Försenad vård för barn med nydebuterad diabetes typ 1

MÅN, AUG 23, 2021 06:50 CET

Fyra av tio barn med nydebuterad diabetes typ 1 fick inte regelmässig akutbehandling på sjukhus samma dag som familjerna sökte hjälp i primärvården. Det visar en studie från Göteborgs universitet. Försenad sjukhusvård kan i svåra fall leda till livshotande tillstånd.

Studien är publicerad i tidskriften Pediatric Diabetes och omfattar 237 individer upp till 18 års ålder med nydebuterad diabetes typ 1. De vårdades alla på sjukhus i Sverige någon gång under 2015-2017 med syraförgiftning, ett tillstånd som kan uppstå om patienten inte får insulin i tid.  

Uppgifterna i studien kommer från frågeformulär ifyllda av antingen föräldrar eller andra vårdnadshavare, ihop med barnen om de var över 15 år, eller av vårdpersonal på sjukhus. Detta kompletterades med registerdata från Nationella Diabetesregistret (Swediabkids).

I de fall där sjukhusinläggningen föregicks av kontakt med primärvården, och där turerna kunnat kartläggas, visade det sig att 43 procent (48 av 112 fall) inte hade fått behandling på sjukhus samma dag, som riktlinjer föreskriver.

Mer kunskaper behövs

Försenade eller uteblivna akutbehandlingar på sjukhus förekom även i fall där föräldrar eller vårdnadshavare redan före första vårdkontakten fattat misstanke om just diabetes typ 1. Dessa misstankar fanns i 39 procent av de studerade fallen (92 av 237).

Patienter med diabetes typ 1, oavsett ålder, drabbas förr eller senare av syraförgiftning (diabetisk ketoacidos) om insulin inte ges. Med tilltagande syraförgiftning ökar symtomen av exempelvis illamående, kräkningar, och orkeslöshet. Tillståndet kan i förlängningen leda till medvetslöshet och i värsta fall död.

Forskarna bakom den aktuella studien drar slutsatsen att förbättrad kunskap om symtom vid nydebuterad diabetes typ 1, och vikten av snabb handläggning, behövs både bland befolkningen i stort och i primärvården.

Sjukhusvård är avgörande

Studiens försteförfattare är Johan Wersäll, doktorand vid Sahlgrenska akademin, Göteborgs universitet och läkare inom anestesi och intensivvård vid Sahlgrenska Universitetssjukhuset.

– Syraförgiftning är en potentiellt livshotande komplikation till diabetes som helt kan undvikas om insulinbehandling startar i tid, och därför är omedelbar handläggning på sjukhus vid misstanke om diabetes av yttersta vikt, konstaterar han.

Symtom att vara uppmärksam på är exempelvis sängvätning hos ett barn som tidigare varit torrt, att barnet är törstigt och behöver dricka ovanligt mycket, trötthet, illamående, kräkningar, buksmärtor, ansträngd andning och viktnedgång.

– Barn som uppvisar symtom som kan relateras till nydebuterad diabetes bör regelmässigt kontrolleras avseende blodsockret inom primärvården. I de fall blodsockret är förhöjt bör dessa barn omedelbart inremitteras till närmaste akutmottagning på en barnklinik, avslutar Johan Wersäll.

Titel: Delayed referral is common even when new-onset diabetes is suspected in children. A Swedish prospective observational study of diabetic ketoacidosis at onset of Type 1 diabetes, https://doi.org/10.1111/pedi.13229

Kontakt: Johan Wersäll, tel. 0707 59 58 88, e-post Den här e-postadressen skyddas mot spambots. Du måste tillåta JavaScript för att se den.

Bild: Johan Wersäll (foto: Cathrin Jonasson)

Nyhetsinfo

press release Göteborgs Universitet

 

Läs hela texten freepdf inkl artikeln

https://onlinelibrary.wiley.com/doi/10.1111/pedi.13229

From the article

Delayed referral is common even when new-onset diabetes is suspected in children. A Swedish prospective observational study of diabetic ketoacidosis at onset of Type 1 diabetes

Johan H. WersällPeter AdolfssonGun ForsanderSven-Erik RickstenRagnar Hanas

Abstract

Objective

Delayed treatment for new-onset diabetes Type 1 (T1D) can lead to diabetic ketoacidosis (DKA) with potentially devastating consequences. This prospective observational study aimed to characterize pediatric patients with DKA at hospital admission, regarding parental awareness of diabetes-related symptoms and delayed referrals from primary health care providers to pediatric emergency wards.

