Are specially trained midwives the right professionals to perform the newborn physical examination?

Newborn infant physical examination

NIPE is a comprehensive clinical assessment of the newborn performed between 6 and 72 hours after birth, and is part of the routine infant screening programme offered universally in the UK, which also includes the initial examination of the newborn and the 6–8 week check performed by the GP (Townsend et al, 2004; Baston and Durward, 2010). While a normal NIPE is reassuring, it does not guarantee normality and has poor specificity: for example, approximately two thirds of neonatal heart problems are unidentifiable at the NIPE (Townsend et al, 2004). Traditionally, the NIPE was performed by a senior house officer (SHO) on paediatric rotation in the hospital or, less frequently, by a GP in the community (Clarke and Simms, 2012; Yearley et al, 2017). However, the extension of the role of the midwife has meant that midwives are the lead professional for low-risk women and neonates, and coordinators of care within the multidisciplinary team for those deemed high risk (Department of Health, 2010). As such, midwives' scope of practice encompasses fields from which they were previously excluded, and they may now receive additional training in such skills as ventouse extractions, sonography and NIPE (Townsend et al, 2004; Fryer and Evans, 2015; Rogers et al, 2017). In 1996, NIPE training for midwives was initiated. Justifications to extend the midwife's role to encompass NIPE include continuity of carer, autonomous midwifery practice, and reducing doctors' working hours (Mitchell, 2003a). NIPE has four key screening components—heart, eyes, hips and testes—and is a holistic examination that also encompasses overall assessment of the neonate, family history, parent education and discussion of parental concerns (Baston and Durward, 2010; Public Health England (PHE), 2016a).

Evidence-based practice

Evidence-based practice is central to woman-centred midwifery care (Cluett, 2005) and enables midwives to provide high-quality care to women and their support networks (Rees, 2011). The Trust in which the author of this article practises invokes guidelines that stipulate that the purpose of a NIPE undertaken by a midwife is to confirm normality and ensure appropriate referrals of abnormalities. Exclusion criteria exist that would preclude a midwife-led NIPE (Ironton, 2012). Such criteria are not standardised nationally, and PHE (2016b) stipulates that an appropriately trained professional must perform the NIPE, without elucidating exclusion criteria or circumstances when it may be inappropriate for certain professionals to do so. That said, it is important to recognise that NIPE is a screening tool and does not offer a diagnosis. It could be argued that during their care for the woman–infant dyad midwives are continually screening; for example, recording fundal height antenatally, fetal heart rate monitoring in the intrapartum period, the initial newborn check after birth and routine care in the postnatal period.

Seminal work

The Evaluation of the Midwife Role in the Examination of the Newborn (EMREN) study was a multifaceted study evaluating the implications of extending the midwife's role to incorporate NIPE, and was the first body of work to measure the appropriateness of this role (Townsend et al, 2004). This seminal piece remains one of a limited number of primary research pieces in the field. It was conducted by the Health Technology Assessment programme, whose research is used by such bodies as the UK National Screening Committee and the National Institute for Health and Care Excellence, which should give the peer-reviewed work credibility, reliability and validity.

The randomised controlled trial (RCT) (Bloomfield et al, 2003) that formed one arm of EMREN aimed to evaluate the quality of both physical and holistic components of NIPE carried out by midwives compared to SHOs. The examinations were recorded, randomised, and independently rated by one of two consultant paediatricians and one of two senior midwifery lecturers using a piloted and modified pro forma. Eleven NIPE-trained midwives and eight SHOs participated, each of whom performed two examinations. The authors concluded that where there was a significant difference between the quality of SHO and midwife examinations (Fisher's exact test P < 0.05) and good agreement between raters' analysis (kappa coefficient ≥ 0.4), the quality of NIPE performed by midwives was rated higher than those performed by the SHOs. Statistically significant differences (P < 0.05) in quality of examination of the heart and lungs, communication skills, soothing of neonates, and overall quality of examination, were observed in favour of NIPE performed by midwives. However, the agreement between raters was moderate–good (kappa coefficient ≥ 0.4) in only 48.2% of items, questioning the consensus between raters and the validity and reliability of results.

This RCT was actually a quasi-experiment: midwives and SHOs were randomised to one of two groups of raters, both of which contained one consultant and one midwifery lecturer; there was no control group; and such a limited dataset (n = 39) is not sufficient for an RCT. Therefore, the significance of findings must be questioned, particularly considering that a power calcu lation was not used to determine the number of parti cipants needed to test the hypothesis, and conclusions drawn must be viewed with caution. However, the ethics of conducting an RCT to test the hypothesis would have been problematic: withholding screening from a group of neonates, or offering a placebo, would contravene a fundamental principle of experimental research on people—that of non-maleficence (Steen and Roberts, 2011).

