How effective and safe is semen washing for HIV-serodiscordant couples?

The issue

The prevalence of the human immunodeficiency virus (HIV) in people of reproductive age, together with improvements in life expectancy and quality of life in HIV-positive people receiving highly active antiretroviral treatment (HAART), have meant that more serodiscordant couples are considering having children.

Since semen is a vehicle for transmitting HIV, the options for parenthood of couples in which the man is HIV-positive and the woman HIV-negative theoretically are:

  • adoption
  • donor insemination
  • biological children.

Adoption and donor insemination are alternative, safe options, but due to legal restrictions they are only offered in a few centres and countries. For couples who choose to have biological children, the present options are conception via unprotected sexual intercourse and conception via semen washing. The rate of HIV transmission for unprotected heterosexual intercourse (seropositive male to seronegative female) is estimated to be around one HIV transmission for every thousand contacts. The transmission rate may be higher in people with a high viral load or in the presence of other sexually transmittable infections. Usually, when attempting to have biological children, couples are advised against unprotected intercourse, as the priority is to prevent infection in the uninfected woman and her fetus.

The application of semen washing to reproduction assumes that spermatozoa are not a major reservoir for HIV. The method involves three-steps:

  1. filtering the liquefied semen through a density gradient;
  2. washing the recovered spermatozoa, to eliminate seminal plasma or non-spermatozoal cells; and
  3. recovering the highly motile spermatozoa by using a swim-up method.

These steps can be combined or used separately. In either case, the resulting sperm sample is divided into two parts: one for HIV testing and, if the test proves negative, the other for assisted reproduction techniques, such as intrauterine insemination, conventional in vitro fertilization and intracytoplasmic sperm injection.


A systematic review was conducted to analyse the effectiveness and safety of washing the semen of HIV-positive men so that they can be used in assisted reproduction techniques (ART). Twenty-three studies (16 clinical series and 7 before-and-after studies) were selected, most of which were performed in Europe, and the methodological limitations of the studies were identified.

Efficacy in reducing or eliminating HIV from semen samples

The range of results of the studies showed that between 0% and 20% of the semen samples tested positive for HIV after semen washing (70 positive cases in a total of 1279 samples). Various semen-washing methods and HIV-detection tests were used. Depending on the method of testing, the lowest limit of detection was 10 copies/ml. Since HIV can, theoretically, remain undetected, semen washing can greatly reduce the risk of infection, but cannot completely exclude it. 

Efficacy and safety in preventing transmission to women and children

In 914 serodiscordant couples treated with semen washing, 1680 cycles of ART were performed (72.7% by assisted insemination and the remainder by in vitro fertilization). The studies did not report any HIV infections in the women using ART or in the resulting children.

Policy considerations

  • While no semen-washing method fully guarantees elimination of HIV in motile spermatozoa, the methods do minimize the risk of transmission.
  • There is sufficient evidence showing that semen washing in HIV-positive men to inseminate women has not produced any seroconversion in the women or in the children, but the strength of the evidence is limited by the number, quality or consistency of the findings of the individual studies. The quality of the evidence is attributed to the small sample size and uncontrolled study design.
  • For an HIV-positive man, an undetectable viral load and highly active antiretroviral treatment (HAART) appear to improve the outcome of clinical pregnancy rate after semen washing.
  • HIV-serodiscordant couples should be given suitable pre-conception counselling, including information about the available alternatives to having children and the benefits and risks of each one.
  • The seminal viral load should be measured before enrolling an HIV-serodiscordant couple in a semen-washing protocol.
  • Washed semen should always be screened for HIV before performing assisted reproduction techniques (ART).
  • A register of the women subjected to this technique (and any children that result) should be created to facilitate follow-up.

Type of evidence

A systematic review was conducted in the main biomedical databases of the scientific literature through August 2005.

The views expressed in this summary are based on a publication of a HEN Network member agency and do not necessarily represent the decisions or stated policy of WHO/Europe.


This summary is an update of the following original report (in Spanish) prepared by the Catalan Agency for Health Technology Assessment and Research.