Research Design and Methods

Patients 0–18 years admitted to hospital with new-onset T1D and DKA between 2015 and 2017 were invited to participate. Questionnaires were filled out separately by the caregivers and by the attending hospital staff. Data from the Swedish National Diabetes Registry (SWEDIABKIDS) were used for comparison. Delayed referral was defined as a primary healthcare contact due to diabetes-related symptoms 0–4 weeks before hospital admission without immediate referral, or registered elevated glucose levels at primary healthcare centers without immediate referral.

Results

The study included 237 patients, among which parental suspicion of new-onset diabetes before healthcare contacts was reported in 39%. Parental suspicion of diabetes was associated with higher pH values at diagnosis. Patients in contact with primary health care providers before hospital admission had a delayed referral in 43% of the cases. Delayed referral was associated with lower pH values at hospital admission. Symptoms leading to primary healthcare contacts were similar regardless of whether delay occurred or not.

Conclusions

Parental suspicion of diabetes was associated with milder DKA at hospital admission. Delayed referral was seen in a considerable proportion of children with primary healthcare contacts for symptoms associated with diabetes. Increased awareness of diabetes symptoms is of paramount importance.

 

INTRODUCTION

New-onset Type 1 diabetes (T1D) is a medical emergency condition. Caregivers should be aware of symptoms of T1D and of the need for prompt medical evaluation of the child if such symptoms appear. According to national and international guidelines, primary health care providers who suspect new-onset T1D in a child should arrange immediate referral to the nearest hospital with pediatric emergency ward capabilities.12 Initial contacts by telephone with medical professionals are becoming increasingly common within the primary health care sector in many countries. It is essential to maintain a low threshold for advising patients or caregivers to seek prompt physical medical evaluation if symptoms that could be attributed to diabetes are reported. In Sweden, the usual way of referral of a child with new-onset diabetes to a pediatric emergency ward is from a general practitioner since children usually do not see a pediatric specialist during an initial primary healthcare visit.

Urine and blood glucose tests as well as ketone tests should be performed liberally in primary health care settings. Primary healthcare centers universally have access to blood glucose testing but usually not to blood ketone meters, and pH levels in blood cannot be measured in primary healthcare centers. All pediatric care is free of charge and paid for by the public healthcare system in Sweden, including patient visits to primary healthcare centers and laboratory tests such as tests for blood ketones (if available) and blood glucose.

Both Swedish2 and international (International Society for Pediatric and Adolescent Diabetes [ISPAD])1 guidelines recommend that children with elevated glucose levels should be urgently referred to a pediatric emergency ward regardless of symptoms. However, several international reports have shown that pediatric patients whose blood glucose levels were above the normal range in a primary health care setting were not referred to a hospital on the same day.3

Untreated new-onset T1D eventually leads to diabetic ketoacidosis (DKA). DKA can have devastating consequences in the acute setting4 and has been associated with long-term adverse effects on glycemic control56 and cognition,7 which makes it imperative to avoid this complication whenever it is possible. Internationally, many cases of DKA among patients with new-onset diabetes are attributed to a missed diagnosis or delays in referral.8 These cases, along with patients' or caregivers' delay in seeking medical evaluation, plausibly explain all cases of DKA as an acute complication to new-onset diabetes. Theoretically, DKA could be entirely avoidable in new-onset diabetes if the general population were aware enough to seek medical care immediately if diabetes-related symptoms are present, and if a delay in the management of such cases were eliminated. Interventional studies aiming at bringing down DKA in pediatric patients with new-onset diabetes by information campaigns have, in some cases, been quite successful.9 In contrast, others have shown less effectiveness.10

The primary aim of this prospective observational study was to describe the status of patients 0–17.99 years of age with new-onset T1D and DKA at hospital admission, with the main focuses to characterize patients whose caregivers had suspected new-onset diabetes before contacting healthcare services, and patients with delayed referral to hospital after contacts with the primary healthcare system. The secondary aims were to analyze the effects of parental suspicion of diabetes and delayed referral on pH levels at hospital admission for DKA.

