Attitudes towards the neurological examination in an unwell neonate: a mixed methods approach

Abstract

Background

The neurological examination of an unwell neonate can aid management, such as deciding if hypothermia treatment is warranted in hypoxic ischaemic encephalopathy or directing investigations in hypotonic neonates. Current standardised examinations are not designed for unwell or ventilated neonates, and it is unclear how confident paediatricians feel about the examination or what aspects they perform. 

Aim

This study aimed to review the confidence of UK paediatricians on the neurological examination in unwell neonates, describe their attitudes towards it, and determine what could improve practice.

Methods

An explanatory sequential mixed methods approach (QUAN → QUAL) with equal weighting between stages. A survey on attitudes to the neonatal neurological examination was sent to all UK neonatal units and members of the British Paediatric Neurology Association. Volunteers were sought for semi-structured interviews. Thematic analysis was used to interpret qualitative data, which was triangulated with quantitative questionnaire data.

Results

One hundred ninety-three surveys were returned, 31.0% from neonatologists, 9.3% paediatric neurologist. The median range for confidence was 4 (IQR3-5). Twenty-three interviews occurred. Thematic analysis revealed three themes: “Current culture on neonatal units”, “ Practicalities of the neurological examination in unwell neonates”, and “Changing the culture”. Most interviewees did not feel confident performing or interpreting the neurological examination in unwell neonates. Many units had a culture of seeing it as low priority, did not see its relevance in the acute management of unwell neonates. A few interviewees worked in units with a positive culture towards the neurological examination who used adapted standardised examinations and provided training. 72% of questionnaire responders wanted a new standardised neurological examination designed for the unwell neonate, which should be short, utilise pictures like the Hammersmith Neonatal Neurological Examination, contain an assessment of consciousness, be developmentally appropriate and achievable in unwell, ventilated neonates, be accompanied by a schematic to aid interpretation, and for greater training and assessments of competence.

Conclusions

There are barriers preventing paediatricians being able to perform a neurological examination in unwell neonates, and a culture of neurophobia is common. A new standardised examination is needed, alongside aids to interpretation, training, and assessment.

Introduction

After assessment and resuscitation, the starting point for any patient evaluation is history and examination. The examination of the neurological system is complex in children, requiring adaptation for their developmental abilities and behaviour [1]. The results can direct treatment; for example, in perinatal hypoxic ischaemic encephalopathy (HIE) the neurological examination determines whether hypothermia treatment is warranted or not [2, 3]. Following commencement of hypothermia, it also helps monitor change and informs prognostication. Outside of HIE, it determines the anatomical site of abnormal signs and supports formulation of differential diagnoses and management plans.

In the UK, postgraduate education begins with two years foundation level training, incorporating a variety of adult specialities and, rarely, paediatric posts. Paediatric training begins thereafter. The first 5 years are focussed on attaining generic core competencies, encompassing 3 years as a junior trainee (ST1-3) and 2 years as a senior trainee (ST4-5), with at least one year in tertiary neonatal services. During core training, paediatricians acquire competencies in examination in a variety of specialities, including general paediatrics, emergency medicine, neurology, community / neurodisability paediatrics, intensive care, and so forth. Thereafter, a trainee continues through 2–3 years of training in general paediatrics (ST6-8). Thus, all UK paediatric trainees should receive the same training on the neurological examination of a neonate, although there is no standardisation to this training.

There is little data on how confident and competent paediatricians are at performing the neurological examination in an unwell neonate. Previous work in our region showed trainees felt confident in perinatal HIE and reported they documented it thoroughly [4]. This was inconsistent with our experience. When asked to list what aspects of the examination trainees performed, they listed only cursory or limited aspects of the examination. At the time, reviewers concluded our results reflected poor training in our geographic area, but our experience was it reflected an attitudinal problem across the country, if not the world. If true, a thorough education programme is needed to improve patient examination and assessment, which could have benefits for patient care. This cannot be created without knowing what training is happening currently, what challenges paediatricians face with the neonatal neurological examination, and what tools are needed to assist them.

