Predicting success of intrauterine insemination using a clinically based scoring system

Anna Lena Zippl, Alfons Wachter, Patrick Rockenschaub, Bettina Tóth, Beata SeeberArchives 2022 · DOI

TL;DR

This paper introduces a novel clinical scoring system designed to predict the likelihood of pregnancy following Intrauterine Insemination (IUI). By analyzing routinely collected patient data, this score aims to empower clinicians and couples with better-informed decision-making and personalized counseling regarding IUI treatment.

What Problem Does This Paper Address?

Intrauterine Insemination (IUI) is a common first-line treatment for infertility, but its success rates can vary significantly between individuals and couples. While IUI is simpler and less costly than In Vitro Fertilization (IVF), predicting which couples are most likely to succeed can be challenging. Existing predictive models often rely on outdated criteria or focus on specific infertility causes, limiting their broad clinical applicability. This paper addresses the need for a simple, clinically applicable scoring system that can estimate a couple's probability of pregnancy via IUI, thereby facilitating more personalized and evidence-based counseling.

Methods

The researchers conducted a retrospective cohort study involving 1437 homologous IUI cycles performed at a single university-based reproductive medical center between 2009 and 2017. They identified predictive factors for pregnancy using a multivariable continuation ratio model, a statistical technique that models the probability of an event occurring at different stages. This model was used to assess the impact of various clinical parameters, including female age, anti-Mullerian hormone (AMH) levels, total progressive motile sperm count (TPMSC), and female infertility diagnosis. Based on the significant predictors identified, a simple clinical score was developed.

Datasets

The study utilized a dataset of 1437 homologous IUI cycles from 758 couples collected retrospectively between 2009 and 2017. This dataset included demographic and clinical data such as female age, body mass index (BMI), anti-Mullerian hormone (AMH) levels, and semen analysis parameters (specifically, total progressive motile sperm count - TPMSC). No new benchmark datasets were introduced.

Key Insights and Ideas

  • Individualized Prediction is Crucial: The study strongly emphasizes that IUI success is not a one-size-fits-all outcome. Couples' characteristics play a significant role, necessitating personalized predictions rather than relying on general success rates.
  • Beyond Basic Metrics: While age and sperm count are known factors, the inclusion of AMH as a key predictor highlights its growing importance in assessing ovarian reserve and predicting fertility outcomes.
  • Simplifying Complexity: The development of a straightforward clinical score from complex multivariable analysis is a significant contribution. This makes advanced statistical insights accessible for practical clinical use.
  • Focus on Safety and Efficacy: The study's approach to developing a score that aligns with current guidelines (avoiding multifollicular growth to reduce multiple gestations) is forward-thinking and clinically relevant.

Key Results

The study found a per-cycle pregnancy rate of 10.9% and a cumulative pregnancy rate of 19.4% after repeated cycles. Key predictors of decreased pregnancy probability included:

  • Female age ≥ 35 years (OR 0.63)
  • Endometriosis, unilateral tubal factor, or anatomical alteration (OR 0.54)
  • AMH < 1 ng/ml (OR 0.50)
  • Total progressive motile sperm count (TPMSC) < 5 million (OR 0.47)

A predictive clinical score ranging from 0 to 5 was developed. Couples with a score of 5 had a cumulative pregnancy probability of nearly 45% after 3 cycles, while those with a score of 0 had only a 5% probability. The study also noted that the type of ovarian stimulation medication (clomiphene vs. gonadotropins) did not significantly influence pregnancy rates in their cohort.

Limitations and Future Work

The authors acknowledge several limitations:

  • Retrospective Design: Data was collected retrospectively, which can introduce potential unknown confounders and missing data.
  • Single Center Study: The findings are based on data from one center, which may limit generalizability to other populations or healthcare settings.
  • Potential Selection Bias: Couples' decisions to continue or discontinue treatment could introduce bias, although the authors found patient characteristics to be stable across early cycles.
  • Pre-wash Sperm Parameters: Only pre-wash sperm parameters were available, and future studies could investigate the correlation with post-wash parameters.

Future work should focus on external validation of the scoring system to confirm its generalizability across different populations and centers. The authors suggest that once validated, this model can significantly facilitate patient-oriented counseling and decision-making in reproductive medicine.

Summary updated

Abstract

IUI success rates vary widely depending on couples' characteristics. A simple to use score could be used to estimate a couple's chance of achieving pregnancy via IUI, facilitating individualized counseling and decision-making.
BibTeX
@inproceedings{zippl2022predicting,
  author = {Anna Lena Zippl and Alfons Wachter and Patrick Rockenschaub and Bettina Tóth and Beata Seeber},
  title = {{Predicting success of intrauterine insemination using a clinically based scoring system}},
  booktitle = {Archives},
  year = {2022},
  doi = {10.1007/s00404-022-06758-z},
}
Figure