Local side effects observed among long-term users of norplant contraceptive implants

Original research article

Local side effects observed among long-term users of Norplantcontraceptive implants

F. Alvareza, V. Brachea, A. Faundesb,*, A. Jorgea, M.H. Sousab

aPROFAMILIA, P.O. Box 1053, Santo Domingo, Dominican RepublicbCemicamp/Unicamp, Cidade Universitaria “Zeferino Vaz” Rua Vital Brasil 200 Caixa Postal 6181, 13084-971, Campinas, Sao Paulo, Brazil

Received 29 October 2002; received in revised form 5 June 2003; accepted 6 June 2003

Abstract

The prevalence of local signs and symptoms related to the site of insertion of implants, and the association of these signs and symptomswith time of use, skin color and body mass index (BMI) was evaluated among users of the Norplant implant system. Three hundred andthree Norplant users attending at the reproductive health clinic of PROFAMILIA in Santo Domingo, Dominican Republic, were asked ifthey had ever had any discomfort at the site/arm of implant insertion and the implant insertion area was examined and hyperpigmentationand hollowing of the surface was recorded. Half of the subjects reported either pain or paresthesia or both, in almost equal numbers, sometime since insertion. Hyperpigmentation was observed in 35.6% and hollowing in 22.4% of the subjects. Report of pain and paresthesia wasinversely associated to time of use and to BMI. Hyperpigmentation was directly associated to time of use and darker skin, and hollowingof the arm surface in the implants area with time of use and BMI. Local signs and symptoms were more frequent than previously reported,although they were of mild nature and appeared not to worry most of the users. © 2003 Elsevier. All rights reserved.

Keywords: Norplant implants; Local side effects; Paresthesia; Skin changes

1. Introduction

Implantable contraception has been the most importantaddition to contraceptive methods during the last few de-cades. It has been the subject of a large number of clinical,epidemiological and sociobehavioral studies involving tensof thousands of subjects in a variety of countries. Currently,it is estimated that over 11 million women have used im-plantable contraceptives worldwide [1].

Last year, the World Health Organization convened aninternational meeting to review different aspects of thiscontraceptive system [1]. In spite of the marked interestin this method, it is remarkable that very few studies havebeen conducted with the specific purpose of obtaininginformation regarding local side effects or complaintsrelated to the insertion site. In most clinical studies,reference is made only to infection and expulsion rates.One study specifically observed the implant site in theimmediate month postinsertion [2], while another re-ported on local reactions at implant sites during the first

year of use [3]. In several clinical studies, reference ismade to complaints related to the implant site mentionedby users in their general follow-up, while in a few stud-ies, reference is made to the incidence of women discon-tinuing use due to these complaints. One study carriedout in the United States requested that a specific questionon reactions at the implant site be asked of every subject.Although it found that 10.6% of users reported changes inthe color of the skin over the implants, this, in no case,resulted in discontinuation [4]. Another study withJadelle reported �5% of women with local complaints,including pain or tenderness, numbness, tingling andhyperpigmentation [5]. We identified only one sociobe-havioral study that interviewed Norplant implant users intheir homes and questioned them about local complaints[6]. As it was published in Portuguese, it has remainedvirtually unknown to those interested in the subject.

It has been observed in our clinical experience that signsand symptoms related to the site where the implants areplaced are present among long-term users of Norplant moreoften than previously reported. Providers in our clinic hadnoted two findings among Norplant implant users: hyper-pigmentation and skin depression or hollowing at the site of

* Corresponding author. Tel.: �19-3289-2856; fax: �19-3289-2440.E-mail address: [email protected] (A. Faundes).

Contraception 68 (2003) 111–115

0010-7824/03/$ – see front matter © 2003 Elsevier. All rights reserved.doi:10.1016/S0010-7824(03)00144-6

the implants. As potential users have the right to know whatside effects may be associated with implant use beforeaccepting the method, we decided to systematically evaluatelocal side effects. For that purpose, Norplant implant usersvisiting the reproductive health clinic of PROFAMILIA inSanto Domingo, Dominican Republic, were evaluated withthe specific purpose of identifying signs or complaints re-lated to the insertion site or arm where implants werelocated.

2. Subjects and method

2.1. Subjects

Norplant implants users who came to the reproductivehealth clinic of PROFAMILIA in Santo Domingo, Domin-ican Republic, for scheduled or unscheduled visits, wereinvited to participate in this study. Three hundred and threewomen who agreed to participate were enrolled; they weremore or less equally distributed with respect to duration ofuse (1–5 years). A range between 39 and 59 women were ineach of the 5 years of use, while a slightly larger number ofwomen (n � 75) had used the implants for more than 5years and had come to the clinic for removal. Sixty-sevenpercent of the volunteers were mulatto, while 23.8% werewhite and 9.6% were black.