This study aimed to review UK paediatricians’ confidence about the neonatal neurological examination, describe their attitudes towards it, the challenges they face, and to ascertain what would improve practice.

Methods

We adopted an explanatory sequential mixed methods approach in two distinct phases (QUAN → QUAL) [5]. A questionnaire (Phase I) was distributed to examine health care professionals’ confidence in the neonatal neurological examination, followed by qualitative interviews (Phase II) to describe the reasons for these results. Equal weighting was given to both phases.

Phase I

In England, several specialities can be involved in the assessment of an unwell term neonate: neonatologists on Level 3 neonatal units, paediatricians on Level 1 and 2 neonatal units, paediatricians receiving referrals from the emergency room and primary care, and paediatric neurologists. Neonatologists and paediatricians are more likely to review acute emergencies where the diagnosis is clear, such as perinatal HIE, whilst neurologists are more likely to be determining the cause and severity of a neonate’s neurological state. Because the initial training of all paediatricians is the same, we hypothesised there would be homogeneity in views on training and performing the neonatal neurological examination. We therefore aimed to sample views from all these paediatricians.

We designed a questionnaire in paper and electronic versions (Supplementary material 1) based on our previous experience [4], and trialled it prior to use. The trial included asking 6 paediatricians (2 consultant paediatric neurologists, 1 speciality doctor in paediatric neurology and 3 paediatric trainees) to complete it either in paper or online format. They were asked to comment on the length of the questionnaire, whether the questions would give us information we wanted, and to identify any ambiguous or confusing language / questions. We rewrote problematic questions and improved the layout of the questionnaire based on these comments. The description of “sick” or “unwell” was outlined as a baby with illness like encephalopathy, weakness, or those who were ventilated / had umbilical or other lines in situ. This encompassed a large number of conditions depending on the specialist interest of the responder. Participants were asked to focus on the term neonate when answering, although they could also comment on preterm examination if they wanted. The questionnaire including questions with Likert scores ranging from 0 (never, not well / easy, or low confidence) to 6 (always, very well / easy, or high confidence) covering how often paediatricians perform the neurological examination, how easy they find performing it and its constituent parts, how well they interpreted its results and used them to make a management plan, and whether they used standardised neonatal neurological examinations or not. A free text box allowed responders to explain what challenges they faced when performing the neonatal neurological examination.

One hundred ninety-six Neonatal units were identified from a national website (ukntg.net/uk-neonatal-units). The clinical lead was asked if they and other staff members would complete the questionnaire. Members of the British Paediatric Neurology Association (BPNA) were asked to complete it via a monthly e-newsletter. Responders were asked to advertise the questionnaire to others. Responders were those who performed the neurological examination in clinical practice. This included Paediatricians and Advanced Neonatal Nurse Practitioners (ANNPs). ANNPs are highly trained nurses with additional qualifications and training, whose daily work is the same as paediatric trainees. We report frequencies, percentages, median, and interquartile ranges of responses. The free text comments were grouped into natural categories, and the frequency of responses falling within each category reported.

Connection of the two phases: the last question of the questionnaire asked for volunteers for a qualitative interview. Our inclusion criteria were paediatric doctors who assessed unwell neonates. We did not recruit Advanced Nurse Practitioners to qualitative interviews because their response rate to the questionnaires was low and recruitment would have been hard. We were also aware their training programme differed from doctors’, which may have led to greater variation in views and the need for a larger sample size. From the list of volunteers, a convenience sample was used to make recruitment easier; however, we selected a spread of grades of doctor and specialities. We did not expect gender to influence answers, but our previous experience was that male doctors were more likely to arrange interviews [6, 7], so we ensured a gender balance in this cohort. We selected interviewees from a wide geographical area of England to avoid results reflecting training issues specific to a particular region. No more than 3 participants were recruited from any single centre. Once we had identified initial participants from the questionnaire, we used a snowball technique, i.e. asking participants to recommend further potential participants, where specific demographics were underrepresented. The purpose of the interviews was to describe attitudes towards the neurological examination in the unwell neonate, how interviewees were trained, and how they reached confidence. We asked interviewees to focus particularly on the term neonate. The topic guide (Supplementary material 2) was influenced by the questionnaire and was trialled in 3 volunteers, and the questions and their order were changed to improve the interviews.