Audit of clinical effectiveness

Bloomfield et al (2003) did not demonstrate whether components of the NIPE, such as auscultating the heart, were carried out effectively and whether referrals were appropriate; clinical effectiveness was not assessed, but rather the raters' perceptions of quality of examination. Appropriate referrals were assessed in another arm of EMREN (Townsend et al, 2004), which concluded that the difference in appropriate referral rates between SHOs and midwives was not statistically significant, but did not specifically follow the neonates from the RCT. A considered point is that if the RCT (Bloomfield et al, 2003) had been designed as a prospective cohort study, it could have been supported by a clinical audit to ascertain whether midwives were clinically effective at detecting and referring abnormalities. Williamson et al (2005) piloted such an audit and collected retrospective data from one hospital in England over an 18-month period (2000–02) by tracking midwives' personal NIPE records against infant medical records to determine whether appropriate referrals had been made. Of 508 eligible infants, 482 were audited; attrition resulted from the examination being signed by an SHO (n = 2), birth at < 37/40 weeks' gestation (n = 2), and incorrect hospital number recorded in the midwife's personal notes (n = 22). Attrition owing to incorrect documentation calls into question the accuracy of NIPE documentation and results.

Clinical effectiveness was defined as the ability ‘to detect relevant family history, congenital abnormalities and make appropriate referrals' (Williamson et al, 2005: 117). An acknowledged limitation of NIPE is lack of specificity and inability to detect all abnormalities (Townsend et al, 2004; PHE, 2016b), thus it is questionable whether an SHO would have detected abnormalities if a midwife had been unable to do so, as those abnormalities may have been unidentifiable at the time of the examination. The audit assessed midwives' clinical effectiveness against appropriate referral rates that had been delineated in previous studies, and Williamson et al (2005) did not aim to compare clinical competence of midwives and SHOs, but rather to ascertain that midwives were clinically effective and able to detect abnormalities at an accepted rate. Bloomfield et al (2003), perhaps mistakenly, sought to compare different health professionals, and it may be more pertinent to test whether midwives meet nationally approved detection and referral rates that are set out by PHE (2016b) (Table 1). The authors concluded that the overall clinical effectiveness of midwives performing NIPE was 97.9%, and that while the number of midwives was limited (n = 8), the number of examinations (n = 482) was sufficient to ‘draw some inference’ (Williamson et al, 2005: 118), although use of a power calculation was not evidenced. Moreover, the rate of congenital abnormalities in the population was limited (Table 2) and it is questionable whether clinical efficacy could be measured if neonates presented with so few abnormalities.

Barriers to maintaining NIPE competency

A further study that appeared to offer an assessment of clinical effectiveness and quality of NIPE was conducted by Lanlehin et al (2011), although in reality the study focused on institutional and organisational barriers to health professionals undertaking NIPE post-qualification. Forty questionnaires were sent to health professionals, including 38 midwives, who undertook NIPE training at a London university during the period 2002–05. A response rate of 20% (n = 8) calls into question whether the results were generalisable to the original sample selected; responses may be biased towards health profes sionals who had had either a significantly positive or negative experience. That said, the findings supported those of previous studies, which suggested that a lack of managerial support and protected time in which to undertake NIPE impeded midwives' ability to maintain competency post-qualification (Steele, 2007; McDonald, 2008). While Lanlehin et al's study (2011) did not measure the appropriateness of midwives performing NIPE, a lack of post-qualification NIPE experience may have detrimentally affected clinical effectiveness and although it is the responsibility of the individual to maintain competency (Nursing and Midwifery Council (NMC), 2015), they cannot realis tically be expected to do so without workplace support.

Pushing boundaries

A more recent study has suggested that barriers to midwives undertaking NIPE have been overcome and that inclusion criteria allowing a midwife to undertake NIPE have been extended to include, for example, neonates born with light meconium or by elective caesarean section at term (McDonald et al, 2012). A grounded theory approach was adopted to explore midwives' attitudes to and experience of NIPE. This was an appropriate methodology as it focuses on the experiences of participants (n = 12 midwives and n = 5 heads of midwifery) from five NHS Trusts, and the limited cohort does not affect generalisability as qualitative research does not need to be generalisable. The authors used individual interviews with comparative analysis and theoretical sampling, and recorded contemporaneous descriptive field notes to aid understanding of the phenomenon by noting participants' body language, for example. Participants were able to review transcripts to ensure that they were representative of their experiences, which may have resulted in unreliable modifications that were not representative of their opinions at the time of interview. The authors concluded that by changing service provision to allocate appropriately trained midwives to run daily NIPE clinics on a rotation basis, NIPE was less likely to be performed by the midwife who facilitated the birth, thus jeopardising continuity of carer. Considering that NIPE is contraindicated within 6 hours of birth, the veracity of this statement may be queried because even if the midwife present at birth was NIPE-trained, care may have already been transferred to another midwife owing to shift change or transfer home or to a postnatal ward at the time of NIPE.

