Article Text

Download PDFPDF

Timeliness of referral of children with new onset type 1 diabetes
  1. Dhaara Iyer,
  2. Deepak Choudhary,
  3. Juliana Chizo Agwu
  1. Department of Paediatrics, Sundwell and West Birmingham NHS Trust, West Bromwich, UK
  1. Correspondence to Dr Juliana Chizo Agwu, Department of Paediatrics, Sandwell and West Birmingham NHS Trust, West Bromwich, B71 4 HJ, UK; chizo.agwu{at}nhs.net

Abstract

Introduction Type 1 diabetes (T1D) is characterised by autoimmune destruction of pancreatic β cells leading to insulin deficiency. Prompt referral to a specialist paediatric diabetes team (PDT) for insulin initiation and further management is important to prevent diabetic ketoacidosis (DKA), which remains the most common cause of death in this condition.

Objective The aim was to study the timeliness of referrals from general practitioners (GPs) to PDT, of children suspected of having TID.

Method We carried out a survey of the practice of GPs when they suspect TID in a young person, to investigate whether they recognised the need for urgent referral to PDT. In addition, we carried out retrospective case notes review of children diagnosed with diabetes mellitus between January 2005 and December 2014.

Results 111/300 (37%) of GPs replied to survey. Of these, 73/111 (65.8%) would have referred promptly to the PDT in accordance with National Guidelines. However, 34.2% would have taken an action that would have led to delay in referral to PDT. 96 children were diagnosed with TID during the audit period. There was a delay in referral in 35 (36.5%) children. 19/35 (54.3%) of these children presented with DKA. Mean duration of delay in presentation to the PDT was 1.8 days. In both survey and audit, the most common reason for delayed referral was GP attempting to confirm the diagnosis by undertaking further diagnostic tests.

Conclusions Our study has identified a modifiable reason for delayed referrals of children with TID.

  • Type 1 diabetes
  • children and young people
  • Diabetic keto acidosis
  • General Practise

Statistics from Altmetric.com

Background

A recent survey in the UK identified 29 000 children (aged 0–16 years) with diabetes mellitus (DM).1 Of these more than 97% have type 1 diabetes (TID). T1D is characterised by autoimmune destruction of pancreatic β cells leading to insulin deficiency. Prompt referral and treatment is important to prevent diabetic ketoacidosis (DKA), which remains the most common cause of death in this condition. Previous studies in Europe have shown that 25%–40% of children at their initial presentation have DKA.2 Many of these children present initially to primary care healthcare professionals with suggestive symptoms and are then referred on to specialist paediatric diabetes team (PDT) for further management. To avoid delay in the initiation of insulin therapy, the National Institute for Health and Care Excellence (NICE) issued guidance in 2004 advising that ‘children and young people with suspected Type 1 Diabetes should be offered immediate (same day) referral to a multidisciplinary paediatric diabetes care team that has the competencies needed to confirm diagnosis and to provide immediate care’.3 This has been reiterated and amended in the recently published NICE guidelines and the recommendation is to ‘refer children and young people with suspected Type 1 Diabetes immediately (on the same day) to a multidisciplinary Paediatric Diabetes team with the competencies needed to confirm diagnosis and to provide immediate care’.4 The aim is to reduce the number of children presenting with DKA.

Aim

We carried out a survey of the practice of general practitioners (GPs) when they suspect DM in a young person. The aim was to investigate whether they recognised the need for urgent referral to PDT. In addition, we carried out a 10-year audit of the timeliness of referrals of children with suspected T1D to PDT, as well as reasons for any delay.

Method

Between January and August 2011, 300 local GPs practising in the Sandwell and West Birmingham area were asked to complete a postal questionnaire (see online supplementary file appendix 1) indicating what actions they would take if they suspect DM in the following scenario: A 14-year-old girl presents to them with short history of polyuria, polydipsia, lethargy and weight loss and point of care blood glucose was more than 11.1 mmol/L.

We also carried out a 10-year retrospective case notes review of children diagnosed with DM between 1 January 2005 and 31 December 2014 and managed by Sandwell Paediatric Diabetes service. We reviewed demographic data, information on source of referral, date of initial presentation to a healthcare professional, date of insulin initiation, cause of delayed presentation (where applicable), date of initial diagnostic blood tests and mode of presentation.

