Methods
We developed algorithms for four case scenarios, based on findings of a rapid review on third stage abnormalities, identification of the most common deviations, severity of related complications or clinical outcomes, and discussion with the WHO technical working group. These were as follows: Approach to Postpartum haemorrhage (PPH) after vaginal delivery, Diagnosis and management of uterine atony, Diagnosis and management of genital tract trauma, Diagnosis and management of retained placenta. A detailed description of the methodology used for the development of the algorithms is provided in Chapter 2
Search strategies The search was directed at research databases (Cochrane, Medline and CINAHL) for evidence based information on the identification and management of third stage abnormalities in low-resource settings, with a focus on PPH. In addition, international and national guidance documents were identified and reviewed (FIGO, RCOG, NICE, ACOG, SA-NCCEMD, ANZCOG, SOGC). Published systematic reviews were also searched and identified. The search was limited to English language literature. Where high quality evidence was lacking, individual trials and some case series were examined.
The literature review was guided by the hierarchy of evidence and prioritised WHO guidelines followed by other international or national guidelines using GRADE methodology. In the absence of guidelines on a case scenario, a combination of existing studies and expert opinion was used to determine key points for consideration in the algorithm. The highest level of evidence found was used to support the decisions along the management pathway, in the order of up-to-date systematic review (with meta-analyses), up-to-date systematic review (without meta-analyses), any available systematic review, validated decision rules, randomized controlled trials, non-randomized controlled trials, observations studies, and consensus documents.
Population and Setting The algorithms were developed to cover the assessment and management of pregnant women with singleton, term pregnancies considered to be low risk, who developed third stage abnormalities after vaginal delivery. Health facilities in low-and middle-income countries were the priority. However, the algorithms are applicable to any health care setting, and possible adaptations that may be required were acknowledged. The target users for these algorithms are skilled health personnel providing care during childbirth working alone or as part of teams, particularly midwives, clinical officers or equivalent, non-specialized clinicians (i.e. clinicians without specialist training in obstetrics but who also provide care for women in labour), and specialists.
Development of the Algorithms After collating the evidence, a selection process for inclusion of the evidence in the algorithm took place. Selection was based on relevance of the evidence to the key decision points and severity of the condition targeted by an intervention. The selection also accounted for the strength of evidence and applicability and feasibility in a LMIC context. If there were inconsistencies among guidelines, the most up-to-date guidelines and evidence were reviewed and used to inform consultation with experts. A list of inconsistencies was discussed at the WHO Technical Working Group meeting and a consensus reached on evidence for the algorithm. Algorithms were structured to cover criteria that should be used to suspect a deviation from normal third stage of labour, initial assessment, probable causes and potential differential diagnosis, further assessments to reach the most probable diagnosis, and management of the condition. Draw.io, an open source diagramming software, was used to construct the algorithm in a flowchart format. The online software facilitated remote working on SELMA by the WHO Technical Working Group. The algorithm was composed of standardised but variable shaped boxes, representing either a clinical state, decision point, action task, or link to a different algorithm. Each box was numbered and joined to other boxes via arrows, to orientate the reader to the direction of flow. The numbers also corresponded to a table of evidence, showing the evidence source for the action and decision points. They algorithms underwent internal peer review by the WHO Technical Working Group and revisions where needed.