Changes in service provision

It was reported in the national survey arm of the EMREN study that only 2% of babies in the UK received NIPE undertaken by a midwife (Townsend et al, 2004). This increased to more than 50% of neonates born under midwifery-led care in over 70% of Trusts in 2014 (Table 3) (Rogers et al, 2015), which evidences a change in NIPE service provision. Results must be viewed with caution because, of the 154 UK heads of midwifery who were sent the 2014 survey, only 64.3% responded (n = 99/154), calling into question the generalisability when extrapolating results. Furthermore, of the 99 respondents, 91.9% (n = 91) provided figures for both total number of midwives (n = 18 050) and number of NIPE-qualified midwives (n = 2467) (Rogers et al, 2015). The majority of respondents (94.9%, n = 94/99) reported inclusion and exclusion criteria for midwife NIPE, although the method of data presentation meant that each criterion was reported as a percentage of Trusts that employed it; each Trust could tick as many criteria as applicable, thus the full criteria adopted by each Trust could not be visualised. Further clarification would have facilitated analysis of results. For example, 52.12% (n = 49/94) reported that suitability of the neonate to have NIPE undertaken by a midwife was decided by the midwife to perform the NIPE, yet it was not clear whether eligibility was checked according to local Trust guidelines or whether the midwife decided which neonates s/he felt competent to examine, independent of inclusion and exclusion criteria. The most significant limitation of the national survey is that it was not declared until the end of the paper that only estimated data were available regarding which health professional performed NIPE, and the percentage of NIPE performed by each different professional; therefore, the accuracy and robustness of the national survey findings must be viewed with caution.

Pre-registration NIPE training

In a recent national survey conducted by Yearley et al (2017), a questionnaire was distributed to each lead midwife for education (LME) in the UK and responses were received from 68.9% (n = 40/58) of the NMC-approved education institutions (AEIs). Ten AEIs provided pre-registration NIPE training, of which nine provided qualitative data outlining their rationale for doing so. Thematic analysis identified themes regarding NIPE qualification: consistency with the philosophy of midwifery; meeting service requirements; and provision of a responsive service (Yearley et al, 2017). It was highlighted that not all AEIs would be able to maintain NIPE competency in midwifery lecturers, and that Trusts may experience a shortage of midwives who are NIPE-trained to supervise students. This could hinder students' ability to successfully complete NIPE training and achieve competency, and the authors noted that there was a distinct variation in numbers of NIPE required by each AEI, which varied from no set number to 50 examinations (Yearley et al, 2017). The quality of NIPE training and students' competence should be questioned when national standardisation of NIPE training does not exist. Research exploring student midwives' subjective experience of undertaking NIPE training might be beneficial in ascertaining whether NIPE competency is a realistic expectation for pre-registrants, and whether adequate support is in place post-qualification to enable preceptors to maintain competency.

Future practice: NIPE SMART

NIPE SMART (screening, management and reporting tool) has been designed to standardise NIPE examination, referral and documentation, and will enable comprehensive and systematic audit of clinical effectiveness once implemented nationally. As of October 2016, 111/139 NHS maternity Trusts in the UK were using the NIPE SMART system, and it continues to be rolled out nationally (Walker, 2016). The tool will not measure the clinical effectiveness of midwives compared to other health professionals, but rather their clinical effectiveness according to national standards delineated by PHE (2016b). That said, while clinical effectiveness will be measured, the holistic elements of NIPE will not: it will not be possible to assess how well health professionals communicate with parents or whether parental concerns are discussed.

Conclusions

While literature concludes that specially trained midwives are the appropriate professionals to perform NIPE on low-risk neonates, the quality and robustness of research is to be questioned and, as such, findings must be viewed with caution. As the lead practitioner in the care of low-risk women and neonates, and coordinator of care for those deemed high-risk, the midwife's role continues to expand and service provision reflects that. An increased number of midwives perform NIPE routinely, and a higher percentage of NIPE is performed by midwives, suggesting that institutional barriers have been overcome. The addition of NIPE training to some pre-registration programmes may further increase the coverage of midwifery provision of NIPE, yet the impact on student experience should be considered.

The difference in quality of NIPE performed by midwives compared with other health professionals is not relevant as long as midwives meet NIPE standards set out by PHE, and the role of the midwife is not to make a diagnosis but rather to recognise deviations from ‘normal’ and make referrals accordingly. It is important to reiterate that NIPE is a screening tool. While it is not ethical to conduct RCTs to determine which neonates are suitable to have NIPE undertaken by a midwife, more than two decades after the publication of the EMREN study (Townsend et al, 2004) in which authors concluded that standardisation of eligibility criteria is necessary, this is still to be rectified. Midwives screen women and neonates throughout the antenatal, intrapartum and postnatal periods, thus it could be argued that midwives who have completed additional training to undertake NIPE are competent in assessing the additional clinical indices for all neonates that are healthy enough to be cared for outside of the neonatal unit.

Key Points