Phase II

A qualitative descriptive methodology was used. A single interviewer performed the semi-structured interviews (AF). The interviewer was a female paediatric neurology trainee. The interviewer may have known the interviewee previously and had preconceived views that could lead to bias. We reduced this risk through training and by ensuring the topic guide contained open questions. Interviews occurred at a time and location of the participants’ choice and in person/face-to-face, where possible. During the COVID-19 pandemic, interviews were performed virtually via video link. Written informed consent was obtained. The interviews were recorded digitally, transcribed verbatim by a medical secretary, and anonymised, with names replaced by participant numbers and working or training locations exchanged for “X”. One of the research team listened to the recorded interviews whilst reviewing the transcript to check it for accuracy. The researcher corrected typing errors. One participant, for whom English was not their first language, asked to review their transcript after it had been reviewed by the research team to ensure it accurately reflected their views.

Thematic analysis was performed as per Braun and Clarke (2006) [8], including familiarisation of data, coding using an inductive approach by two researchers (QC and ARH), review of initial codes, agreement on a coding structure, reflexive changes to coding structure as more data was analysed, and identification of a thematic structure. Themes were developed using an iterative process. With relation to researcher reflexivity, ARH was a consultant paediatric neurologist, who was aware he had preconceived ideas about the neonatal neurological examination in clinical practice. He ensured he questioned his assumptions regularly during the analytical process. QC was a non-medical research fellow with training in qualitative research, who was previously aware of the challenges faced in neonatal care from discussions with health care professionals. Both researchers reviewed and discussed their coding structure together repeatedly during data analysis and refined them after discussion. The themes were formulated independently, and then discussed and revised, with both researchers able to challenge the other’s assumptions.

We reached data saturation at 20 participants. Data saturation was defined as occurring when additional data did not significantly change or refine the coding structure, nor thematic framework, i.e.thematic exhaustion occurred [9, 10]. We recruited a further 3 to ensure no new codes or themes arose, and to ensure a fair balance of interviewee characteristics. The concept of data saturation is controversial, along with how many participants are needed to reach this [11, 12]. We thought our 23 participants was sufficient because our study had relatively narrow aims, i.e. to determine views on performing the neonatal neurological examination, and participants’ training of the neonatal neurological examination would have been similar. Furthermore, our interviews were relatively long, ranging from 55 min to over 120 min, contained rich in-depth data, and the same interview schedule was used for all interviews. NVivo for Mac version 12 (QSR International PTY Ltd, 2018) was used for analysis. The results of the quantitative and qualitative phases were integrated by determining the messages provided by both sections of the study and identifying areas in which they agreed or disagreed. Where the messages differed, we looked at our qualitative data to determine if there was an explanation. We also studied the results of our previous quantitative questionnaire [4], and other previously published data in this field to explain our results and to inform our discussion. Ethical approval was obtained from the Nottingham 1 Research Ethics Committee (IRAS 259,148) and informed consent obtained from all interviewees.

Results

Quantitative data

One hundred ninety-three questionnaires were returned with no duplicates. Responders worked in 60 units across the UK, although 98 responders did not report location. Multiple responses from single units were received, with the largest being our own: 6 from our Children’s Hospital and 3 from the Maternity Hospital. 60/193 (31.0%) responders worked in neonatology, 111 (57.6%) in paediatric specialities other than neurology, 18 (9.3%) in paediatric neurology, 4 (2.1%) in Paediatric Emergency Medicine / Anaesthesia. Ninety-two (47.7%) were consultants, 57 (29.5%) ST4-8 trainees, 25 (13.0%) ST1-3 trainees, 8 (4.1%) Advanced Neonatal Nurse Practitioners, and 11 (5.7%) other grades.

