Background: Complications of unsafe abortion are a major contributor to maternal deaths in developing countries. This study aimed to evaluate the clinical assessment for life-threatening complications and the following management in women admitted with complications from abortions at a rural hospital in Uganda. Methods: A partially completed criterion-based audit was conducted comparing actual to optimal care. The audit criteria cover initial clinical assessment of vital signs and management of common severe complications such as sepsis and haemorrhage. Sepsis shall be managed by immediate evacuation of the uterus and antibiotics in relation to and after surgical management. Shock by aggressive rehydration followed by evacuation. In total 238 women admitted between January 2007 and April 2012 were included. Complications were categorized as incomplete, threatened, inevitable, missed or septic abortion and by trimester. Actual management was compared to the audit criteria and presented by descriptive statistics. Results: Fifty six per cent of the women were in second trimester. Abortion complications were distributed as follows: 53 % incomplete abortions, 28 % threatened abortions, 12 % inevitable abortions, 4 % missed abortions and 3 % septic abortions. Only one of 238 cases met all criteria of optimal clinical assessment and management. Thus, vital signs were measured in 3 %, antibiotic criteria was met in 59 % of the cases, intravenous fluid resuscitation was administered to 35 % of women with hypotension and pain was managed in 87 % of the cases. Sharp curettage was used in 69 % of those surgically evacuated and manual vacuum aspiration in 14 %. In total 3 % of the abortions were categorized as unsafe. Two of eight women with septic abortion had evacuation performed during admission-day, one woman died due to septic abortion and one from severe haemorrhage. Conclusions: Guidelines were not followed and suboptimal assessment or management was observed in all but one case. This was especially due to missing documentation of vital signs necessary to diagnose life-threatening complications, poor fluid resuscitation at signs of shock, and delayed evacuation of septic abortion.
This study was carried out at a rural hospital in South-western Uganda. The hospital serves a population of 100,000 inhabitants providing 40,000 consultations annually. Clinical officers were responsible for the outpatient department where the women initially were assessed and management initiated. Two medical officers were in charge of the maternity ward, the adult ward, the HIV-department, the outpatient department and the surgical ward. Only medical officers performed the uterine evacuations. The maternity ward was further staffed with midwives, nurse-midwives and nurses. Nursing staff was present 24 hours a day and a medical officer was on call during night shifts. Induced abortion “on demand” was not performed at the hospital. Contraceptives were offered for free at family planning counselling before discharge. The sample size included all women admitted during a five-year retrospective period from January 2007 to April 2012. This period was chosen as the resources and opportunities to manage abortion complications at the hospital did not change during this period. Thus, uterine evacuation could not be performed prior to the inclusion period. Furthermore, the five-year period may eliminate selection bias such as seasonal variation in pregnancy incidence, opportunity to reach the hospital and variations in staff composition and experience. Inclusion criteria were all women registered as admitted with incomplete, threatened, inevitable, missed and septic abortion. Cases were excluded if management only took place at the outpatient department, the women were referred during treatment or the abortion showed to be complete (Fig. 1). In these cases the results would be misleading since full management could not be assessed. Women with a gestational age above 28 weeks were excluded as deliveries from this point are considered a birth in Uganda. The diagnosis and gestational age was based on ultrasound in less than one third of the cases although not routinely used for threatened abortion. In cases where ultrasound was not applied the gestation age was established by last normal menstrual period or uterine size estimated by clinical or medical officers. Research material collected by inclusion of five abortion types and defined exclusion criteria. OPD: outpatient department Steps one to three of the classic five-step CBA cycle were performed. These steps consist of establishment of criteria for good quality care, data collection and analysis of the findings. We could not complete the audit cycle by implementation of changes (step four) and re-evaluation (step five) because of time constraints. Preliminary results were instead presented to the staff and recommendations on how to improve practice in the future were discussed in plenary. Step one was to establish realistic and relevant criteria for optimal assessment and management of abortion complications. The hospital’s local standards for management turned out to be inaccurate, missing, incomplete and remarkably different from international standards (Table 1). Thus, manual vacuum aspiration was only part of the hospital’s guideline regarding “unsafe abortion”, and the guideline did not mention the importance of immediate evacuation at signs of sepsis or excessive bleeding, but only that it should be considered after stabilization of the patient with IV fluids and antibiotics. Furthermore, oxytocin and ergometrine should not be used for medical abortion and finally fluid resuscitation at a blood pressure < 100 mmHg should be managed with two litres of fast running fluids. The hospital’s guideline for management of abortion A guideline for missed abortion did not exist IM intramuscular, IU international units, IV intravenous, L litre, mcg micrograms, mg milligrams, ml millilitres, mmHg millimetres of mercury, MVA manual vacuum aspiration, NS normal saline, POC products of conception, SC sharp curettage aNo recommendation for type of evacuation procedure bNo recommendations for procedure for inevitable abortion above 16 weeks of gestational age and active bleeding Our criteria were selected based on acceptability, simplicity, feasibility and critical importance after reviewing international guidelines and national guidelines (Table 2) [12, 16, 17]. The national standards are published by the Ugandan Ministry of Health and based on WHO’s guidelines [17]. The medical doctor in charge of the hospital ensured prior to the data collection that our audit criteria were realistic according to the local setting. Audit Criteria for acceptable management aTo be measured at admission. Oxygen saturation is an important vital to monitor, but the ward did not have the equipment to perform the measurement bAxil measurement, 0.5° was therefore added to the raw data to compensate The data collection was performed from February to April 2012. Potential eligible cases were identified by provisional diagnosis from the maternity ward’s admission-register, e.g. abortion, vaginal bleeding or abdominal pain. Based on the admission date from the register, we could identify the medical records anonymously from the archive and include or exclude due to the final diagnosis. To identify cases either not registered in the admission-books or registered differently in the admission book and the medical file all gynaecological case files in the archive were cross-checked. The contraceptive care registration system showed major differences compared to the maternity ward’s register, and consequently we could not reliably identify the women’s post-abortion contraceptive use. Accordingly, this resulted in the decision to exclude family planning as a part of the partially completed CBA. A standardized data collection form was pre-made and used to screen the medical records. In case a procedure was not documented, it was assessed as ‘not performed’. The staff did not ask the women consistently whether they had had an illegal induction of the abortion, but when the women informed the staff it was documented in the medical file. The third step was to compare the actual practice to the selected audit criteria. After the data collecting period the hospital received the data collection form and was encouraged to continue the cycle by implementing the suggested changes and re-evaluate the management. Data were continuously entered into Microsoft Access. IBM SPSS 20.0 was used to calculate mean, median and standard deviation. The hospital and its personnel have been kept anonymous. Ethical permission from the hospital to conduct the study was obtained prior to the study from the hospital´s authorities and made available for review by the Editor of this journal. The Danish National Committee on Biomedical Research Ethics in Denmark have assessed the project and found that no further ethical approvals for the analyses were necessary according to Danish law.
N/A
DIMA AI Care
