Revising acute care systems and processes to improve breastfeeding and maternal postnatal health: a pre and post intervention study in one English maternity unit

Each year, the majority of the three-quarters of a million women who have a baby in the United Kingdom (UK) are admitted to hospital for the birth, following which they are transferred to a postnatal ward prior to discharge home. For many women, this will be their first experience as a hospital in-patient [1]. Care on the postnatal ward is led by midwives, supported by maternity care workers who may or may not have undergone a period of training [2]. Despite evidence of widespread and persistent maternal morbidity [3, 4], there has been limited revision to the organisation, timing and content of postnatal care which continues to focus on a small number of physical health outcomes, for example vaginal blood loss and assessment of uterine involution. Over the last two decades, in addition to higher patient turnover intervals [5, 6] the number of women in the UK who have a caesarean birth has increased [7], outcomes of which are associated with higher risk of severe maternal morbidity relative to vaginal birth [8]. Although caesarean section (CS) is classed as major abdominal surgery, the content of routine postnatal care is unlikely to incorporate specific aspects of post-operative care, with limited guidance on wound management and effective pain management [9, 10]. Rates of severe morbidity in general are also increasing [11], with evidence of more complex health needs among women who become pregnant leading to poor pregnancy outcome [12].

Breastfeeding rates in the UK are among the lowest in Europe. Early results from the Infant Feeding Survey 2010 [13] showed that overall initiation rates in the UK increased from 76% in 2005 to 81% in 2010 (for England, rates were 78% and 83% respectively) however, the definition of breastfeeding used in the survey is contentious with data on women who only ever put their baby to the breast on one occasion included [14]. Data on the duration of feeding from the 2010 survey were not available at the time of writing, although the previous survey found a dramatic reduction in the proportion of UK women still exclusively breastfeeding at one week post birth (45%) and less than 1% of women exclusively breastfeeding at six months.

The National Institute for Health and Clinical Excellence (NICE) issued a guideline in 2006 to inform routine postnatal care within the National Health Service (NHS) in England and Wales for women and their babies. Guideline recommendations included the need for individualised care, tailored to the needs of each woman and informed by her pregnancy and birth history, with evidence based recommendations to inform maternal and infant health, infant feeding and the planning and organisation of care. Data from a recent survey of first-time mothers found many women were not receiving care as recommended by NICE [15] with the potential that postnatal care standards in England declined during the last decade [16].

Following an increase in complaints from women about the quality of inpatient postnatal care, including lack of support for breastfeeding, senior clinical managers at one large maternity unit in the south of England aimed to improve the provision of routine postnatal care, supported by implementation of NICE guidance [15]. To address how systems and processes across the organisation could be revised to promote and sustain a ‘seamless’ transfer of women from birth to discharge home and identify barriers and facilitators to enhance postnatal care, a decision was made by the multi professional team (which included the authors) to adapt a model of continuous quality improvement (CQI). The CQI model views quality improvement as an ongoing activity integrated within the organisation and focused on activities determined by those seeking to inform change. It is an approach which emphasises the role of senior management engagement with project teams, and importance of measurement [17], with many CQI interventions reflecting the Plan Do Study Act (PDSA) cycle.

As it became apparent that changes to support evidence based postnatal care needed to be made across the organisation, the importance of engaging as many staff as possible was recognised. Results from a pre and post implementation survey of women at 10 days and 3 months to collate data to compare breast feeding, maternal health outcomes and experiences of care are presented here. The study was funded by The Burdett Trust for Nursing.

Of 1160 women recruited pre intervention, 741 (64%) returned a completed questionnaire at 10 days and 616 (54%) returned a completed questionnaire at three months. Post intervention, 725 (63%) of 1153 women recruited returned a 10 day questionnaire and 575 (50%) returned a three month questionnaire. There were no differences between the two time periods and maternal characteristics (Table 1) and no differences in maternal or obstetric characteristics of women who were not recruited during the study period compared with those who were. The majority of women were white European, including a large number of women from Eastern Europe with Polish the second most common language spoken by women. There was a significant increase in the number of births by caesarean section pre and post intervention, with a concurrent decrease in instrumental births.

Using a CQI approach to identify where change could be introduced into routine systems and processes, it was possible to improve breastfeeding initiation and duration, some maternal physical health outcomes and enhance women’s views of their inpatient care. This is the first study to have addressed how postnatal services in an acute UK maternity setting could be revised in line with evidence.