Weight and height was obtained in 240 of the partici-pants for whom the body mass index (BMI) was calculated.The mean weight of the 240 women was 62.3 � 12.4 kg,while the mean body mass index (BMI) was 24.5 � 4.3.

2.2. Research protocol

This was a cross-sectional, observational study. The pro-tocol consisted of observation of the implant site, and spe-cific questioning regarding complaints or symptoms relatedto the location of the implants, mostly done by a singleinvestigator. Skin color of the woman (black, mulatto orwhite), as well as duration of use of the implants wererecorded for all subjects. A single observer subjectivelydefined the skin color of the subjects. The criteria used wasto define as white those women who had the lightest coloredskin (excluding sun tan) and as black those with the darkestskin. All others were classified as mulatto, which resultsfrom the racial mix of persons with white and black skin,and which is the most common skin color in the DominicanRepublic.

For recording complaints related to the insertion site orarm, one single question was asked: Have you ever felt anydiscomfort in the arm in which you have your implantsinserted? The answers were recorded and later classified asno discomfort, pain, numbness, poking, tingling, pruritusand loss of strength in the arm.

The area of the arm where the implants were placed wasthen carefully examined for hyperpigmentation and skindepression or hollowing at the site of the implants. Hyper-pigmentation was defined as darkening of the skin over theimplants area. In some cases, there was a defined darkeningexactly over each of the individual implants, while in othercases there was a diffuse darkening over the entire area ofthe implants (Fig. 1). Skin depression or hollowing of theimplants area was evaluated with the patient’s arms ex-tended forward, palms of the hand facing upward. Compar-

Fig. 1. Photo of the arm at the site of the implants, showing the darkening of the skin over each of the capsules.

112 F. Alvarez et al. / Contraception 68 (2003) 111–115

ing both arms, it was possible in some implants users toidentify a clear depression at the site of the implants, withthe appearance of loss of the subcutaneous fat in the entirearea.

2.3. Data tabulation and analysis

The data regarding symptoms and complaints were tab-ulated first independently and later separately in two groups:one group with women who reported pain in the arm whereimplants were inserted and another with complaints classi-fied as paresthesia, such as numbness, weakness in the arm,pruritus, tingling and poking sensation in the arm. Some ofthe subjects reported more than one symptom or discomfortsince insertion of the implants. Thus, both the number ofwomen reporting each symptom and the number of womenwho reported one or more symptoms were tabulated.

Duration of use, skin color and BMI were treated asindependent variables (predictors) to study their associationwith local signs and symptoms treated as dependent vari-ables (outcomes). Prevalence ratio and 95% confidence in-terval were used to study these associations.

The ethical review committee of the institution approvedthe research protocol and the users were free to participateor not in the study.

3. Results

Half of Norplant implant users reported having had somekind of discomfort or symptom related to the implants siteor arm during use (since insertion until the time of theinterview). The most common complaint was pain in thearm, reported by over one of every four users. Almost thesame number of women reported ever having one or moreparesthesia-like symptoms (Table 1). Among these symp-toms, poking sensation was reported by 14% (n � 43),pruritus by 6% (n � 19), tingling sensation by 5% (n � 14),weakness in the arm by 4% (n � 12) and numbness by 2%(n � 7). Direct evaluation of the arm identified hyperpig-mentation in over one third of the users and skin depressionin 22% (Table 1).

The frequency of ever having pain in the arm decreasedwith time of use, was not associated with skin color and was

more often reported by women with abnormally low BMI(Table 2). Similar associations were found with report ofever having paresthesia-like symptoms, except that heavierwomen were not different in reporting paresthesia than thegroup with the lowest BMI (Table 3).

Both hyperpigmentation and skin depression were sig-nificantly more frequently observed among long-term users(�2 years of use) than among women within the first 2years of use of the method (Tables 4 and 5). Moreover, skindepression was observed in only 1 out of 21 women exam-ined within the first 6 months of use (data not shown intable).

Table 1Women ever reporting local complaints or symptoms and findingsobserved upon inspection of implant site

Complaint/sign na %

Pain 86 28.4Paresthesia 88 29.0Any complaint 152 50.2Hyperpigmentation 108 35.6Skin depression/hollowing at implant site 68 22.4

a Total number � 303.

