSS, combined with freezing all embryos (strong, very low-quality evidence) 1.
Should all embryos be frozen next time?
Only for a reason. A 2021 Cochrane review of 8 trials (4,712 women) found probably little or no difference in cumulative live birth between freezing all embryos and a conventional fresh transfer: where the fresh approach gives 58%, freeze-all would give 57–63% 4. Freeze-all reduced OHSS but may raise the risk of high blood pressure in pregnancy and of larger-than-expected babies, and, by design, takes longer to reach a pregnancy 4.
ESHRE's 2023 add-ons recommendations say elective freeze-all is not recommended for routine use, and should be considered when there is a clear indication such as higher OHSS risk, endometrial problems, or PGT 5. ESHRE's 2025 stimulation guideline recommends freeze-all to minimise late-onset OHSS (strong, low-quality evidence) 1.
Would a mild or natural cycle be better?
ESHRE's 2025 guideline no longer uses 'mild stimulation' as a category, because studies defined it differently and often used total doses similar to conventional stimulation 1. For low responders, it says a modified natural cycle is probably not routinely recommended over conventional stimulation (conditional, very low-quality evidence) 1.
It adds that low responders are a varied group, and that for women with very low ovarian reserve, clinicians could choose a modified natural cycle (good practice point) 1. That is a choice for specific situations, not a general upgrade.
What about adding growth hormone, DHEA, LH or other drugs?
These are often suggested after a poor cycle. ESHRE's 2025 positions 1:
- Growth hormone: probably not recommended for low responders (conditional, very low-quality evidence); not recommended for normal responders or women with PCOS (strong) 1.
- DHEA: not recommended for low or normal responders (strong, low-quality evidence) 1.
- Testosterone: probably not recommended for low responders (conditional, moderate-quality evidence) 1.
- Adding letrozole or clomiphene to injections: probably not recommended for high or normal responders; for low responders, letrozole is probably not recommended, and clomiphene with or without injections is rated about equal to injections alone 1.
- Adding recombinant LH to FSH: probably equally recommended as FSH alone, including for low responders and women 35 and over (conditional, low-quality evidence) 1.
- Switching between recombinant FSH and hMG: equally recommended (strong, moderate-quality evidence) 1.
- Aspirin, sildenafil and myo-inositol for low responders: not recommended 1.
It depends on…
What happened in your last cycle?
Too few eggs (low response)
ESHRE: a low response alone is not a reason to cancel; a dose above 300 IU is not recommended for predicted low responders; mid-cycle dose changes are probably not recommended; growth hormone, DHEA and testosterone are not supported 1. Cochrane found no clear case for a higher dose 3. Double stimulation can be used to collect more eggs when no fresh transfer is planned (strong, low-quality evidence) 1. See our low egg retrieval page.
Too many follicles, OHSS risk or OHSS
This is where ESHRE supports change most clearly: the antagonist protocol for predicted high responders, a GnRH agonist trigger with freeze-all, and dopamine agonists to lower early OHSS (all strong recommendations) 1. A reduced dose is probably recommended 1; a dual trigger is probably not 1.
Eggs were fine, but few fertilised or embryos were poor
ESHRE's recommendations are judged on live birth and OHSS, and none is framed around embryo quality 1. Swapping FSH for hMG or adding LH is rated about equal 1, and adjuvants such as DHEA and growth hormone are not recommended 1. Look at sperm, fertilisation and lab factors first: see our poor fertilisation and poor embryo quality pages.
Was your protocol chosen using a test of ovarian response?
Yes, AMH or antral follicle count was measured
That is what ESHRE recommends for predicting low and high response (strong recommendation) 1. Ask how your result shaped the protocol and dose.
No, or only age, FSH or other hormones were used
ESHRE does not recommend age, BMI, basal FSH, inhibin B, oestradiol, progesterone or LH for predicting response 1. Ask whether AMH or AFC should guide the next plan.