Improving breastfeeding initiation and the duration of exclusive breastfeeding are important public health priorities for the UK [22], given shorter and longer-term health benefits for women and their infants [23‐25]. It was reassuring that the QI initiative impacted positively on breastfeeding. At the time of the study, the unit was not Baby Friendly accredited or working to obtain a certificate of commitment. A recent systematic review of structured compared with non structured programmes to support the initiation and duration of breastfeeding found that although most included studies reported statistically significant increases in breastfeeding initiation in hospital settings, studies were methodologically poor, effect sizes differed and not all found statistically significant differences in breastfeeding duration [26].

There are a number of reasons to support why our multi-faceted intervention made a difference, including an approach which supported ‘doing the right thing at the right time’. An important factor was recognition of the need to target the content and timing of care and provision of information offered as a continuum across pregnancy, birth and the postnatal period. Support for breastfeeding commenced with antenatal information, with subsequent clinical care and processes to support breastfeeding promoted immediately following birth, with midwives on the labour ward asked not to transfer women for at least two hours following the birth to enable women to have longer ‘quiet’ time with their babies to initiate skin to skin care. The 2005 UK Infant Feeding Survey found breast feeding initiation higher for babies exposed to early skin to skin contact (79% immediately and 87% within an hour, compared with 57% of babies with no such contact) and initial incidence of skin to skin contact correlated with breastfeeding prevalence at one and two weeks post birth [14]. In line with NICE guidance [15] the new postnatal record prompted the early identification and management of breastfeeding issues as part of in-patient care. A combination of these factors could have impacted on women’s confidence to breastfeed.

Mode of birth was adjusted for within the analysis, although it was of note that there was a large increase in CS births pre and post intervention. CS births can negatively impact on breastfeeding initiation, with evidence that women who have had operative birth are less likely to commence breastfeeding [27, 28], but are no more likely to discontinue breastfeeding than women who had a vaginal birth [14, 15].

The NICE postnatal care guideline [15] recommended that all maternity care providers in England and Wales should implement an externally evaluated structured programme that encourages breastfeeding, using the UK Baby Friendly Hospital Initiative as a minimum standard. Currently, evidence of which type of structured support is more likely to increase the initiation and duration of breastfeeding in England and Wales is unclear. Although more women were exclusively breastfeeding at three months post intervention, this was not statistically significant. Nevertheless the revisions to care reflected ‘best practice’ within routine service provision and could be a useful approach to adopt elsewhere given the increased demands on finite health service resources.

The inclusion of a symptom checklist in the new postnatal records appeared to impact on early identification of physical health outcomes, which could then be managed by the midwife or referred as appropriate. Women post-intervention were less likely to report offensive vaginal loss within the first 10 days of the birth, an important indicator of potential serious maternal infection [15]. Use of Modified Obstetric Early Warning Scores (MEWS) have been recommended by CMACE [12, 29] to detect early onset of severe maternal morbidity, although evidence to support the effectiveness of these tools in obstetric populations is not currently available. Data on use of the MEWS were not evaluated as part of this study, and it is not possible to assess if use prompted identification of potentially serious morbidity. As the MEWS tool in use did not include ‘triggers’ of specific signs and symptoms of severe postnatal morbidity, for example, temperature, lower abdominal pain or increased vaginal blood loss, it is more likely that use of the symptom checklist prompted identification. As other significant differences in reported physical morbidity, including backache (in hospital), constipation, difficulty passing urine and mastitis (within first 10 days at home) were also found post-intervention, would confirm this.

At three months, it is less clear that the QI intervention impacted on longer-term physical health, with evidence that resolution of physical health problems, even with directed community based midwifery interventions and longer-term follow up, may be difficult to achieve. MacArthur et al. [30, 31] conducted a large cluster randomised trial of a new model of midwifery-led postnatal community care, which extended midwifery contact to 28 days, with a final midwife contact at 10–12 weeks which replaced the six – eight week appointment with the family doctor. Primary trial outcomes at 4 and 12 months post-birth included physical and mental health and well-being as assessed using the Physical and Mental Health Components scores of the SF36 (PCS and MCS) and the EPDS. There were no differences in PCS scores at 4 and 12 months post-birth. Secondary trial outcomes included physical morbidities at 12 months post-birth, with no statistically significant differences in outcomes other than fatigue and haemorrhoids. In the current study most women would have been discharged by their midwife at around 10 – 14 days post birth, with no further follow up of their maternity care until the 6 – 8 week appointment with the family doctor. Our findings suggest that some symptoms are amenable to early midwifery intervention, with a need to promote ways in which longer-term physical morbidity can be addressed.