Table 2Association of duration of use of implants, skin color and body massindex with ever having pain at implants site/arm

% RRa (95% CI)b

Duration of use�24 mo of use (n � 94) 35.1 1.0025–48 mo of use (n � 92) 33.7 0.96 (0.65–1.43)�48 mo of use (n � 117) 18.8 0.54 (0.34–0.85)Weighted RR 0.74 (0.55–0.99)

Skin colorWhite (n � 72) 31.9 1.00Mulatto (n � 202) 25.2 0.79 (0.52–1.19)Black (n � 29) 41.4 1.30 (0.75–2.24)Weighted RR 0.93 (0.67–1.29)

Body mass indexUp to 20 (n � 27) 48.1 1.00�20–24 (n � 100) 29.0 0.60 (0.37–0.99)�24–29 (n � 79) 20.3 0.42 (0.23–0.76)�29 (n � 34) 29.4 0.61 (0.32–1.17)Weighted RR 0.54 (0.39–0.75)

a RR � Relative risk.b CI � confidence interval.

Table 3Association of duration of use of implants, skin color and body massindex with ever having paresthesia at implants site/arm

% RRa (95% CI)b



Body mass index�20 (n � 27) 44.4 1.00�20–24 (n � 100) 18.0 0.41 (0.22–0.73)�24–29 (n � 79) 27.8 0.63 (0.36–1.09)�29 (n � 34) 41.2 0.93 (0.52–1.66)Weighted RR 0.62 (0.45–0.86)

a RR � relative risk.b CI � confidence interval.

113F. Alvarez et al. / Contraception 68 (2003) 111–115

The risk of hyperpigmentation was low among whitewomen, and increased considerably for women with darkerskin, in such a way that the proportion of black women withhyperpigmentation was sevenfold higher than among whitewomen. While �10% of white women had hyperpigmen-tation of the skin at the site of the implants, almost twothirds of black women had these local signs (Table 4). Onthe other hand, there was no association between skin colorand hollowing of the arm surface at the site of the implants(Table 5).

The women’s BMI was associated with both hyperpig-mentation and skin depression at the site of the implants.

Both signs were about 2-1/2 times more frequently observedamong women with BMI �30 than among those with BMI�20. A trend of higher prevalence of these symptoms withincreased BMI was observed (Tables 4 and 5).

4. Discussion

The results of this study indicate that local signs and/orsymptoms that may have cosmetic implications and causesdiscomfort occur in many implant users. This is in contrastto the very limited attention given to these local problems inthe abundant literature on this method. While it appears thatusers do not attach great importance to these local problems,women have the right to be informed of the risk of experi-encing any of these local symptoms and signs.

These results confirm and broaden the previous report ofpain in the arm or site of the implants, and of numbness inthe arm, reported among users of all implant systems. The28.2% of users reporting pain in this study, is similar to the28% found in a US Norplant study that also specificallyquestioned women about there symptoms within the firstyear of use [7]. The Chilean cohort of a large study withImplanon also reported a relatively high 16.8% of womenwith pain at the implant site [8]. Most studies with differentimplant systems, however, report a much lower proportion(�1.0–7.1%), possibly because only spontaneous com-plaints were recorded [3–5,8–12].

There has been a report of two cases with neurologicalcompromise in the arm where the implants were inserted,which have been associated with the method and its inser-tion [13]. What we evaluated were less dramatic symptoms,which nonetheless represent a real source of discomfort tothe users, including numbness, loss of strength, pruritus andtingling. While the proportion of women reporting pares-thesia-like symptoms in this study is much higher thanreported in previous publications of clinical trials, it issimilar to the percentage of women who reported them in asociobehavioral study interviewing Norplant users or pastusers at their homes [6].

The high proportion of users who reported pain and/orparesthesia in this study should be viewed with cautionbecause it refers to occurrence at least once in the entireperiod of use, from insertion until the time of the interview.It is expected that the longer the time of exposure to have acomplaint, the higher the proportion of users reportingthem, even if its yearly incidence decreases with time. Theobservation of an inverse association between duration ofuse and complaint, should be interpreted as influenced byrecall bias; the longer the time since it occurred, the higherthe chance of disappearing in the user’s memory. In can beargued that the proportion of women reporting complaintsmay have been affected by selective discontinuation of use,by women with local symptoms or complaints. However,the percentage of women reporting these symptoms as areason for removal is negligible, no greater than 2% [12],

Table 4Association of duration of use of implants, skin color and body massindex with hyperpigmentation at implants site/arm

% RRa (95% CI)b





Table 5Association of duration of use of implants, skin color and body massindex with skin depression or hollowing at implants site/arm

% RRa (95% CI)b





114 F. Alvarez et al. / Contraception 68 (2003) 111–115

and cannot explain the reduction in reported symptoms bywomen with longer duration of use. Consequently, the factthat the proportion did not increase with time is a strongindication that its yearly incidence has to be decreasing thelonger the period of use.

The evidence of a recall bias, on the other hand, is afurther indication that these symptoms do not interfere in ameaningful way in the woman’s life. Thus, the high fre-quency of these local problems, reported by half of the usersinterviewed, should be weighted against the minimal rele-vance that women seem to attribute to their presence. Astudy of these local complaints with a prospective researchdesign would allow to confirm or negate the above hypothesis.

