Article Text

Download PDFPDF

Missed opportunities for earlier treatment? A qualitative interview study with parents of children admitted to hospital with serious respiratory tract infections
  1. Nick A Francis1,2,
  2. Joanna C Crocker1,2,
  3. Arla Gamper3,
  4. Lucy Brookes-Howell1,2,
  5. Colin Powell4,
  6. Christopher C Butler1
  1. 1Department of Primary Care and Public Health, School of Medicine, Cardiff University, Cardiff, UK
  2. 2South East Wales Trials Unit, School of Medicine, Cardiff University, Cardiff, UK
  3. 3School of Medicine, Cardiff University, Cardiff, UK
  4. 4Children's Hospital for Wales, Department of Child Health, School of Medicine, Cardiff University, Cardiff, UK
  1. Correspondence to Dr Nick A Francis, Department of Primary Care and Public Health, School of Medicine, Cardiff University, Neuadd Meirionnydd, Heath Park, Cardiff CF14 4XN, UK; francisna{at}


Objective To identify potential opportunities for earlier intervention among children who develop a complicated respiratory tract infection (RTI).

Design Qualitative, face-to-face, individual interview study, either in hospital or at home, with parents of children admitted to hospital with a complicated RTI.

Setting Participants were recruited from a large UK teaching hospital, and described events (largely) prior to hospital admission.

Participants Parents of 22 children (12 with empyema, 8 with pneumonia, 1 with peritonsillar abscess and 1 with mastoiditis).

Results Parents' accounts revealed missed opportunities for timely treatment resulting from parental and health service associated factors. Themes relating to parental factors included problems assessing the severity of the illness (5 parents), beliefs about accessing health services (10 parents; including fear of appearing ‘neurotic’, belief that their concerns would not be taken seriously, and belief that their child would not be prescribed antibiotics or would be prescribed antibiotics too readily) and feeling powerless to challenge clinical authority (7 parents). Health service associated factors included perceived problems accessing healthcare services (13 parents; including inadequate primary care triage, barriers to accessing timely consultations and past experience of problems accessing healthcare leading to failure to consult) and perceived poor quality clinical encounters (11 parents; including inadequate assessment and communication).

Conclusion Addressing, where appropriate, these parental (skills, fears and beliefs) and health service (access and consultation quality) associated factors may lead to more prompt care for seriously ill children.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Several countries have reported an increase in the incidence of empyema since the late 1990s,1,,5 which has continued in the USA despite the introduction of pneumococcal conjugate vaccines.6 7 The incidence of pneumonia has not shown similar widespread increases. In the USA, there was a reduction in the incidence of pneumonia in children under 1–2 years of age and a stable rate in older children between the late 1990s and mid 2000s.8,,10 In the UK, a Scottish study found a relatively constant incidence of pneumonia between 1981 and 2005 in most children, but an increase in the 1–4-year-old age group.2 The reasons for the increasing incidence of empyema remain unclear. Retrospective analyses of routinely collected data have been used to demonstrate an association between lower antibiotic prescribing rates and increased rates of mastoiditis11,,13 and peritonsillar abscess,11 and these support a common concern among some that reductions in primary care antibiotic prescribing for common infections may be exposing patients to greater risk of complications. Recent primary care studies suggest an association between antibiotic prescribing and rates of mastoiditis, peritonsillar abscess and pneumonia, but more than 4000 patients need to be treated to prevent each complication.13 14 However, these studies do not provide evidence that efforts to reduce overall antibiotic prescribing are causing increased complications, and cannot be used to identify whether there are missed opportunities for earlier intervention in those who develop complications.

What is already known on this topic

  • Observational studies have demonstrated an increase in the incidence of serious respiratory tract infections (empyema and mastoiditis) in children over the past decade.

  • The causes of these increases are not clear, and there is little evidence on whether there are remediable factors delaying early treatment (especially early antibiotic treatment) in these patients.

What this study adds

  • This qualitative study identified problems in the interpretation of symptoms by parents, delays in consulting resulting from parental beliefs about accessing healthcare, and parents feeling powerless to challenge clinical authority.

  • Parents also perceived problems accessing healthcare and poor quality clinical assessment and communication in primary and secondary care consultations.

A confidential enquiry into child deaths in England and Wales included some data on deaths from acute infections,15 and a case–control study in Mexico explored the role of primary care processes in infant deaths from acute respiratory tract infections (RTIs).16 However, these data may not be relevant to complicated RTIs in the UK. No previous study has examined parental accounts to describe factors that may have delayed appropriate care for children developing complicated RTIs. We set out to examine parental accounts of the illness course, help-seeking behaviour and interactions with healthcare services of children admitted to hospital with complicated RTIs in order to identify barriers and potential missed opportunities for early intervention.