The results to questions on general attitude and practice relating to the neurological examination in the unwell neonate are summarised in table one. Some respondents did not answer all of the questions (see Tables 1 and 2). Variation in practice was noted with reference to the neurological examination in an unwell neonate: trainees and Advanced Neonatal Nurse Practitioners (ANNPs) reported they performed a neurological examination in around half of neonates, and consultants in most. Within the consultant responses ranges, some reported they hardly ever performed the examination. The median range for confidence was high (Table 1), although a wide range was seen. Responders reported that a high-quality documentation of a neurological examination was found in the medical records of an unwell term neonate around half of the time, with the range of scores extending to ‘never’. Neurologists scored this lower. A small proportion of responders routinely used a standardised neurological examination, including the Hammersmith Neonatal Neurological Examination (HNNE), to assess unwell neonates.

Table 1 Responses to the survey questions on attitudes to the neurological examination in an unwell neonate

Table 2 Responses to the survey questions confidence of specific aspects of the neurological examination in unwell neonates

Table two shows data on how easy responders thought individual aspects of the neurological examination in an unwell neonate were to perform. The following aspects were rated as a median of 5 or more by the whole cohort, indicating ease of examination: anterior fontanelle, pupillary size, and suck. The aspects considered the hardest to perform, defined as a median score of 2 or less were: fundal examination and cranial nerve function. Despite rating cranial nerve function as being hard, responders rated most individual aspects of cranial nerve function as being easier, including pupillary responses, vision, eye movements, facial expression, suck, and gag.

Responders were asked to document the challenges they faced when performing the examination in an unwell neonate. The most common answer was the effect of sedation and muscle relaxants on clinical signs (n = 61), followed the physical barrier provided by lines, ventilation tubes, incubators and other equipment and concern about dislodging them (n = 56). The other challenges were grouped into natural categories and the most frequent were:

• Cardiovascular instability or the baby needing ‘minimal handling’ (n = 25)

• Not knowing how to interpret the findings (20)

• Difficulty eliciting the abnormal signs (13)

• Absent or poor training on the neonatal neurological examination (11)

• Time constraints / the examination takes too long (12)

• A lack of experience or opportunity to practice the examination (12)

• Not knowing what to do (11)

• Lack of an appropriate standardised neonatal neurological examination for unwell neonates (10)

• Understanding normal from abnormal findings at different gestations in preterm neonates (10)

• Lack of confidence (9)

• Difficulty determining whether abnormal signs are a result of a primary neurological disorder or multisystem illness, like septicaemia (9)

• Stabilisation / other procedures take priority (8)

• Subjectivity and reproducibility (7)

• Equipment is unavailable, especially tendon hammers (7).

When responders were asked whether a new standardised neurological examination specifically designed for unwell neonates would be useful, 124/172 (72.1%) agreed, 39/172 (22.7%) were unsure and 9/172 (5.2%) thought it would not be useful.

Qualitative data

Twenty-three interviews were performed. Nine volunteers via questionnaire did not respond to invitations to arrange an interview, and 1 was unavailable. Two identified via snowballing did not respond to invitations and 1 cancelled owing to changing work patterns. No repeat interviews occurred. Fourteen interviewees were consultants, 7 ST4-8 trainees, and 2 other grades of doctors. Ten worked in neonatology, 7 in paediatrics, and 6 in paediatric neurology. Twelve were female. Six worked in Yorkshire and the Humber, 3 in North England, 3 East of England, 6 the Midlands, and 5 in London. The length of interviews ranged from 53–122 min, median 84 min. Three themes emerged (Fig. 1):

• 1. “Current culture on neonatal units”

• 2. “Practicalities of the neurological examination in unwell term neonates”.