Unlike the current study, MacArthur et al. [30, 31] found significant differences in maternal mental health outcomes at 4 and 12 months post-birth following their community based intervention. We found no difference in EPDS scores at 3 months post-birth and no difference in emotional health needs in hospital, an area previously reported as a neglected aspect of women’s postnatal care [16, 32]. It is possible that the women in the current study did not consider emotional health a priority when a hospital in-patient, that in the short space of time prior to discharge they did not consider their feelings, or care to identify these outcomes was not sufficiently different pre and post intervention. Mental health disorders among women following birth are an important area for consideration, and the need to review strategies to prompt early identification and management is recognised [33, 34]. The trial by MacArthur et al. [30, 31] remains the only universal intervention to have impacted on maternal mental health outcomes and highlights the potential benefit of planned, effective midwifery care.

An earlier study from Melbourne, Australia [1] which compared women’s views pre and post intervention of a number of initiatives across a network of four public maternity hospitals, also attempted to improve outcomes of routine in-patient postnatal care. The intervention package included discussion visits with midwives in the third trimester of pregnancy, written information for service users, rotation of staff across intra-partum and postnatal clinical areas and introduction of evidence based guidelines and protocols. A four year evaluation based on outcomes from three units included postal surveys at three months post-birth of women at baseline (1922/1256, 65.3%) and post-implementation (1050/1829, 57.4%). Post-implementation there were statistically significant improvements in overall ratings of hospital postnatal care and level of advice offered. There were no differences in morbidity outcomes, but fewer women post intervention required help and advice on their own health at home suggesting strategies had benefited women’s health.

As part of our work, we asked women what information, support and advice they needed to better prepare them for going home [18]. Women were more positive about their experience, satisfaction with and planning of care post intervention, emphasising that service providers considering revisions to routine care should reflect what users consider important if positive feedback is to be achieved, including provision of practical demonstrations of infant care on the wards. There has been a major policy shift in the NHS to engage service users in decisions about their care [35] and a move for greater involvement of patients, carers and clinicians in setting priorities for research in areas where treatment uncertainties exist [36]. Service user engagement in planned service revision is clearly important for maternity service development within a QI initiative as women may highlight areas for change which clinicians and managers may not necessarily have deemed a priority. Midwives viewed overall changes as positive for the women and for their practice, although there was some concern that the new postnatal records took longer to complete as they were more comprehensive than the previous notes used at the study site [37].

The limitations of our study have to be considered, as well as the strengths. It was disappointing that there was some loss to follow up of women at three months despite sending two reminder questionnaires, however our response rates are comparable with similar postnatal studies [1]. The study design within a CQI framework meant that any potential confounding factors could not be controlled for, and influences on outcomes as a result of other changes across the organisation could not be accounted for. Data to inform economic analyses were also not planned as part of this work, although no additional resources were requested to implement the CQI intervention. That there could have been ‘knock-on’ effects from some of the revisions to care (for example, asking women to be kept on the labour ward for a longer period of time) cannot be discounted.

The study was not designed to specifically compare the occurrence of maternal health problems pre and post-intervention. A much larger sample would be required for this purpose. Nevertheless, that we had a positive impact on a number of important outcomes particularly when current staffing levels and skill mix within NHS maternity services are of concern [38] demonstrates that a CQI approach to revise routine systems and processes could support improved outcomes of routine care in other UK maternity units.

Achieving sustained change in an area of the maternity service where women may only be admitted for a few hours post-birth is an on-going challenge. Using a CQI approach provided an excellent opportunity to engage all relevant stakeholders, including women, to consider and identify where systems and processes could be revised within current resources to improve the quality and outcomes of in-patient care following birth. The improvement initiative was developed in response to women’s views of care and recognition of an ‘evidence to practice’ gap. Work to identify the issues, how these could be resolved and how best to implement and sustain change were important first steps before adopting a CQI approach to address the ‘knowing-doing’ gap [39]. We have shown that CQI can support implementation of evidence to enhance breastfeeding outcomes, reduce maternal morbidity and enhance women’s views of in-patient postnatal care and would encourage other maternity providers to consider a similar approach.