Patients and procedures

Paediatric nurse practitioners and clinical staff on paediatric and ear, nose and throat teams in a large teaching hospital (University Hospital of Wales) were asked to identify children aged 6 months to 16 years admitted with pneumonia, empyema, mastoiditis, peritonsillar abscess or acute rheumatic fever. The nurse practitioners obtained informed consent from the parents of eligible children willing to participate, and a researcher then contacted these parents to arrange a semi-structured interview.


Interviews took place either in a quiet room on the ward or at the parents' home. Children old enough to participate, and interested in doing so, were encouraged to take part in the interview along with their parent(s). Interviews were conducted by one of three interviewers (two general practitioners (GPs) (NAF, Naomi Cadbury) and one research assistant (Jenny Shearman)). All interviewers had experience in conducting qualitative interviews but did not receive specific training for this study. An interview guide based on the study aims and a review of research evidence on acute infections in children was developed by the study team and used to guide all interviews. The guide comprised open questions followed up by specific prompts, and prompted discussion of the time-course and features of the illness, all consultations with healthcare providers prior to admission, sources of information used (family, friends, etc), use of antibiotics prior to admission, the child's past medical history and the family's social circumstances. Interviews were not constrained by topics in the guide and participants were encouraged to elaborate on any of the topics or related topics they thought broadly relevant to the study. All interviews were audio recorded and transcribed verbatim with personal details omitted. Interviewers also recorded field notes during or shortly after the interviews. Interviews were conducted, transcribed and coded concurrently. Recruitment continued until no new ideas were emerging (data saturation).


Data were analysed using framework analysis.17 This approach involves the following steps: familiarisation, creating a thematic framework, indexing, charting, mapping and interpretation. The thematic framework was developed iteratively and was informed by the data and agreed during study team meetings. NVivo 8 software was used for the indexing stage. Three researchers (NAF, JCC, AG) conducted the analysis, and eight (36%) transcripts were indexed by more than one researcher (six by NAF and JCC, two by NAF and AG). Participating children made only limited contributions, largely correcting or clarifying details provided by their parent(s), and therefore their data were not analysed separately from the parent data. Emerging themes related to potential missed opportunities for earlier treatment identified in participants' accounts of their child's illness. Judgements made by researchers were noted. Disagreements about coding were resolved, where necessary, in a series of study group meetings.

Ethics approval

The study was approved by the South East Wales Local Research Ethics Committee Panel B (05/WSE02/186).


Participants and consultations prior to admission

The parents of 22 children were recruited and interviewed (table 1). Fourteen interviews were with only the child's mother, five with the child and mother, and three with the child's mother and father. Sixteen interviews took place at the family home and six on the hospital ward. Twelve of the children were female, the median age was 4 years (IQR 2.25–9), and 12 had empyema, eight pneumonia, one peritonsillar abscess and one mastoiditis and lateral sinus thrombosis. Median symptom duration prior to admission was 7 days (range 2–49 days). Of the six children who had an illness duration of 3 weeks or more, three had a gradually progressive illness, two had a biphasic illness with a period of improvement followed by a worsening (one of which was sudden), and one had a mild cough for about 6 weeks and then a sudden deterioration. Most (16) parents had had a face-to-face consultation with a healthcare professional at some point prior to hospital admission. One parent had had only a telephone consultation with an out-of-hours nurse. Six patients had not had a face-to-face consultation prior to the day they were admitted. Ten patients (45%) had received a prescription for antibiotics prior to the day they were admitted. Two of these had been advised not to give the antibiotics immediately (delayed prescribing approach). Of these, one parent reported collecting the prescription and starting the antibiotics within 24 h, and the other did not collect the prescription.

Table 1

Description of participants

Main themes

Three themes relating to parental factors (parent assessment of illness, parent beliefs about health services and feeling powerless to challenge clinical authority) and two themes relating to health service associated factors (accessing healthcare services and quality of the primary care clinical encounter) emerged from the data.

Parental factors

Parental assessment of illness

Five parents reported difficulties in assessing the severity of their child's illness or interpreting their child's symptoms (table 2).

Table 2

Parental assessment of illness

All parents described potentially serious symptoms in their children. The most common were rapid breathing, panting or grunting. Although most parents regarded these symptoms as unusual and often worrying, nearly half (10 parents) described a delay of 24 h or more between first identifying the symptom(s) and consulting a healthcare professional, suggesting a failure to recognise and act on their importance. For example, one parent described her child making a ‘panting noise’, but waiting 5–6 days before consulting because she had difficulty knowing whether he was ‘putting it on’.