The objective physical changes at the site of the implantsmay have a different significance for users. It is interestingthat so few references to hyperpigmentation at the site of theimplants are published, while we found it in 36% of thesubjects of this study. Part of the reason for the lack ofreporting may be due to the increase in the appearance ofthis sign with longer duration of use, and to the low pro-portion of hyperpigmentation among white women (8.3%).Possible differences in skin color of the populations of otherstudies may partially explain the infrequent report of hyper-pigmentation during implants use. It would be interesting tosee if our findings are confirmed among African women andif these local changes are also observed among Asian pop-ulations in China, Indonesia or India, for example, where alarge number of users can be found.

Such differences in skin color do not explain the lack ofreport of skin depression at the site of the implants, whichwas found in over one fifth of our subjects, and in one fourthof white women. It is true that this sign appears morefrequently with longer time of use, but it was alreadypresent in 15% of the subjects at 2 years, but only in 1 of 21women during the first 6 months of use. Thus, studiesfocusing on short-term users may have less probability offinding women with this local change. It appears, however,that the main reason for not reporting is the lack of attentionthe site of the implants insertion has received in previousstudies, including our own, with implant contraceptive sys-tems.

What causes skin depression and hyperpigmentation inimplant users? Skin depression of the surface of the arm inthe area where the tubes are implanted has the appearance ofloss of the subcutaneous fat in that region. The higherprevalence of this problem among women with a largerBMI may be attributed to their thicker subcutaneous layer offat, if, in fact, this was the cause of the observed hollowing

in the arm surface. Likewise, the appearance of hyperpig-mentation seems to be related to the amount of melatonin inthe skin. The causes of these changes are unknown, al-though they may be related to the local release of the steroidor correspond to a foreign-body reaction.

This study has the limitations inherent in a cross-sec-tional observational design. Future prospective studies,comparing with appropriate controls, should be carried out,allowing identification of the incidence of pain and pares-thesia-like symptoms and the evolution of both hyperpig-mentation and hollowing with different durations of use.Such studies should also evaluate the reversal of thesesymptoms and signs after removal of the implants.

References

[1] Meirik O. Implantable contraceptives for women. Contraception2002;65:1–2.

[2] Diaz J, Rubin J, Faundes A, Diaz M, Bahamondes L. Comparison oflocal signs and symptoms after insertion of Norplant implants withand without scalpel. Contraception 1991;44:217–21.

[3] Klavon SL, Grubb GS. Insertion site complications during the firstyear of Norplant use. Contraception 1990;41:27–37.

[4] Sivin I, Mishell DR, Darney P, Wan L, Christ M. Levonorgestrelcapsule implants in the United States: a 5-year study. Obstet Gynecol1998;92:337–44.

[5] Sivin I, Alvarez F, Mishell DR, et al. Contraception with twolevonorgestrel rod implants. A 5-year study in United States andDominican Republic. Contraception 1998;58:275–82.

[6] Hardy E, Goodson P, Rodrigues de Souza T. O Sistema anticoncep-tivo Norplant do ponto de vista das usuarias. Rev Reproducao 1991;6:27–45.

[7] Frank ML, Poindexter AN, Cornin LM, Cox CA, Bateman L. One-year experience with subdermal contraceptive implants in the UnitedStates. Contraception 1993;48:229–43.

[8] Urbancsek J. An integrated analysis of nonmenstrual adverse eventswith Implanon. Contraception 1998;58(Suppl):1095–155.

[9] Cravioto MC, Alvarado G, Canto-de-Cetina T, et al. A multicentercomparative study on the efficacy, safety, and acceptability of thecontraceptive subdermal implants Norplant and Norplant-II. Contra-ception 1997;55:359–67.

[10] Mascarenhas L. Insertion and removal of Implanon. Contraception1998;58(Suppl):79S–83S.

[11] Croxatto HB, Urbancsek J, Massai R, Coelingh Bennink H, van BeekA. A multicentre efficacy and safety study of a single contraceptiveimplant Implanon. Implanon Study Group. Hum Reprod 1999;14:976–81.

[12] Brache V, Faundes A, Alvarez F, Cochon L. Nonmenstrual adverseevents during use of implantable contraceptives for women: data fromclinical trials. Contraception 2002;65:63–74.

[13] Hueston WJ, Locke KT. Norplant neuropathy: peripheral neurologicsymptoms associated with subdermal contraceptive implants. J FamPract 1995;40:184–6.

115F. Alvarez et al. / Contraception 68 (2003) 111–115

Documents

Local side effects observed among long-term users of norplant contraceptive implants