Physiological patterns in pregnancy

Thesis

The Confidential Enquiries into Maternal Deaths in the UK have highlighted an urgent need for development of a national Modified Early Obstetric Warning System (MEOWS) to aid the more timely recognition and treatment of complications in pregnancy. Six vital signs are used to assess the clinical status of women presenting acutely during and immediately after pregnancy: heart rate, oxygen saturation, systolic and diastolic blood pressure, temperature and respiratory rate. In the case of MEOWS, these define the thresholds that determine if a woman requires further medical review. Dynamic changes in maternal vital sign physiology associated with pregnancy complicate the use of vital signs to detect physiological deterioration. These changes are currently poorly described in the literature.

The aim of this work is to establish the normal physiological response to pregnancy in low-risk pregnant women, in terms of expected changes in vital signs. Using this knowledge, we will propose the first evidence-based, gestation-specific MEOWS chart.

To achieve this, we have set up the Prediction of Physiological Patterns in Pregnancy (4P) study. Using a custom-designed digital health system, healthcare professionals and women self-monitoring at home collect large numbers of vital sign measurements electronically. In particular, as respiratory rate currently is the only vital sign measured manually in routine clinical care, we propose two methods for electronic measurement of respiratory rate: using the in-built accelerometer of a smartphone and using the photoplethysmogram signal output by a pulse oximeter. These methods are validated against the current reference standard derived from a 15-second nurse count of breaths.

This thesis uses a dataset of 2,645 sets of vital sign measurements, collected from 439 women during pregnancy, in the 4P study. Using the electronic methods for measurement of respiratory rate, we obtain respiratory rate estimates with a mean absolute error of 2.9 respirations per minute compared to the reference measurement. From these data, we design reference ranges for changes in all six vital signs conditional on gestational age. Using a centile-based approach, we propose gestation-specific thresholds of normality for each vital sign, which will form the basis of a new MEOWS score to alert for physiological deterioration. Defining normality is the first step in recognising abnormality.

Authors: Loerup, L

• DOI: 10.5287/ora-z6pjb7r8x
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: Vital signs; Pregnancy; Patient monitoring
• Subjects: Biomedical engineering