Common protocol changes and ESHRE's 2025 position
| Change | What it is meant to do | ESHRE 2025 position (strength, evidence) |
|---|---|---|
| Antagonist instead of agonist protocol | Same live birth with less OHSS 2 | Recommended over agonist protocols (strong, moderate) 1 |
| Higher dose for predicted low responders | More eggs | Above 300 IU not recommended (strong, very low) 1 |
| Higher or lower dose for normal responders | More eggs or fewer side effects | Neither probably recommended over conventional dose (conditional, very low) 1 |
| Lower dose for predicted high responders | Less OHSS | Probably recommended (conditional, very low) 1 |
| Changing dose mid-stimulation | Correct an unexpected response | Probably not recommended (conditional, very low) 1 |
| Dual trigger (agonist plus hCG) | Better egg maturation | Probably not recommended for normal, low or high responders 1 |
| Agonist trigger with freeze-all | Avoid severe OHSS | Recommended for women at risk of OHSS (strong, very low) 1 |
| Freeze-all for everyone | Higher live birth | Not recommended for routine use (ESHRE add-ons 2023) 5; no cumulative gain in Cochrane 4 |
| Modified natural cycle | Gentler cycle for low responders | Probably not routinely recommended; an option in very low reserve 1 |
| Growth hormone, DHEA, testosterone | Better response or eggs | Not or probably not recommended 1 |
What could change next cycle
9 options, by strength of evidence
- Good evidence it helps 1
- Mixed, or some groups only 3
- Not enough studies yet 1
- Evidence says it doesn't help 4
Ratings follow the evidence reviews cited beside each option, not our opinion.
- Antagonist protocol instead of long agonistMixed, or some groups only
Same live birth, less OHSS; slightly more cancellations for poor response 1,2.
- Agonist trigger with freeze-all when OHSS risk is highGood evidence it helps
ESHRE 2025: strong recommendation to minimise severe OHSS 1.
- Higher dose for predicted low respondersEvidence says it doesn't help
ESHRE: above 300 IU not recommended 1; Cochrane: no clear justification for raising the dose 3.
- Adjusting the dose partway through stimulationNot enough studies yet
ESHRE: probably not recommended, given the lack of evidence of value 1.
- Dual triggerEvidence says it doesn't help
ESHRE 2025: probably not recommended for normal, low or high responders 1.
- Freezing all embryos without a specific reasonEvidence says it doesn't help
Cochrane: little or no cumulative live birth difference 4; ESHRE add-ons: not recommended routinely 5.
- Modified natural cycleMixed, or some groups only
ESHRE: probably not routine for low responders, but an option with very low reserve 1.
- Double stimulation to bank eggsMixed, or some groups only
ESHRE: can be used to accumulate eggs or embryos when no fresh transfer is planned (strong, low-quality evidence) 1.
- Growth hormone, DHEA or testosteroneEvidence says it doesn't help
ESHRE 2025: not recommended or probably not recommended 1.
Questions to ask your doctor
- What exactly fell short in my last cycle: the number of eggs, their maturity, fertilisation, embryo development, or the transfer?
- What change are you proposing, and which problem is it meant to fix?
- Where does that change sit in ESHRE's 2025 stimulation guideline, and how strong is the evidence 1?
- Was my protocol chosen using my AMH or antral follicle count, and does the result change your plan 1?
- If you suggest freezing all embryos, what is the specific reason in my case 5?
- If nothing about the protocol needs to change, what is the more likely explanation, and what would you do instead?
Questions people ask
Is a short protocol better than a long protocol?
Will a different brand of stimulation injection work better?
My progesterone was high on trigger day. Should I have frozen my embryos?
ESHRE says measuring progesterone on trigger day is probably recommended when a fresh transfer is planned, and that high levels should prompt counselling about possibly lower pregnancy rates 1. It adds that a decision to delay transfer should also weigh the number of eggs and embryos and embryo quality 1.
Do I need blood oestradiol tests during stimulation?
ESHRE says adding oestradiol measurements to ultrasound monitoring is probably not recommended, and it does not recommend timing the trigger on oestradiol levels alone 1. Scans remain the main way stimulation is monitored.
Will CoQ10 or other supplements improve my next cycle?
ESHRE's add-ons group states plainly that antioxidant therapy is not recommended in ART 5. The Cochrane review behind that pooled 63 trials in 7,760 women of antioxidants including CoQ10, melatonin, myo-inositol and vitamins, and rated the evidence on live birth very low quality 8. Cochrane's editors have since flagged nine of the included studies, seven of them retracted, while retaining confidence in the conclusion 8.
Should we do ICSI instead of IVF next time?
Not by default. ESHRE's add-ons group states that ICSI is not recommended for non-male factor infertility 5. In the largest trial it cites, 1,064 couples without male factor were randomised to ICSI or conventional IVF, with comparable live birth (relative risk 1.11, 95% confidence interval 0.93 to 1.32) and comparable fertilisation failure 5. Where the sperm is the problem, ICSI is a different question.
Most of my eggs were immature. Was the trigger given at the wrong time?