Beliefs about health services

Seven parents described not consulting (four with an out-of-hours service, two with a GP, and one with an accident and emergency department (A&E)) at an earlier time because of a fear of ‘overreacting’ or ‘appearing neurotic’ or ‘being seen as a pain’. Five of these described this as not wanting to ‘bother’ the healthcare service (table 3 – 3.1).

Table 3

Parental beliefs about health services

Three parents described not consulting, or consulting later than they would otherwise have (two with a GP in hours and one with a GP out of hours) because, based on past experience, they did not believe their concerns would be taken seriously. One of these did not consult because of a belief that her child had to be sick for a number of days before their GP would prescribe antibiotics (table 3 – 3.2). In contrast, one parent delayed consulting because she believed that antibiotics would be prescribed too readily (table 3 – 3.3).

One parent reported not taking her child to an out-of-hours service because she felt that they would not know her child's past history. Another, whose child was experiencing difficulty breathing, did not feel she was allowed to take her child directly to A&E (table 3 – 3.4).

Parent feeling powerless

Some parents expressed frustration at the outcome of their consultations. Seven felt unable to challenge clinical decisions, either because they did not want to challenge the clinician's authority or they assumed that the trained clinician must be right (table 4).

Table 4

Parent feeling powerless

Health service associated factors

Accessing health services

Nine parents described delays resulting from their general practice's appointment system. These included difficulties in getting through on the telephone, prolonged waits for emergency appointments, being told that there were no more emergency appointments that day, and having to wait for a doctor to call them back, sometimes hours later in the day. These factors sometimes directly resulted in a delay in consulting, and at other times resulted in parents deciding to delay consulting (table 5 – 5.1).

Table 5

Accessing health services

Some of these perceived problems may have been failures of appropriate triage. One parent felt the telephone triage conducted by her practice nurse was inadequate, resulting in a delay in her child being seen. Other parents described problems with triage, including one parent who described being asked to assess features of her child's illness which she felt incapable of doing (table 5 – 5.2).

Quality of the clinical encounter

Nine parents reported feeling dismissed after consulting in primary care, out-of-hours, A&E and secondary care settings (table 6 – 6.1). Six parents considered advice to give their child over-the-counter medicines as dismissive (table 6 – 6.2). Two parents believed that their child had been inadequately examined, two described leaving the consultation with a poor understanding of their child's illness, and three described consultations as ‘rushed’ (table 6 – 6.3).

Table 6

Quality of the clinical encounter

In some instances, parents described consulting with a healthcare professional after their child had developed one or more serious symptoms (rapid breathing, panting, grunting), but with no hospital admission or further investigation resulting from the consultation (table 6 – 6.4). One parent reported taking her child to A&E with difficulty breathing, and being told that this was caused by constipation (table 6 – 6.5). Another reported receiving a prescription for antibiotics over the phone (without being seen), even though the child had developed breathlessness and panting since a previous consultation the week before. One parent reported not using a delayed prescription because she was confused by the doctors advice that this was ‘a viral infection’ and that the antibiotics were ‘as a precaution’.


Summary of the main findings

Parents' accounts of events prior to the admission of their child to hospital with a serious RTI reveal potential delays in receiving appropriate healthcare. We found evidence of problems in interpreting the importance of certain signs, delays in consulting because of problems in accessing healthcare, parents feeling powerless to challenge clinical decisions, and poor quality clinical assessment and communication in primary care consultations. Seventeen of the 22 children had consulted with a healthcare professional at some point prior to the day of their admission. Less than half had received antibiotics. The median duration of symptoms prior to admission was 7 days, but some children appeared to deteriorate very rapidly, while others had a gradual deterioration over weeks.

Strengths and limitations of the study

By conducting face-to-face, semi-structured interviews with the parents of children who had developed serious complications of RTIs, we have been able to tap into rich accounts of the illness course, home management decisions, consultations with healthcare professionals and potential barriers to earlier management. Parents were asked to describe their version of the events and were encouraged to speak freely.

These were retrospective accounts, given by parents who had recently been through the turmoil and upset of having a child admitted to hospital with a serious illness, and this may have influenced their accounts. Nevertheless, many of the identified potential barriers are relatively concrete events (eg, being prescribed antibiotics over the phone, and child panting or grunting for a number of days before consulting).

Those who had experienced problems in the pre-hospital management of their child's illness might have been more likely to agree to participate in the study, seeing this as an opportunity to express a grievance. We were not able to record information on eligible children who were not recruited during the study period. However, the aim of this qualitative study was to identify important categories of information that may be relevant to achieving earlier intervention for complicated RTIs, rather than to generate statistically representative data.