• 3. “Changing the culture”

Fig. 1

Summary of main themes and subthemes

Theme 1: Current culture on neonatal units

Illustrative quotations are shown in Table 3. Neonatologists viewed the neurological examination through the lens of HIE, whilst neurologists focussed on the aetiology of neurological signs. Two quotations highlighted the perceived purpose of the neurological examination: “the history gives you the mechanism [of injury], and the examination gives you the site” and “The history tells you more than the examination. The examination is just to grade the severity of it”. The value of the examination formed the first subtheme, with two cultures adopted: low priority and important.

Table 3 Illustrative quotations for Theme 1—current culture on neonatal units

In the low priority culture, decisions on whether to start hypothermia therapy in HIE were obvious without an in-depth assessment of the baby’s neurology. Clinicians were focussed on resuscitation, stabilisation, and time-critical procedures, such as securing venous access. The need to obtain competencies in procedures, and the perception that procedures were “fun”, meant trainees prioritised these above the neurological examination. Some interviewees did not think the neurological examination gave them useful information at all, “Is me disturbing this child so much going to add to the clinical picture?”, and this was seen particularly in interviewees who considered the neurological examination time consuming. Other interviewees noted that, once resuscitation and stabilisation had occurred, the neonate may have had sedative or paralysing medications, which rendered the neurological examination useless. When they felt unable to assess the neonate neurologically, interviewees relied on information from other sources, such as blood gases, cranial ultrasound, and amplitude integrated electroencephalography (aEEG).

Other interviewees considered the neurological assessment important. In the context of HIE, they acknowledged the need for ‘fire-fighting’, i.e., focussing on emergency aspects of care, but felt the neurological assessment was important to justify decisions about hypothermia treatment and to provide a baseline against which change could be monitored. For these interviewees, the neurological examination was not time consuming, did not disturb a baby excessively, and interviewees thought aspects could be performed at the same time as procedures, such as assessing response to pain. In non-HIE cases, the neurological assessment was important to determine the neuroanatomical site of signs, form differential diagnoses, formulate management plans, and was described as “fun”.

The second subtheme outlined challenges to examination, including sedation, cardiovascular instability, negotiating arterial / venous catheters, ventilation, and finding a suitable time. The latter included disturbing the baby as little as possible by grouping tasks around the time of developmental cares. Other challenges related to understanding the findings of a neurological examination over the telephone, specifically discussions between units on the suitability of hypothermia therapy. The greatest challenge faced was that most interviewees did not know what to do to perform the examination or how to interpret the findings. There was no accepted structure to the examination, consultants and trainees had received no training in it during their career, they struggled to extrapolate adult-style examinations to the sick neonate, and they did not know how to document their findings. Trainees wanted more training, but the consultants themselves did not feel confident in the examination or communicating what they had done. Trainees rarely watched consultants examine and did not receive assessments on their technique. Some interviewees noted this was a problem in the whole of paediatric training and not just neonatal training.

The response to these challenges took two forms, forming a third subtheme. The first was avoidance, where the neurological examination was either not performed or a cursory examination was undertaken: the “AF [anterior fontanelle] normal, tone okay” phenomenon. Legitimate challenges to examination then became excuses, including projecting the need to coordinate examination with cares and procedures on nurses, who allegedly would not let doctors examine a baby. Others defended nursing staff from these accusations and noted the nurses allowed examinations if there was good communication on why it was needed at that time. Some health care professionals thought a neurological examination was pointless in a child who had received sedation, or could not be done owing to lines and ventilation. As a result of avoiding examination, interviewees gave examples of missed or delayed diagnoses.

A smaller proportion of interviewees did not accept these excuses and sought to overcome challenges. They would perform a limited examination, working around lines and equipment, and noting the importance of serial examinations in sedated babies to monitor change. They had typically received training outside the UK or had self-taught themselves, adapting their style over years. One had created proformas for the examination and introduced them into their unit or network to promote better examination and communication. Two interviewees worked in units with a focus on neurological care that used standardised examinations, such as the HNNE [13,14,15], and provided teaching for trainees. Whichever response to the challenges was adopted, this became the culture embedded in local practice.

Theme 2: The practicalities of the neurological examination

Illustrative quotations are presented in Table 4. The first subtheme comprised attitudes towards different aspects of the neonatal neurological examination, which broadly fell into two categories: the achievable or important, and the impossible or unimportant.

