The effects of increasing the copper load on IUD performance: a systematic
review
P. O’Brien
Westside Contraceptive Services, London, UK
Objectives: Most modern IUDs contain more than 300mm2 of exposed
copper to increase efficacy. We undertook a systematic review of randomised
trials that had differing copper loads but identical plastic frames to assess
the effects of increasing the copper load.
Design and methods: Medline and the Cochrane Library (2004) were
searched for all randomised trials of IUDs, the abstract were reviewed and
possible articles retrieved. Trials were included that compared IUDs with
identical frames but copper loads that differ more than 100 mm2 in copper
surface area. Trial quality was assessed to see if effects varied with quality.
Ratios of cumulative rates, or ratios of Pearl rates when cumulative rates were
not available, were combined using the general variancebased method, and a fixed
effect model. A random effects model was used for the one meta-analysis with
heterogenecity.
Results: 9 trials, involving 14,625 women, were included, 3 using the
Multiload frame, and the remainder used the TCu frame, all comparing TCu380A
with either TCu200 (2 trials) or TCu220 (4 trials). None of the Multiload trials
showed any difference in efficacy at any time period to 3 years. The two-year
combined pregnancy rate ratios (RR) for high versus low copper load was 0.88
(95%CI 0.45 to 1.65, 3 trials). In contrast the TCu380A IUD had consistently
lower pregnancy rate than TCu200 (RR at one year was 0.31, 0.11 to 0.52, 2
trials). There was marked heterogeneity (p=0.007) for trials comparing TCu380A
and TCu220. At 3 years follow-up there was no statistically differences in
pregnancy rate (RR 0.99, 0.07 to 14, 2 trials). The expulsion rates of high and
low dose copper devices were statistically similar and homogenous for all
devices compared (RR 1.22, 0.74 to 1.99, 6 trials at one year; RR 0.95, 0.74 to
1.22, at 3 years, 3 trials). There were more removals for bleeding and pain with
the high load devices at 1 and 2 years (RR 1.26, 1.11 to 1.43 at two years, 5
trials), although by the third year of use the rates were statistically similar
(1.02, 0.89 to 1.19). Only one trial continued beyond 5 years, so the benefits
of extra copper on duration of use are difficult to gauge.
Conclusions: The effect on pregnancy prevention of incorporating more
copper onto the IUD frame varied with the carrier frame. The lack of effect with
Multiload may be because of the smaller increase in copper load, while the
benefit for TCu380A may also be related to placing the extra copper on the arms.
Excess removal for bleeding and pain can be expected with higher load devices
but the effect did not persist in the trials in this review.
