The development and testing of an algorithm for diagnosis of active labour in primiparous women

Abstract

Objectives

to describe the development and testing of an algorithm for diagnosis of active labour in primiparous women.

Design

qualitative and quantitative methods were used. A literature review was first conducted to identify the key cues for inclusion in the algorithm. Focus groups of midwives were then conducted to assess content validity, finally a vignette study assessed the inter-rater reliability of the algorithm.

Setting

midwives from two study sites were invited to participate. Data were collected during 2002 and 2003.

Participants

midwives from the first site took part in the focus groups (n=13), completed vignettes (n=19), or both. Midwives from the second site then completed vignettes (n=17).

Findings

an algorithm, developed from the key informational cues reported in the literature, was validated in relation to content validity by the findings from the focus groups. Inter-rater reliability was tested using vignettes of admission case histories and was found to be moderate in the first test ($K=0.45$). However, after modifying the algorithm the kappa score was 0.86, indicating a high level of agreement.

Key conclusions

diagnosis of labour may be straightforward on paper but is frequently problematic in practice. This may be because the diagnosis of labour is made in a high pressured environment where conflicting pressures of workload, limited resources and emotional pressures add to the complexity of the judgement.

Implications for practice

we offer a valid and reliable decision-support tool as an aid for diagnosis of labour. The evaluation of the implementation of this tool is under way and will determine whether it is effective in reducing unnecessary admissions and improving clinical outcomes for women.

Introduction

Within the UK and across much of the developed world, home birth is now uncommon (US Department of Health and Human Services, 1999; Kitzinger, 2000; Scottish Executive, 2002). Despite continuing debate about the relative safety and efficacy of hospital or home birth (Olsen, 1997; Vedam, 2003; Macfarlane, 2004), and world-wide concern among health-care professionals over steadily increasing rates of intervention in normal labour (WHO, 1996), most women experience labour and birth in hospital. This predominantly institutionalised model of care requires a clear cut, if somewhat artificial, distinction to be made between the latent phase of labour (a poorly defined period from onset of regular contractions, during which the woman might be expected to remain at home) and the active phase, the phase in which there is increasing cervical dilatation (Austin and Calderon, 1999), when most women would be admitted to hospital.

The judgement about whether a woman is in labour or not is an important issue for midwives, who are the principal care providers for women through normal pregnancy and childbirth. Although, on paper, it would seem to be a straightforward judgement, there is evidence that, in practice, it is often problematic and that there can be serious clinical and resource implications if misjudgements are made. The extent of the problem is illustrated by the findings of an audit of a workforce planning tool for midwifery services (Ball and Washbrook, 1996), which reported that up to 30% of women admitted to labour wards in the UK were subsequently found not to be in labour. Challenges include identifying the way in which midwives make a judgement about whether or not a woman is in labour and, if appropriate, developing ways of supporting them with this process.

Following the framework suggested by the Medical Research Council for evaluating complex interventions in health care (MRC, 2002), we conducted a series of studies in which we deconstructed the process of judgement and decision-making by women and midwives about onset of labour in order to explore these in more depth (Cheyne et al., 2004a, Cheyne et al., 2004b, Cheyne et al., 2006). Reported in this paper are the findings of one of these studies, the development and testing of an algorithm to assist with the diagnosis of active labour in primiparous women.