Our sample consisted largely of children with empyema and pneumonia with only one case of peritonsillar abscess, one mastoiditis and no cases of acute rheumatic fever. Therefore, the results may not be relevant to children with these latter complications. However, most of the barriers identified are likely to be relevant to a wide range of acute complications (access problems, difficulty interpreting symptoms, etc).

Comparison with existing literature

Parent associated factors

We were not able to identify any other qualitative studies exploring parental accounts of complicated RTIs prior to hospital admission. However, the parental difficulties in understanding and interpreting their child's illness, deciding when to consult and dealing with healthcare professionals, that were identified in this study have been described in qualitative studies of parents of children with uncomplicated RTIs.18 19 Parents commonly express a desire to have greater information about recognising serious illness in their children,20 and use of a booklet providing such information has been shown to be feasible in a pragmatic trial.21 However, impact on complications and hospitalisations was not assessed.

Health service associated factors

A confidential enquiry of child deaths in England in Wales found that 18 (72%) of the 25 deaths from acute infections had primary care involvement.15 Recognition of illness severity was identified as a critical element. In addition, avoidable primary care factors were identified in the reports of 20% of the deaths, and 72% of these involved death from an infectious cause.

Consistent with our findings, a case–control study examining the influence of primary care processes on infant mortality from acute RTI in Mexico found that maternal delay in seeking healthcare, lack of confidence in healthcare services and problems with access to services, were all important contributory factors.16

A study of clinician management of children presenting to an emergency department with fever found that physicians tend to underestimate the likelihood of serious bacterial infection in young children with fever, leading to under-treatment with antibiotics.22

Importance of breathing abnormalities as a clinical feature

The predominance of abnormal breathing as a clinical feature in children who go on to be admitted to hospital with a serious RTI was notable in our study. A recent systematic review of clinical features associated with the presentation of serious illnesses in children in developed countries23 and a large WHO sponsored study in developing countries24 both identified rapid breathing (as well as reduced consciousness, convulsions, cyanosis and reduced capillary refill) as an important feature. A qualitative study that sought to describe the clinical features of children with serious infections reported breathing abnormalities in all four who had respiratory infections (pneumonia).25

Grunting was identified as an important marker of serious bacterial infection in a recent case–control study of hospitalised children,26 and major challenges in the identification of children who go on to develop more severe illness have also been demonstrated in this setting.27

Implications for clinical practice and research

This exploratory study highlights areas for concern and further exploration of certain processes by which children developing serious infections are recognised and managed. Our findings suggest that there is a need for greater parental education about the importance of certain clinical features (notably breathing problems, rapid breathing, grunting and panting). Public education campaigns have highlighted the importance of recognising a petechial rash. A similar education campaign about the importance of abnormal breathing patterns or breathing problems may address this knowledge gap. Further research to clarify the importance of this problem and identify effective methods for increasing the identification of these features is needed. Problems with accessing general practice and out-of-hours consultations were highlighted. Balancing the requirements to manage workload, encourage self-care where appropriate and provide rapid emergency care where appropriate, is an ongoing challenge for general practice. Our results suggest that some general practice emergency access systems may not working as well as they should in achieving sufficient ‘sensitivity’ in identifying those developing complicated RTIs. Finally, there are potential important messages for clinicians about recognising the signs of serious illness. Differentiating children with RTIs who will go on and develop serious complications from the majority who have self-limiting illnesses is a challenging task that is the focus of ongoing research. Nevertheless, our results suggest that well described features, such as tachypnoea and grunting, are possibly being missed. This is an important concern that needs further investigation, and may indicate the need for ongoing clinical skills training in primary and secondary care.


We would like to thank the parents and children who participated in the study and shared their experiences. We thank the staff in the Department of Child Health and the ENT Department, and in particular the nurse practitioners who helped identify and recruit eligible patients. We acknowledge the contribution made by Jenny Shearman, who interviewed the parents of 10 children, and Naomi Cadbury, who interviewed the parents of six children and helped in the early analysis of data. We thank Kerenza Hood who helped in the design and implementation of the study, and Mike Robling who helped in the early analysis.



  • Funding The Cymru Alliance of Primary Care – Oriented Research Network (CAPRICORN, now AQUARIUS) provided NAF with a Priorities and Needs Fellowship and a Dissemination Fellowship, which funded this project. The authors declare complete independence from the funder.

  • Competing interests None.

  • Ethics approval This study was conducted with the approval of the South East Wales Local Research Ethics Committee Panel B (05/WSE02/186).

  • Provenance and peer review Not commissioned; externally peer reviewed.