Implementation and Evaluation of Obstetric Early Warning Systems in tertiary care hospitals in Nigeria

Blog 11 Jun 2021
47

What is this about?

Early Warning Systems (EWS) have been developed to facilitate the timely presence of appropriately skilled staff to attend clinically deteriorating patient. They provide the opportunity to aggregate the impact of sometimes subtle deteriorations in physiological observations into an overall score that, when abnormal, is used to prompt a clinical response. However, the early warning systems designed for the general population does not account for the unique physiology of pregnant women, and it does not effectively identify at-risk obstetric patients.

Using secondary data on obstetric inpatients admitted to 42 Nigerian tertiary hospitals, Umar A. et al (2020) developed and internally validated a simple obstetric diagnostic prediction model and EWS for use in resource-limited settings using recommended methodologies. The resulting EWS model performed excellently in predicting Severe Maternal Outcome and with a potential to improve adverse pregnancy outcomes in low resource settings.

What did we do?

We assessed the effectiveness of this validated obstetric EWS compared to routine monitoring systems and explored the experience of health care providers implementing it. The study was implemented in 3 tertiary care hospitals across the northern regions of Nigeria. EWS was implemented in the 600-bed multispecialty University of Ilorin Teaching Hospital, a public tertiary health care centre located in the north-central region. The control sites, National Hospital Abuja, and Abubakar Tafawa Balewa university teaching hospital Bauchi are teaching hospitals in the north-central and northeast regions, respectively.

The intervention was the implementation of a statistically developed obstetric EWS.This is a simple score-based recording chart for vital signs. It includes seven clinical parameters (temperature, pulse rate, respiratory rate, systolic blood pressure, diastolic blood pressure, consciousness level (based on the AVPU (alert, voice, pain and unresponsive) scale) and mode of birth for post-partum women). Each parameter is scored as 0 for normal, 1 for mild and 2 for severe derangements. A guideline at the top of the chart guides frequency of patient monitoring and when to trigger clinicians’ review; scores of 0 or 1 are reassuring; hence require 12-hourly monitoring or as routine for post-operative patients. A score of 2 indicates the need to repeat observations after 30 minutes; if the score remains the same or rises, doctors should be informed for review. Those with scores of 3 or more are likely to deteriorate clinically and require immediate review. This tool was implemented in one of the three study sites (intervention arm) while two control sites continued with the existing practices of clinical monitoring.

In this mixed-method study, we included 2400 obstetric admissions to inpatient wards between 1 August 2018 and 31 March 2019 at these three tertiary care hospitals. The quality of patient monitoring and prevalence of outcomes were assessed through retrospective review of case notes before and 4 months after EWS was introduced. The primary outcomes were outcomes were frequency of vital signs recording, the ration of observed/expected frequency of vital sign observations and while the secondary outcomes were maternal death, direct obstetric complications, length of hospital stay, speed of clinical review, caesarean section (CS) and instrumental birth rates.

What did we find?

There was a significant improvement in the frequency of vital signs recording in the intervention site: observed/expected frequency improved to 0.91 from 0.57, p<0.005, but not in the control sites. CS rate reduced from 39.9% to 31.5% (chi-square p=0.002). No improvement was observed in the other outcomes. Health workers reported that the EWS helped cope with work demands while making it easier to detect and manage deteriorating patients. Nurses and doctors reported that the EWS was easy to use, and that scores consistently correlated with the clinical picture of patients. Identified challenges included rotation of clinical staff, low staffing numbers and monitoring equipment.

Our conclusion

The implementation of EWS improved the quality of patient monitoring, but a larger study will be required to explore the effect on health outcomes. With modifications to suit the setting, coupled with regular training, the EWS is a feasible and acceptable tool to cope with the unique demands faced in low-resource settings.

Implication for future research

All health facilities included were tertiary hospitals that provided comprehensive emergency obstetric care services. The scope and budget allowed us to have only one hospital in the intervention arm. However, this is a large university teaching hospital servicing a state with a population of 2.37 million. The feasibility and utility of implementing the EWS chart in smaller centres, including primary healthcare facilities, with smaller staff numbers was not investigated in this study. To improve generalisability of these findings, further multicentre studies with multiple intervention sites and across different levels of care (including primary and secondary care hospitals) are needed.

This work was part of Dr Aminu Umar’s PhD under the supervision of Dr Charles Ameh and Professor Matthews Mathai. An abstract describing this work has been accepted as a poster presentation for the RCOG virtual world congress 2021 and will be featured in the online World Congress edition of the BJOG journal.

Aminu is now a Specialty Registrar, Lead Employer Trust, Health Education Northeast, Newcastle.