It cannot be read off one cycle. ESHRE says the association between follicle size at triggering and outcome has not been sufficiently studied, and that the timing decision is multi-factorial; most often the trigger is given when several leading follicles are 16–22 mm 1. It advises against timing the trigger on oestradiol levels alone, and rates adding an agonist to hCG as probably not recommended 1.
Do I have to take birth control pills before starting the next cycle?
ESHRE's 2025 guideline does not recommend combined oral contraceptive pre-treatment in an antagonist protocol with FSH alone, because of reduced efficacy (strong recommendation, moderate-quality evidence) 1. It adds that oestrogen or progesterone pre-treatment can be used for scheduling purposes, and that a washout period may be applied if the pill is used to programme a cycle before a fresh transfer 1.
Should my progesterone support after the transfer be changed?
ESHRE recommends progesterone for luteal support and says any of the non-oral routes — vaginal, intramuscular or subcutaneous — can be used 1. Dydrogesterone is probably recommended 1. Adding oestradiol, hCG or GnRH agonist injections on top of progesterone is probably not recommended, and adding LH belongs only in a trial 1. So the route can change; the extras are not supported.
My cycle was cancelled before egg collection. Can that be avoided next time?
Ask first why it was cancelled. ESHRE states that a low response to stimulation alone is not a reason to cancel a cycle, and that an unexpected low responder should be counselled about pregnancy prospects with the decision made individually 1. A Cochrane review found cancellation for poor response somewhat more common with antagonist than with agonist protocols (odds ratio 1.32) 2.
Is acupuncture or herbal medicine worth adding to the next cycle?
ESHRE's add-ons group does not recommend them: the acupuncture evidence on live birth is contradictory, Chinese and herbal medicines need trials with long-term follow-up, and for other complementary therapies there are too few studies to conclude anything about efficacy or safety 5. That is a statement that they are not shown to work, not that they are known to be harmful 5.
Do my AMH and antral follicle count need to be repeated before the next cycle?
Ask your doctor, because the guideline sets no repeat interval. ESHRE strongly recommends AMH or antral follicle count for predicting low and high response, and does not recommend age, BMI, basal FSH, inhibin B, oestradiol, progesterone or LH for that purpose 1. If your last protocol was not chosen on AMH or AFC, that is the thing to raise 1.
Sources
- ESHRE Guideline Group on Ovarian Stimulation; Ata B, Bosch E, Broer S, Griesinger G, Grynberg M, et al. ESHRE guideline: ovarian stimulation for IVF/ICSI: an update in 2025. Hum Reprod. 2026;41(4):498-514. ESHRE / Human Reproduction
- Al-Inany HG, Youssef MA, Ayeleke RO, Brown J, Lam WS, Broekmans FJ. Gonadotrophin-releasing hormone antagonists for assisted reproductive technology. Cochrane Database Syst Rev. 2016;4:CD001750. Cochrane
- Ngwenya O, Lensen SF, Vail A, Mol BWJ, Broekmans FJ, Wilkinson J. Individualised gonadotropin dose selection using markers of ovarian reserve for women undergoing in vitro fertilisation plus intracytoplasmic sperm injection (IVF/ICSI). Cochrane Database Syst Rev. 2024;1:CD012693. Cochrane
- Zaat T, Zagers M, Mol F, Goddijn M, van Wely M, Mastenbroek S. Fresh versus frozen embryo transfers in assisted reproduction. Cochrane Database Syst Rev. 2021;2:CD011184. Cochrane
- ESHRE Add-ons Working Group; Lundin K, Bentzen JG, Bozdag G, Ebner T, Harper J, et al. Good practice recommendations on add-ons in reproductive medicine. Hum Reprod. 2023;38(11):2062-2104. ESHRE / Human Reproduction
- Teede HJ, Tay CT, Laven J, Dokras A, Moran LJ, Piltonen TT, et al. Recommendations from the 2023 International Evidence-based Guideline for the Assessment and Management of Polycystic Ovary Syndrome. Hum Reprod. 2023;38(9):1655-1679. Monash University / ASRM / ESHRE / Endocrine Society / ESE (Human Reproduction)
- ESHRE Working Group on Recurrent Implantation Failure; Cimadomo D, de los Santos MJ, Griesinger G, et al. ESHRE good practice recommendations on recurrent implantation failure. Hum Reprod Open. 2023;2023(3):hoad023. ESHRE / Human Reproduction Open
- Showell MG, Mackenzie-Proctor R, Jordan V, Hart RJ. Antioxidants for female subfertility. Cochrane Database Syst Rev. 2020;8:CD007807. Cochrane