DISCUSSION

The frequency of DKA at onset is inversely related to the prevalence of T1D.14 Sweden has a high incidence of T1D but a low relative incidence of DKA compared with most other countries.15 In Sweden, DKA at onset has varied between 20% and 25% during 2015–2018.16The proportion of DKA, according to SWEDIABKIDS at diagnosis, was 21.7% during our study period. This proportion is higher than the previous figure of 16%, reported in a similar Swedish study during 1999–2000.17 The findings in this study indicate that a considerable proportion of patients with healthcare contacts due to diabetes-related symptoms before admission for DKA was not immediately referred to an emergency ward. In our study, severe or moderate DKA (pH < 7.2) seemed more common in patients with a delayed referral, which could indicate that delayed referral contributed to a worsened metabolic status at presentation. Since pH values were not measured at the time of contact with the primary healthcare providers, it is not possible to know how many patients who had developed DKA already at the time of consultation. Although it could be assumed that immediately referred patients were those with more typical symptoms, it is nonetheless evident from Table 2 that symptoms at presentation were quite similar regardless of whether the patient was immediately referred or not.

Patients in the lowest pH group had more specific symptoms of acidosis (tachypnea, vomiting from ketonemia) but lower frequencies of the classic diabetes symptoms of thirst and polyuria. The latter finding could have contributed to an overlooked diagnosis of more severe ketoacidosis, resulting in not being referred. Patients not expressing or showing obvious signs of dehydration, for example, thirst or polyuria, thus, seem to be at risk of a missed DKA diagnosis. Although classic symptoms were relatively common in the two groups with higher pH, a large proportion of these patients were not immediately referred.

Many children in whom elevated blood glucose levels were confirmed, were not directly referred to hospital. One explanation for the delay could be that some general physicians and other health care staff are not fully aware of the existence of diabetes in children and adolescents. Another explanation could be that general practitioners assume the diagnosis of Type 2 diabetes when an elevated blood glucose level is found, thus missing the urgency of the situation.

A growing concern is the risk of long-term neurological consequences of DKA. Schoenle et al. reported in 2002 that impaired intellectual development in boys with diagnosis <6 years was associated with DKA at diagnosis.18 A study from 2019 found that a single episode of moderate or severe DKA (as was more common in the group with delayed referral in our study) was associated with lower cognitive scores and altered brain growth.19 Similarly, an Australian study from 2014 showed significant impairment on cognitive tests after DKA.7 A delayed DKA diagnosis may thus result in both a short-term risk of treatment complications and life-long neurological impairment for the child, making it paramount to find measures to avoid DKA both at diagnosis and afterwards. Both primary health care providers and the general population therefore need to achieve and maintain higher awareness of symptoms that could be related to new-onset diabetes. Misdiagnosis has been shown to be significantly associated with the incidence of DKA in other studies.20

Studies describing interventions to bring down the incidence of DKA among patients with new-onset T1D have shown mixed results. Some interventions have shown remarkable effects. For example, an Italian poster campaign which lowered the incidence of DKA to nearly zero in individuals above the age of 6 years,9 and a Turkish study showed a two-fold decrease in the rate of DKA after a 2-year national diabetes awareness program was initiated.21 However, interventions in Austria10 and Wales3 did not show any significant effects of diabetes awareness interventions. The scope of interventions has commonly been to educate primary health care providers, caregivers and children about the possibility of a diabetes diagnosis in patients with typical symptoms or to remind primary health caregivers to immediately refer patients with an elevated blood glucose value to the nearest hospital with a pediatric emergency department.

Since the only way to avoid DKA in patients with new-onset T1D is to initiate insulin therapy urgently, a high level of alertness among health care providers regarding diabetes-related symptoms must be maintained. Blood glucose and ketone tests and immediate referral must be carried out in cases where diabetes is suspected.22 Many patients in this study (59%) had primary healthcare visits before admission, a proportion which is substantially higher than what was found in a large meta-analysis in which 39% of children who presented with DKA had been seen at least once by a doctor before diagnosis.23 A delay for more than 48 h has been associated with increased risk of DKA.24 An increase in routine glucose testing of children in the primary health care setting would likely be able to better identify cases of new-onset diabetes.2526 A specialized pediatric outpatient system with pediatricians serving mainly for this age group instead of general practitioners may have been better equipped to diagnose children with new onset diabetes earlier. However, reports from countries that have this system implemented still show high percentage of DKA at onset.15

In our study, 39% of the caregivers reported that they suspected diabetes before any contacts with medical professionals were taken. In these cases significantly higher pH values were recorded at hospital admission. Such awareness of symptoms related to diabetes was associated with a lower risk of DKA in a UK study.26 For the group with primary healthcare contacts before hospital admission, it is therefore remarkable that delayed referral was seen in 43% of the cases where caregivers had suspected diabetes prior to taking contact with a primary health care service.