To study the factors that may be associated with presentation of DKA, we calculated the OR (and CI) by carrying out univariate analysis of each factor of interest. Forward stepwise binary logistic regression was used to perform a multivariable analysis. p Value <0.05 was considered significant. Statistical analysis was carried out using Minitab 17 software.

Results

Results of survey

The response rate was 111/300 (37%). Of these 73/111 (65.8%) would have referred promptly to the PDT in accordance with the NICE guidelines. However, 38/111 (34.2%) would have taken an action that would have led to delay in referral to PDT. Of these, 33/38 (86.8%) would have arranged a fasting blood glucose and only refer urgently if DM was confirmed, 2/38 (5.2%) would organise an oral glucose tolerance test prior to referral to the PDT while a further two GPs would refer the child for an outpatient review. One GP would initiate insulin therapy prior to referral to the PDT. Only 33/111(29.7%) of GPs in the survey had been involved in the diagnosis of a child with DM in the previous year. Of those that declared an interest in adult DM, 32/49 would take the correct action compared with 41/62, who did not declare such an interest; OR (95% CI) 0.96 (0.44 to 2.12) p=0.93.

26/111(23.4%) of GPs felt that childhood DM should be managed by themselves in primary care. This again had no relationship as to whether they would have referred in a timely manner or not; OR (95% CI) 0.79 (0.32 to 1.95) p=0.61.

Results of audit

A total of 117 children were diagnosed with DM in the study period. 21/117 (17.9%) were excluded from further analysis as 9/117 (7.7%) had type 2 DM, 1/117 (0.85%) had glucokinase deficiency and 11/117 (9.4%) children were diagnosed elsewhere and transferred later to our service. Ninety-six children with T1D were included in the study. Their mean (SD) age was 8.6 (3.6) years. 49/96 (49%) were female, 70 (73%) Caucasian, 15 (16%) Asian, 11 (11%) African Caribbean or mixed race. The source of referral was a GP for 72(75%) and the emergency department (ED) for 24(25%). 7/24 (29%) of those that presented to ED had previously been seen by their GP.

Presentation of DKA

37/96 (38.5%) newly diagnosed children in our study presented with DKA. In our cohort, younger age of presentation was the only factor that significantly increased the risk of presenting with DKA (see table 1).

Table 1

Factors associated with presentation of diabetic ketoacidosis

Delay in referral to PDT

There was a delay in referral in 35/96 (36.5%) children. 19/35 (54.3%) of these children presented with DKA. Mean (SD) duration of delay in presentation to the PDT was 1.8 (3.3) days.

Reasons of delay

In 25/35 (71.4%), the GP waited to confirm the diagnosis of DM (with an abnormal fasting blood glucose) prior to referral to secondary care. 10/35(28.6%) of children were diagnosed as having other conditions but the presence of TID at the time of presentation was missed. The conditions commonly diagnosed are viral illness (4), oral thrush (2), vaginal thrush (1), throat infection (2) and gastroenteritis (1).

Discussion

The survey and audit was done to study the timeliness of referrals of children suspected of having TID to specialist PDTs. Ninety-seven per cent of children in UK with DM have TID which is due to autoimmune destruction of the pancreatic β cells that produce insulin; children with TID can therefore develop DKA if insulin therapy is delayed. 34.2% of GPs did not recognise the need for urgent referral and would have taken actions that would have delayed referral and therefore insulin initiation. In both the survey and audit, the most common reason for delayed referral was due to the GP carrying out further investigations, for example, fasting blood glucose to confirm the diagnosis prior to referral. To our knowledge, this is the first survey exploring the intentions of GPs in this scenario. While there is a risk of non-response bias (with under representation of GPs who are not interested in the topic), our findings, however, are similar to a UK wide survey that prospectively studied the experience of children newly diagnosed with TID and found that 46% of children with a delayed presentation to secondary care had non-urgent investigations.5 A possible reason for this may be a lack of familiarity of some GPs with childhood diabetes and the increased risk of developing DKA in this condition. Our study showed that only 29.7% of GPs had been involved in diagnosis of childhood diabetes in the previous year. This is not surprising as estimated annual incidence of childhood TID in UK is 21–34.9 per 100 0006 with a prevalence in England of 186 per 100 000.7 A general practice with 2500 child patients can expect one new diagnosis of T1D every 2.5 to 3 years.8 This is in contrast to adult onset diabetes (which is mainly type 2 DM) which is more common. In 2013, 2.7 million or 6% of the adult population had diagnosed diabetes in England, 90% of them had type 2 diabetes.9

