top of page

Diana Green Foster, PhD

Desire for control over IUD removal video

Q: In the 2014 paper, only 59% willing to try self-removal and then 19% successful is low.  How do you see self-removal fitting into wider removal services?  


A: Not all women will be interested in removing their own IUDs.  The clinician placing the IUD has to ask the woman if she might be interested in self-removing the device later. If she does not like the idea or if she has concerns about a partner removing her IUD against her will, then the provider can trim the strings short. If she is interested in self-removal, the clinician can trim the strings 2 inches (5cm) or longer from the external os.  


Q: If IUD strings are left long (5-7 cm), is there higher incidence of accidental removal or partial expulsions, from tampon use or perhaps sex? 


A: There aren’t hard data on this but it seems reasonable to assume that there is a higher risk of accidental removal with longer strings.  This may be a trade off the woman is willing to make in order to have control over removal. Women who want to use a menstrual cup, who don’t trust their partner not to remove it, or who are sure that they want to use the method for a long time may prefer to have shorter strings. Women who want control over discontinuation may want longer strings.  


Q: How long does the string need to be to be successfully removed?   


A: More work should be done on this. Based on our data, 2 inches (5 cm) from the external os may be long enough for successful removal.  


Q: Why was the third position, with a person standing with one leg up on a step, less successful for self-removal? 


A: I don’t know why one foot up was less successful.  Note that the trial occurred inside a clinic.  At home, a woman could try a variety of positions until she finds one where her ability to grasp the string is greatest.  


Q: For patients whose strings are left longer,  has any qualitative data been collected rated to “bothersome strings” (reported by pt or partner)?  


A: I am not familiar with the complete literature on IUD complaints.  Anecdotally, I know that both long strings and short strings can be bothersome. One can feel long strings but short strings can poke. Short strings also can retreat into the cervix and require a more painful removal process. Only the woman would know which she prefers.  


Q: What were the characteristics of self removal failure?  Were most partially removed, trapped in cervical canal?  If partially removed this could potentially decrease efficacy of the method until the pt is able to obtain an appointment for removal or transition to another method.   


A: In our trial, nobody partially removed their IUD.  (This was not a concern in Dr. Amico’s data either.) All the failures to self-removal involved women who were not able to grasp the string and therefore could not pull on it.    


Questions for both Dr. Foster and Jennifer Amico  


Q: Do you have any concerns about self-removal, especially in the LMIC context or areas where STI  and other vaginal infections may be high.  In addition, is there any concern that the removal appointment is an important opportunity for provider contact and self-removal would remove that opportunity for healthcare provisions for other issues  


A: Amico: I do not have any concerns about uterine infection resulting in self-removal attempts. This was not mentioned in the data as an outcome. Also, since self-removal (and provider-removal) does not require instrumentation into the cervix at all, this clinically would not be a concern.  


A: Amico and Foster: We need to keep in mind that people can discontinue most other reversible contraceptives without seeing a clinician first. Self-removal simply makes IUD discontinuation as accessible as discontinuation of short acting methods.  Requiring a visit for IUD discontinuation is like requiring a pap for an oral contraceptives prescription or an IUD insertion: while ideally people will have cervical cancer screening and access to information about effective contraception, it does not make sense to hold one service hostage in order to  access  another.  

bottom of page