Pelvic inflammatory disease among users and non-users of an intrauterine device

Abstract

The correlation between pelvic inflammatory disease (PID) and a present intrauterine device (IUD) has been debated. We aimed to evaluate the differences between IUD users and non-users among women hospitalised with a diagnosis of PID. Our hypothesis was that the role of a present IUD among PID patients is minimal, if any. We performed a retrospective cohort study during 2010–2018 in a tertiary university hospital. Overall, 474 hospitalised patients were diagnosed with PID. Of these, 121 patients were IUD users. The patients without an IUD were younger and had lower gravidity and parity. Among the patients without an IUD, higher rates of prior history of PID and fever at presentation were noticed. In 23.9% (29/121) of women, the IUD was inserted less than four weeks prior to the PID diagnosis. The patients with an IUD insertion-associated PID, had lower rates of tubo-ovarian abscess (2 (6.9%) versus 24 (26.0%), OR [95% CI] 0.18 (0.04–0.84), p = .02) at presentation, as well as a shorter length of stay (LOS) (median 4 versus 5 days, p = .05). In a patient in whom the IUD was retained, hospitalisation period was shorter (median LOS 4 days versus 5 days, p = .007). PID inpatients who carry an IUD represent a specific subset of patients with a milder disease.

Impact statement

• What is already known on this subject? The correlation between pelvic inflammatory disease (PID) and a present intrauterine device (IUD) is debateable.

• What the results of this study add? PID inpatients who carry an IUD represent a specific subset of patients with milder disease.

• What the implications are of these findings for clinical practice and/or further research? Our results show that in IUD users with PID, the practice of IUD removal as part of their PID treatment is of little benefit.

Keywords:

• Intrauterine device
• intrauterine device culture
• intrauterine device removal
• pelvic inflammatory disease
• tubo-ovarian abscess

Disclosure statement

The authors declare no conflict of interest.