A systematic review of 46 studies involving more than 24 000 children in 31 countries showed that factors associated with increased risk of presenting DKA included younger age (OR 1.59 CI (1.38 to 1.84) for children <5 years vs older), diagnostic error (OR 3.35 CI (2.35 to 4.79)), delayed treatment (OR 1.74 CI (1.10 to 2.77)) and ethnic minority of origin among others.10 In our study, only younger age of presentation was significantly associated with the risk of presentation of DKA. Our inability to show a significant association of other factors in increasing the risk of presentation of DKA is most likely to be due to a type 2 error as our sample size of 96 has an 80% power to detect an OR of only 3.5 or more.

In our audit, 54.3% of children with delayed referral presented to secondary care in DKA. This is a cause for concern especially as a recent case control study of children aged between 6 and 18 years showed that DKA at TID diagnosis results in morphological and functional brain changes on MRI scan. In that study, relative total white matter volume and relative frontal white matter volume were significantly greater in the DKA group at baseline compared with the non-DKA group. Although these MRI changes resolved in the first week after diagnosis, these early brain changes were associated with persisting alterations in attention and memory 6 months later.11 In addition, a rare but dangerous complication of DKA especially in the paediatric population is the development of cerebral oedema which can lead to death or permanent neurological disability. It is therefore extremely important to avoid ketoacidosis in children.

Our study has highlighted a modifiable factor that could be targeted to reduce the incidence of DKA at diagnosis of TID in children. It is important to raise awareness of the NICE guidelines through campaigns and study days. Publicity campaigns in Italy have significantly reduced the incidence of ketoacidosis at diagnosis in children with TID.12 A similar campaign targeting pharmacies, schools, GPs and parents in Wales had limited impact (with only 21% of GPs being aware of the campaign) and failed to lead to a reduction in the proportion of children presenting with DKA.13 In view of the fact that the average GP only makes a diagnosis of childhood diabetes once every 2 years, it is important that any strategy does not depend on GPs’ knowledge and experience. We are currently piloting a strategy where a ‘popup’ alert reminds the GP of the NICE guideline, which recommends prompt same-day referral of all children suspected of having DM, every time they want to place an electronic order for blood glucose on any child aged <18 years. We hope this will reduce attempts by GPs to confirm the diagnosis of childhood DM prior to referral to specialist PDTs. Future evaluation will clarify whether this strategy has helped in the medium term to reduce the rate of delayed referrals and ultimately the incidence of DKA at presentation of TID in children.

Main messages

  • Majority of children with type 1 diabetes in whom referral from primary care to specialist paediatric diabetes team (PDT) was delayed presented with diabetic ketoacidosis.

  • A modifiable cause of delayed referral of children suspected with type 1 diabetes is due to general practitioner's wishing to confirm the diagnosis (by carrying out fasting blood glucose) prior to referral to PDT for insulin initiation.

Current research questions

  • What is the impact of introduction of an automated electronic message (that reminds general practitioners (GPs) of the National Guideline which recommends prompt same-day referral of all children suspected of having diabetes mellitus, every time they want to place an electronic order for blood glucose on any child aged <18 years) on the rates of diabetic ketoacidosis (DKA) in children with new onset type 1 diabetes?

  • What is the impact of publicity campaigns that target specifically GPs on the timeliness of referral of children with new onset type 1 diabetes?

  • What is the impact of introduction of primary care referral guidelines on rates of DKA?

Acknowledgments

The authors acknowledge the advice of Dr Peter Nightingale in the statistical analysis.

References

Footnotes

  • Contributors DI wrote the initial draft of manuscript and carried out the audit. DC carried out the survey. JCA came up with the idea of the survey and audit, reviewed and approved the final draft of manuscript.

  • Competing interests None declared.

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

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.