Table 4 Illustrative quotations for Theme 2 – practicalities of the neurological examination in unwell term neonates

The assessment of neonatal conscious level was considered important and an area participants spoke about at length. In the context of perinatal HIE, the assessment of consciousness determined the degree of encephalopathy and suitability for therapeutic hypothermia. Some participants reported it was obvious if a neonate was alert or comatose and the decision about therapeutic hypothermia was easy. Other participants noted that some neonates fell into a “grey zone”and this was difficult to quantify. No interviewee used a formal consciousness score system, although some reported they used a broad method of categorisation similar to the AVPU system (Alert – responds to Voice – responds to Pain – Unresponsive) advocated in the Advanced Paediatric Life Support course [16]. The disadvantage of the AVPU scale was its insensitivity in detecting subtle changes over time.

No interviewee used the Glasgow Coma Scale (GCS). Participants thought the GCS was not developmentally appropriate, too detailed, and did not provide useful clinical information. When asked how they assessed consciousness in a neonate, participants subconsciously adopted the same categories as the GCS and made it developmentally appropriate. They were surprised, when faced with a chart showing the modified GCS for children, how much of the scale they adopted, but they did not formally score their findings. Without a score, interviewees found it hard to quantify the degree of consciousness, document it, and explain their findings to others. This subjectivity meant it was impossible to assess consciousness serially over time by different team members. Participants noted that nurses and parents were better at detecting subtle changes over time because of the consistency and regularity in who was assessing the baby and they could note when they were “quiet”. This suggests some quantification of consciousness is possible in neonates. After reviewing the GCS and its scoring, participants thought a modified score could capture what the nurses and parents were detecting instinctively, although they were unsure whether they would use it without evidence to show it identified deteriorating babies or provided prognostic information. A small proportion of participants were uncomfortable causing pain in babies during an assessment of consciousness, whilst others saw this as being an important component that could be performed at the same time as other routine painful procedures. Two interviewees noted there had been a recent recommendation for including serial monitoring of consciousness in neonates who had fallen or been dropped on the postnatal wards [17], and the development on a new neonatal coma scale would support this.

Determination of limb muscle tone was considered both achievable and important. The assessment of power was also thought important, although participants noted trainees experienced difficulties in differentiating tone from power and the Medical Research Council (MRC) muscle grading system [18] was inappropriate for neonates. Power was generally assessed instead by observing the presence or absence of antigravity movements and resistance to procedures and examinations. One participant discussed the relationship between conscious level and assessment of power, noting a neonate had to be alert to assess power. Participants felt the assessment of the fontanelle was achievable, but a small number questioned what useful information it gave; it was generally used to diagnose raised intracranial pressure and one participant objected to a “sunken” fontanelle being assumed to correlate with dehydration. Logistically, the fontanelle examination was difficult to perform where apparatus or hats were attached to the neonate’s head. Posture, quantity, and quality of spontaneous movements were also considered important, but assessing quality of movements was felt to require considerable experience. Head circumference, rooting, grasp, and plantar reflexes were also considered possible and important.

Features of the neurological examination that fell within the impossible or unimportant category included head lag and ventral suspension, primitive reflexes, such as the parachute reflex, the assessment of cry, which was either impossible in a ventilated neonate or too subjective, and examination of sensation outside the context of a spinal lesion. There was a strong culture that the Moro reflex was a vital component of the standard neonatal neurological examination, although no participant clarified what useful clinical information it gave. It was considered dangerous in an unwell / ventilated neonate. Participants’ opinions on deep tendon reflexes were divided. Some thought they were important, but examination was hampered by the lack of available tendon hammers and the practice of using a stethoscope instead, whilst others thought they were too hard to obtain and did not offer useful information. Participants’ answers about the cranial nerve examination were interesting: non-neurologist participants were resistant to performing the cranial nerve examination and thought it was too hard or did not provide useful information. However, when faced with individual components, such as pupillary responses, eye movements, facial expression, suck and gag, the same participants reported these aspects were important and easy to perform. Neurology participants saw the cranial nerve examination as both possible and important.

