For 3 months, the entire change package interventions were piloted in Hospital B maternity wards supervised by the hospital local improvement team members, the project clinical lead and the executive of nursing operations.

The forms related to haemorrhage risk assessment, and maternal haemorrhage clinical pathway were fine-tuned according to the workflow in the maternity wards and the valuable feedback received from the end user experts/direct patient care staff (online supplemental appendixs 4; 5)

The first learning session was conducted over 3 days; 5, 6 July 2020 the faculty team explained the change package interventions and reflected on the pilot completed in Hospital B.

On 12 July 2020, the participating hospitals were given the chance to present their action plans and recommendations for change package interventions modifications.

Improvement teams quickly adopted the change package due to early project engagement.

The second learning session was conducted on 1/10/2020 where participating hospitals’ teams presented their key performance indicators (KPIs) results, updated action plans, implementation barriers and lessons learnt.

It was noted that hospital E did not report any cases of maternal morbidity in their results, The faculty team emphasised the role of the trigger tool in identifying morbidity cases and directed the team to validate the sources of data used.

Hospital B, being the pilot venue showed the best performance in both the outcome and process measures.

Due to transition from paper based medical record to Electronic Medical Record in hospital A, the team found some difficulty in retrieving the medical records of flagged cases by ACOG criteria for review in September which led to low compliance in the chart review KPI.

Faculty team provided feedback and guidance based on each hospital’s progress status.

Medical record forms implementation was delayed due to logistical and printing problems in hospitals A, C and D; addressed immediately.

It was noted that there was inter-hospital difference in interpretation and action regarding the ACOG trigger criterion for unplanned admission to higher level of care. All cases requiring higher level of care regardless of the physical location of their admission were flagged as mandatory triggers of SMM, after initially were excluded from the numerator in hospital E.

Some SMM cases (postpartum haemorrhage) had inadequate antenatal care (incorrected iron deficiency anaemia) in Hospital B, addressed in the next cycle.

Instances of excluding admissions to high-dependency unit (HDU) were corrected. On review, 79% of unplanned HDU admissions had SMM in hospital A; corrected in the following cycle.

The third learning session was conducted on 24 December 2020 where participating hospitals’ teams presented their KPIs results, updated action plans, implementation barriers and lessons learnt.

Hospital C showed positive outlier in reporting SMM that was attributed to leadership enforcement of the implementation of trigger tools and enhanced communication with the blood bank to flag cases with four or more units of blood transfusion.

All hospitals showed improvement in SMM reporting and to capitalise on this, hospitals were advised to identify their most frequent SMMs and start local improvement projects addressing clinical care.

The faculty team provided feedback and guidance based on each hospital’s progress status.

Hospital A team observed intentional physicians’ underestimation of blood loss manipulating the triggers for the haemorrhage pathway. The faculty team suggested to add a KPI that monitor physician compliance with the pathway implementation through retrospective medical record review for a representative sample and include it in the Ongoing Physician Performance Evaluation.

Hospital B recommended to share the results of the maternal haemorrhage risk assessment with the patient antenatally to enhance patient engagement and compliance which was supported by the faculty team.

To promote shared learning, establish maternal safety network and validate low reporting rates, the faculty team suggested on site visits among hospitals’ teams in which high performing teams were matched with low performing teams (two visits were conducted).

In the Summative Congress the faculty and the participating hospitals’ teams reflected on the entire project covering the following points; review of the project progress in achieving the aims, The best performing hospital (hospital B) shared its success factors, brainstorming sessions tackling the issue of sustaining improvement after concluding the project, lessons learnt and future recommendations presented by the project clinical leads, and planning for project publication and celebration of the project conclusion.

The faculty team along with hospitals’ teams agreed on keeping track of the project KPIs as part of the departments’ quality monitoring systems to ensure sustainability of improvement and initiate local improvement projects when needed.

Due to the success of the project a recommendation was raised to the governing body to approve a strategic improvement roll out of the project change package to all the network’s hospitals providing maternal care.