O&G Forum

OBSTETRICS & GYNAECOLOGY FORUM 2021 | ISSUE 3 | 9 and no IUD was seen in 5.8% (3/52) ( gure 5). e IUD removal rate at follow up was 24.7% (19/77). Reasons for removal included symptoms (6/19; 31.6%), symptoms and ultrasound (5/19; 26.3%), ultrasound alone (5/19; 26.3%), clinical malpositioning (IUD stem visible on speculum) (1/19; 5.3%), and patient request (1/19; 5.3%). In one case the reason for removal was unclear in the notes. Two women opted for IUD reinsertion at the same visit, the others chose a di erent contraceptive method. Surgeons with the best continuation rates were the registrars (67.6%; 23/34), followed by the medical o cers (59.3%; 16/27), the community service doctors (57.1%; 4/7), and interns (33.3%; 1/3) In 8 women who attended follow-up it was unclear from the notes who the inserter was. e amount of previous caesarean sections (p=0.053), degree of cervical dilatation (p=0.249) and HIV status (p=0.474) had no statistically signi cant e ect on continuation rates. Discussion Patients who opted for intra-caesarean placement were mostly between the age of 20 and 34, with the highest uptake in the 25-29 years age group. e uptake (14.3 – 17.4%) is similar in other studies. 21 e high proportion of single, unemployed women choosing the IUD con rm the need for LARCs in these women. e low rate (0,4%) of unbooked women is encouraging, and we can assume that contraceptive methods would have been discussed with most women at some stage during their pregnancy. e high percentage of patients who did not attend follow-up (83.5%) is concerning. e slightly better follow-up rate at NSH might be because patients could follow up at the place of insertion and thus were familiar with the setup, whereas MMH patients attend follow- up at GSH. e employed proportion of women who followed up was almost double that of the unemployed. is underscores the nancial barrier to follow-up, and rates might improve if patients could follow up closer to their homes. Follow-up rates in the literature ranges between 48 and 88.5%, but this was in prospective studies where patients were counselled extensively and may have received remuneration for transport. 8;21;22;24 Some of our patients may have attended in the private sector or at their local clinic, in which case we would not be able to see it on the hospital computer system. A study exploring reasons for the poor follow-up rates and whether women attend elsewhere may be insightful. e continuation rate of 71.6% is lower than expected. In the literature it ranges between 80 and 91%. 2;8;9;13;21;23 One could postulate that women would be more likely to attend follow-up if they experienced side-e ects or wanted the IUD to be removed, which would make the complication and removal rate higher in the follow-up than in the lost-to-follow-up group. However, this is our opinion and cannot be accurately ascertained. Most women (84.4%) experienced no immediate side-e ects, which is reassuring. Many of the side-e ects could also be attributed to the normal in ammatory response to surgery or pain, e.g. a transient tachycardia (10.7%) and/or unexplained temperature (3%). e immediate expulsion rate was 1.1 %. Four out of the ve immediate expulsions were a er emergency C/S. It may be postulated that cervices are o en dilated in the case of emergency C/S and the amount of postpartum bloodloss increased, explaining the higher expulsion rate. e incidence of endomyometritis (0.4%) and minor wound sepsis (0.6%) were low. Major complications were rare – only one woman (0.2%) had a hysterectomy secondary to endomyometritis. It is di cult to prove that the IUD caused the endomyometritis although it may have been a contributing factor. Hysterectomy does not appear to be increased a er intra-caesarean IUD insertion and there were no maternal deaths in our patient sample. Immediate complications or side-e ects did not prolong hospital stay, which is important if cost, resources, and patient satisfaction are considered. e commonest side-e ect at follow-up was a vaginal discharge (26%), with pain the second most common symptom (11.7%). In the literature pain and bleeding are the most common symptoms, with an incidence of up to 55%. 8 e high incidence of leukorrhea could be physiological, since very few patients required antibiotics. In 40.3% of cases IUD strings were not visible at follow-up which was expected. Visibility of strings can be as low as 28% a er caesarean insertion. 8 In a large observational trial by FIGO conducted in 6 low- income countries (2018), missing strings were 2.88 times more common following insertion at C/S as compared to vaginal delivery. 10 An ultrasound was performed in 67.5% of women. e GSH family planning unit scan most women at the 6 weeks visit, whereas an ultrasound was only performed in the instance of lost strings or symptoms at NSH. e continuation rate between the two hospitals was not signi cantly di erent even though the scan rate was higher at GSH. Our follow-up groups were however too small to draw any de nitive conclusions regarding this. ree patients (3/52) in whom ultrasounds were performed were found to have had undetected expulsions. e absence of visible strings is an important clue to this, and availability of ultrasound is important in these patients. One patient presented with a twin pregnancy three months a er insertion due to undetected expulsion. e importance of follow-up must be emphasised in our patient population, since they may experience pregnancy as failure of the device, whereas it was actually failure of attending F/U leading to undetected expulsion. e IUD was normally positioned in 63.5% of women where ultrasound was performed. e removal rate due to ultrasound ndings alone was 26.3%. is con rms that patients may be asymptomatic in some cases of malpositioning. With our high removal rate due to asymptomatic malpositioning, routine ultrasound at the 6-weeks visit should be considered in all women a er intra-caesarean IUD insertion. An important nding is the signi cantly decreased continuation rate a er emergency C/S (60.5%). Many factors could contribute – prolonged labour and multiple vaginal examinations increasing the risk of endomyometritis, the degree of cervical dilatation at the time of insertion, and the in ammatory response characteristic of labour triggering contractions. To our knowledge no studies have speci cally looked at this variable yet and it may need further investigation. e expulsion rate was 11.7% (9/77) if we only look at the patients who attended follow-up. e partial (5/9) and complete (4/9) expulsion rates were similar. Our expulsion rate is slightly higher than in the literature (3.9-10.9%). is may be explained by our small follow- up group, as well as the possibility of patients with complications or symptoms being more likely to attend. Expulsions detected at follow up were higher a er emergency C/S, even though not statistically signi cant (p=0.576). Clinician experience at follow-up may play a role in the partial expulsion rate – some doctors may be more experienced in performing ultrasound for IUD positioning, and may have more strict criteria for the IUD being correctly positioned. eir IUD removal rate may be higher where ultrasound is performed. Some may scan everyone at follow-up, whereas others may have a higher threshold for O&G Forum 2021; 31: 06 - 10 ORIGINAL RESEARCH Figure 5: Ultrasound findings at follow-up NOT SPECIFIED 8% NORMALLY PLACED 63% MALPOSITIONED 23% IUD NOT SEEN 6%