The second subtheme related to participants’ views on standardised neonatal neurological examinations, specifically the HNNE. The participants who worked in units with an interest in neonatal neurology used this tool and adapted it to the clinical condition of the baby. Others had experience of it in the past, such as during training to estimate the gestation of a preterm baby but did not use it routinely on their neonatal unit. Participants reported it was too long and repetitious, the relevance and importance of several signs were unclear, it required significant training, and it included aspects that were impractical in unwell neonates. Participants felt the pictures on the HNNE proforma were helpful, which enabled them to know how to perform the parts of the examination and to indicate results quickly by circling without the need for writing. However, participants thought the proforma was too cramped and busy.

Theme 3: changing the culture

Illustrative quotations are shown in Table 5. A culture change was thought necessary: “There should be a change in culture in the neonatal units until it becomes the norm.” This theme outlined recommended steps to achieve this. The first subtheme was elucidation: convincing health care professionals the neurological examination could improve care. The second was development, summarised by one participant as “I think it is about time we had some sort of good clinical examination”. Participants thought a new simplified standardised neurological examination would improve the quality of neurological examination, communication of findings, documentation, objectivity, and monitoring of change over time. Any new examination needed to be quick, simple, feasible, and reliant mainly on observation. To aid with interpretation, interviewees recommended a schematic or flowchart. Views on scoring systems were mixed, with some considering they would help monitoring over time, and others feeling the focus would be on the score and not the meaning of the findings. One neurology consultant did not see the value of a new examination, although reflected that “there’s always scope for improvement and one probably doesn’t realise it till it actually happens.” probably because they felt their skills were good and did not see the neonatologists’ perspective. Once a suitable examination was created, interviewees recommended teaching courses and videos, with an emphasis on the neurology examination, in the paediatric under- and postgraduate curriculum, which formed the third subtheme. Assessments were thought important to demonstrate competency. The final subtheme was that interviewees thought the examination should be embedded in practice by consultants and through guidelines and research protocols.

Table 5 Illustrative quotations for Theme 3 – changing the culture

Discussion

In 2017, ST4-8 paediatric trainees in our region reported they felt confident in performing the neurological examination in neonates with HIE [4], and the data from our national survey showed comparable results. In comparison, our interviewees lacked confidence. At face value, this data appears at odds with each other, but the discrepancy can be explained. Two observations suggest the confidence levels in our preliminary data were likely suggestive of over-confidence: the high frequency trainees claimed they documented the neurological examination in the medical notes and the limited aspects of the examination they said they performed [4]. In this study, responders to the questionnaire also reported a detailed neurological examination was found in the medical notes of half of unwell neonates, which was also contrary to both the authors’ and interviewees’ experience. The paediatric neurology responders reported lower rates, probably reflecting differences in what they thought constituted a “high-quality examination”. Another observation was that standardised neurological examinations were rarely used by participants in either our questionnaire or interviews, with most adapting the neurological examination for older child or adult to the unwell neonate. However, interviewees frequently stated they did not know what aspects of the examination they should be doing, suggesting variation in what is done. In this situation, paediatricians can feel confident in their practice without being competent, especially if not trained, appraised, or challenged by those with more experience in neurological examination.

There was also a pervasive attitude that the neurological examination was unimportant and yielded little useful clinical information in unwell neonates, particularly amongst those trained in the UK. When the examination was seen through the prism of perinatal HIE, a detailed neurological examination did not add useful information when care was focussed on stabilisation. At these times, it does not matter what parts of the examination are done, or if it is cursory, clinicians can feel confident they have obtained all the information they need. However, when faced with an interviewer with neurological expertise and questions focussed on interpretation and aetiology of signs, confidence fell. Interviewees admitted trepidation in performing and interpreting the neurological examination in an unwell neonate, which correlated with the lower scores for the interpretation of signs in our questionnaires. Further evidence that reported levels of confidence were not genuine was the fear we found around certain aspects of the neurological examination, particularly the cranial nerve examination. In both the questionnaire and interviews, participants reported the cranial nerve examination was hard to perform but reported the individual aspects of it were “easy”. This suggests paediatricians see components of the neurological examination as part of a holistic assessment of the baby without considering their neuroanatomical significance. Thereafter, the concept of “formal” neurological examination in a neonate becomes intimidating, and a culture of avoidance results.