We found that classic symptoms were common both in PRIM-D and PRIM-ND, in similar proportions as in a UK study (polyuria 76%, polydipsia 86%).26 Tiredness and weight loss were also very common with similar proportions in both groups. Alertness to these symptoms could probably speed up a referral, thus preventing DKA in many cases and leading to a milder DKA in others. The similar distributions of symptoms in PRIM-ND and PRIM-D could indicate that delay in diagnosis is more of a doctor's delay than patient's delay. The higher HbA1c values in DKA patients versus non-DKA patients in SWEDIABKIDS were similar to findings in other studies.56

When looking at registry data for DKA, one might assume that pH would not be measured in patients whose clinical condition did not indicate DKA. Missing data would then indicate an absence of DKA. If this were true, the frequency of DKA in SWEDIABKIDS would have been 299/1710 (17.5%). However, when combined with our study data, 30 cases with missing data turned out to be non-registered DKA cases, giving a DKA frequency of at least 329/1710 (19.2%). Thus, it is evident that valid DKA frequencies in registries can only be calculated based on cases with registered pH values, that is, missing values cannot be interpreted as non-DKA as discussed by Cherubini et al.15

The reason for a patient being included in our study, but not in SWEDIABKIDS, may be due to missing data in SWEDIABKIDS. One explanation is that some patients with diabetes (refugees, asylum seekers, etc.) have temporary national identification numbers, while only permanent national identification numbers are registered in SWEDIABKIDS.

This study has some limitations and strengths. The study did not include patients with new-onset Type 1 diabetes without DKA. It would have been of interest to investigate this group to increase power in determining the incidence of delayed referral for the total group of children with new-onset T1D. However, if all patients with new-onset T1D would have been included during the study period, it would not have been possible to determine how many cases of DKA that could have been prevented with an immediate referral, since pH and ketones are not measured at primary care contacts. Misdiagnosis was significantly associated with the incidence of DKA in a Polish study27 and DKA was shown to be 74% more common if the diagnosis was delayed in a meta-analysis.23 Another limitation is that caregivers, if they reported suspicions of diabetes before contacting healthcare services, did not report where they got knowledge about diabetes prior to contacting healthcare services. This information would have been useful also to discuss possible ways of improving public knowledge about new-onset diabetes. Further limitations related to methodology are uncertainty whether “diabetes-related symptoms” were communicated the same way in the questionnaires as they were in the contacts between the health care provider and the patients/caregivers, and the possibility of bias since all participants in the study already knew they had diabetes at the time they answered the questionnaire. There are also limitations in the absence of secondary enuresis among the choices in the list of symptoms in the questionnaire (no parents mentioned this as a symptom), and in that only patients with primary health care were asked about symptoms.

A strength of this study is that it was national and population-based, and that the response rate of the questionnaire included 72% of reported cases of DKA in SWEDIABKIDS. Furthermore, we included 2 years of data to account for year-to-year differences. Another strength is that we included all types of primary healthcare providers, not only general practitioners. However, the results of this study are produced in the health care environment that is present in Sweden, and may not be applicable to countries with a health care system that differs markedly.

CONCLUSIONS

Even in a country with a very high incidence of T1D like Sweden, symptoms of T1D are not always recognized by general practitioners and other healthcare providers, and the majority of caregivers seem to be unaware of the possibility of diabetes when their children present such symptoms. Most children presented with classic diabetes symptoms, with similar proportions and elevated blood glucose levels in both the delay and non-delay groups at the primary healthcare contact. Nonetheless, as many as 43% of these children were not immediately referred to a pediatric emergency ward, indicating a substantial doctor's delay. Improved knowledge and general awareness of diabetes symptoms among both caregivers and among medical professionals working in the primary healthcare sector are paramount in improving this situation and preventing DKA.

Nyhetsifno

www red DiabetologNytt