Category: Management

Which of the following is a standardized tool used to outline daily nursing care to clients who have common needs or medical conditions?
Concept map
Critical pathway
Protocol

Randomised controlled trial to show that access to emergency obstetric care reduces maternal mortality, include advantages and disadvantages of randomised controlled trial
Overview
To test whether improved access to emergency obstetric care (EmOC) reduces maternal mortality, a randomized controlled trial (RCT) can be designed at the cluster or facility level rather than randomising individual women, because access interventions operate at system/community levels and individual withholding of emergency care would be unethical. A pragmatic, cluster-randomised or stepped-wedge RCT is typically the most appropriate approach.
Possible trial designs
Cluster-randomised trial:
Randomise health facilities, health districts, or communities to receive a package of EmOC interventions now versus later (control).
Stepped-wedge cluster trial:
All clusters start as controls and cross over to the intervention in random order over time; allows every cluster to eventually receive the intervention and is often more acceptable ethically and politically.
Pragmatic trial:
Implement the intervention under routine programmatic conditions to maximise external validity.
Key elements of an RCT to evaluate EmOC
Intervention package:
Ready access to comprehensive EmOC (skilled birth attendants, surgical capability for obstetric emergencies, blood transfusion, anesthesia, referral/transport systems, 24/7 availability, essential drugs).
Primary outcome:
Maternal mortality (maternal deaths per 100,000 live births) measured at cluster level over a prespecified follow-up period.
Secondary outcomes:
Severe maternal morbidity, perinatal mortality, facility delivery rates, time-to-intervention for obstetric emergencies, process indicators (e.g., availability of blood, surgical capacity).
Population and sampling:
Define catchment populations for each cluster; ensure adequate sample size accounting for intracluster correlation (ICC).
Randomisation and allocation concealment:
Randomise clusters with transparent procedures; consider stratification by baseline maternal mortality or facility level.
Blinding:
Blinding of participants/providers is generally not possible; outcome assessment can be blinded where feasible (e.g., independent review committees).
Data monitoring:
Independent Data Safety Monitoring Board (DSMB) with preplanned interim analyses for harm/benefit.
Ethical safeguards:
Maintain standard care in control clusters, obtain community and individual informed consent processes per local regulations, ensure rapid referral pathways for emergencies in all clusters if possible.
Sample size and analysis considerations
Calculate sample size using expected baseline maternal mortality, anticipated relative reduction, desired power, significance level, average cluster size, and ICC. Use cluster-level or mixed-effects regression models to account for clustering, time effects (especially in stepped-wedge), and potential confounders. Conduct intention-to-treat analysis and prespecified subgroup analyses.
Ethical and practical considerations
Equipoise:
There must be genuine uncertainty about the magnitude of benefit from the proposed package in the specific context.
Risk of withholding care:
Directly withholding established emergency care is unethical; designs should ensure control clusters continue to receive the best available standard care and consider stepped-wedge to eventually offer the intervention to all.
Contamination:
Women may seek care outside their cluster, diluting effect estimates; define and monitor catchment boundaries and use analytic methods to handle contamination.
Implementation complexity and cost:
Strengthening EmOC requires infrastructure, staffing, supplies, and supervision—expensive and logistically demanding.
Timeframe:
Maternal mortality is a relatively rare outcome; long follow-up and large samples may be required.
Advantages of an RCT for this question
Strongest design for causal inference—randomisation balances measured and unmeasured confounders across arms.
When cluster-randomised or pragmatic, results are applicable to real-world health systems and programmatic implementation.
Can provide reliable estimates of effect size and inform policy decisions and scale-up.
Stepped-wedge designs enhance acceptability by ensuring all clusters receive the intervention eventually.
Disadvantages and limitations of an RCT in this context
Ethical constraints:
Withholding or delaying potentially life-saving services raises important ethical challenges.
Large sample and cost:
Maternal death is a rare event, so trials may require many clusters and substantial resources.
Contamination and spillover:
Movement of patients and providers between clusters can bias results toward the null.
Complex implementation:
Strengthening systems (training, supply chains, blood services) is multifaceted and hard to standardise across sites.
Long timelines:
Time needed for implementation, uptake, and sufficient outcome accrual can be lengthy.
External validity constraints:
Trial conditions (support, supervision) may differ from routine scale-up, potentially inflating effectiveness estimates.
Practical recommendations
Use a cluster-randomised or stepped-wedge design with clear ethical safeguards and DSMB oversight.
Define robust process indicators and intermediate outcomes (e.g., timely intervention rates) in addition to maternal mortality to improve power and interpretability.
Engage communities, health authorities, and stakeholders early to ensure acceptability and sustainability.
Plan for rigorous monitoring, quality assurance, and analytic methods that account for clustering, time trends, and potential contamination.