Fear and avoidance of neurological assessment amongst health care professionals is not a new observation. “Neurophobia”, where neurology is feared and perceived to be the most difficult clinical speciality, was first described in 1994 [19] and has since been found by others, including in the UK [20,21,22,23,24,25]. In paediatrics, this is likely to be compounded by additional developmental and behavioural challenges [1]. In unwell term neonates, these challenges are further magnified by the emergency nature of many conditions, cardiovascular instability, lines and intubation tubing, neonates’ even more limited developmental abilities, and the practice of considering consciousness only in broad categories [26, 27] or not at all. The latter may explain why interviewees noted the assessment of tone and power could be easily confused by trainees: health care professionals caring for older patients would not try to assess MRC grades of power in a comatose patient who is unable to perform the motor tasks asked of them, and so the assessment of power in neonates needs to be linked more closely to consciousness.

Neurophobia does not need be a part of neonatal care. Several interviewees had gained additional training outside the UK, where there was a more positive attitude and training towards the neurological examination. Some UK units had adapted standardised examinations to the unwell neonate. These interviewees valued the information it gave them, particularly during serial monitoring. Proactive training had led to a more positive culture on their unit. This leads to the question on whether standardised examinations, like the HNNE, should be introduced across all neonatal units. There was a lack of enthusiasm for this amongst interviewees, who thought the HNNE was not designed for unstable neonates, its focus was more directed towards detecting abnormality rather than interpretation of signs, it was too long and intimidating, and the proforma cramped and busy. Interviewees and questionnaire responders wanted a new, fit-for-purpose standardised neurological examination for unwell neonates.

A number of recommendations can be made from our data:

• 1. Development of a neonatal consciousness score – this should be developmentally appropriate, utilising the broad categories of the GCS, have a clear scoring structure, and be objective and useful in clinical practice

• 2. Development of a standardised neurological examination for unwell term neonates which is short, safe in ventilated neonates, includes only relevant components, and makes it clear which neuroanatomical site is being examined

• 3. Creation of an interpretation aid to make interpretation of signs, formulation of differential diagnoses and management plans easier

• 4. Improved training and education of all grades of medical staff on the neurological examination with elucidation of why it is useful

• 5. Assessment of competencies once new examination tools are introduced into clinical practice to ensure trainees know how to perform and interpret it

• 6. Research into the new tools to demonstrate their clinical utility and expand their use.

There are limitations to our study. These include that we cannot be sure the results reflect the full range of views amongst paediatricians. Certain attitudes may have led participants to complete the survey or interview, and the snowball technique may identify participants with similar attitudes. The results will also reflect the authors’ experiences to some degree, which is a well-known aspect of qualitative research. Similarly, paediatric training in other health care settings may be different than in the UK and our data may not be generalisable in different contexts.

In conclusion, although confidence levels of performing a neurological examination in an unwell neonate are reported to be high, this confidence does not appear to be genuine. Paediatricians do not know how to perform a high quality neurological examination in unwell neoantes or how to interpret the signs. There is a culture of neurophobia in UK paediatric services, and the examination has secondarily become unimportant and is avoided. Poor training has contributed to this phenomenon. A small number of units have sought solutions to these problems, adapted standardised examinations, and organised training, which has led to a more positive culture. However, our interviewees did not think the current standardised neurological examinations were fit for purpose. A change in the culture is desired, which would start by the development of simple standardised examinations of unwell neonates and a neonatal coma score, alongside training of why it is important, how to perform and interpret the results, and formal assessments of competency.