Association of Diabetic Ketoacidosis in Childhood New-Onset Type 1 Diabetes With Day of Presentation in Germany

Diabetic ketoacidosis (DKA) at the onset of type 1 diabetes (T1D) represents a serious complication that occurs because of delayed diagnosis and initiation of insulin replacement therapy (1,2). Prevention of DKA is an important goal, but recent data indicate increasing DKA rates (3,4). Delayed initiation of insulin treatment may be due to lack of knowledge of diabetes symptoms by either parents or caregivers, or failure to recognize early signs of diabetes in children presenting to primary health care providers (5).

Primary care for children differs between weekdays and weekends or public holidays. On weekdays, medical care for children in Germany is ensured by regular consultation hours in pediatric practices, while, on weekends and public holidays, the general medical on-call service of the statutory health insurance plays a decisive role for urgent medical needs.

The aim of this study was to investigate the influence of day of presentation on diabetes diagnosis and DKA rate.

We used data from the German prospective diabetes follow-up registry (DPV registry) of children and adolescents aged 0.5 to <18 years at diagnosis (±14 days) of T1D between 1 January 2013 and 31 December 2022. The nationwide coverage of the DPV registry for pediatric patients with T1D in Germany has been estimated at 93% (6). Data are documented at each participating institution using the standardized DPV documentation software (7). The ethics committee of Ulm University (Ulm, Germany) approved the analysis of anonymized data from the DPV registry (ethics approval 314/21); local review boards approved data collection.

Demographic data included age at diabetes onset, sex, and immigrant background (patient or at least one parent born outside Germany).

DKA was defined as pH <7.3 and/or serum bicarbonate <15 mmol/L and as severe if pH was <7.1 and/or bicarbonate was <5 mmol/L (1).

The variables public holiday and school vacation were evaluated separately for each federal state. Public holidays were defined as work-free days (excluding weekends), regardless of religious or political background. Delayed treatment initiation was defined as time interval ≥1 day between diagnosis and treatment initiation.

Unadjusted outcomes were compared using Wilcoxon rank sum test for continuous outcomes or Fisher exact test for dichotomous outcomes.

Logistic regression models were used to determine the odds ratio (OR) of diagnosis of diabetes or DKA on different days of the week, adjusted for age groups (<6 years, 6 years to <12 years, and 12 to <18 years), sex, and immigrant background (yes/no).

P values were adjusted for multiple testing using the Bonferroni-Holm method.

A two-sided P value <0.05 was considered statistically significant. All analyses were performed with SAS version 9.4 (SAS Institute Inc., Cary, NC).

Access to the data is possible by remote data processing upon request.

We analyzed data of 30,717 children and adolescents (55.1% males) with new-onset T1D. DKA at diagnosis of T1D was present in 7,916 children (25.8%). Table 1 provides an overview of the study cohort.

T1D was diagnosed on weekdays in 86.6% of children with new-onset T1D (i.e., mean 17.3% per day), with the highest proportion at the beginning of the week—Monday and Tuesday accounting for 40.9% of all diabetes diagnoses, and 13.4% at weekends (i.e., in mean 6.7% per day) (Table 2). This corresponds to an adjusted OR of diabetes diagnosis on weekends compared with weekdays of 0.39 (95% CI, 0.38–0.41; P < 0.001). Furthermore, T1D was less likely to be diagnosed on public holidays (OR 0.57 [0.53–0.63; P < 0.001]) and during school vacations (OR 0.83 [0.80–0.85; P < 0.001]) than on workdays (Table 3).

A total of 6,503 out of 7,915 cases of new-onset diabetes with DKA (82.2%) were diagnosed on weekdays (including 3,129 cases [39.5%] on Mondays and Tuesdays), compared with 1,412 cases on weekends (17.8%) (Table 2). This corresponds to an adjusted OR of DKA diagnosis on weekends compared with weekdays of 0.55 (95% CI, 0.52–0.59; P < 0.001). The adjusted OR of DKA diagnosis during a public holiday and school vacation compared with a workday was 0.73 (0.63–0.84; P < 0.001) and 0.85 (0.80–0.90; P < 0.001), respectively (Table 3). Diabetes diagnoses decreased more than DKA diagnoses during weekends and holidays compared with weekdays. This resulted in a higher proportion of DKA among those diagnosed with diabetes on weekends and holidays (Fig. 1 and Table 3).

Among those with DKA, 2,790 (35.3%) had severe DKA. Of these, most were diagnosed on Mondays (N = 540 [19.4%]) and Tuesdays (N = 429 [15.4%]) (Table 2).

The proportion of children with delayed treatment initiation was 9.1%. The DKA rate was 28.0% in children treated immediately and 21.2% in those in whom treatment was delayed (P < 0.001).

Diabetes diagnoses on weekends, school vacations, and public holidays resulted in higher rates of delayed treatment initiation of 14.4%, 12.4%, and 21.8%, respectively, than diagnoses on weekdays (8.3%) or workdays (8.9%, all P < 0.001). The mean time interval between diabetes diagnosis and treatment initiation was 0.31 days on workdays, 0.55 days on weekends, and 1.17 days on public holidays (both P < 0.001 versus workdays).

An increase in the frequency of DKA and severe DKA was observed during the coronavirus 2019 (COVID-19) pandemic (Table 1). However, outcomes were not significantly affected by the pandemic (Table 3).

This study found lower ORs of diabetes diagnosis, including those with DKA, in children during weekends, public holidays, and school vacations, and a higher proportion at the beginning of the week, probably due to missed diagnoses over the weekend. Therefore, our results imply that a large proportion of patients with new-onset T1D who experience DKA are missed on weekends and holidays.

T1D is the most common form of diabetes in children and adolescents, putting them at high risk of progressing to life-threatening DKA (1). Nevertheless, insulin treatment was not directly initiated after diabetes diagnosis in 9.1% of cases. This is consistent with other reports of delayed insulin treatment (8). Importantly, one in five patients who did not start insulin treatment on the day of diabetes diagnosis developed DKA.

Delayed treatment initiation was more frequent when diabetes was diagnosed on weekends and public holidays. Thus, these cases represent potentially preventable, life-threatening metabolic derangements. Reasons for delayed therapy may include ignorance of the peculiarities of diabetes diagnosis in children with rapid progression to DKA or lack of experience with insulin treatment in children. The higher rate of DKA in those who were treated immediately is likely due to the more severe symptoms in those with DKA, making them more likely to receive immediate treatment.

We assume that the differences in the detection of children with new-onset diabetes between the days of presentation are related to both family and primary health care factors. In addition to lack of family awareness, potentially modifiable risk factors for new-onset DKA include low medical suspicion for new-onset diabetes in children, leading to delayed diagnosis (9).

Family factors refer to behavior around weekends and holidays, as families may choose to wait until the next working day to visit the doctor (10). The likelihood of being diagnosed with diabetes was about twice as high during school vacations as during weekends, suggesting that the family factor is not the main contributor.

In Germany, the majority of outpatient care for children and adolescents is provided by pediatricians in primary care settings (11). On weekends and public holidays, outpatient care is provided by the on-call service. This service is staffed by practitioners with a wide range of specialties, therefore—with a few exceptions—specific pediatric expertise is usually not available at these times. Our study suggests that the lack of pediatric expertise outside of regular working hours may be responsible for underdiagnosis of childhood diabetes and missed cases of DKA. Accordingly, a Swedish study found that referral was delayed in a significant proportion of children presenting to primary care with diabetes-related symptoms (5). In addition, children who were not initially evaluated by a pediatrician had a higher DKA rate (12). Compared to international rates, the rate of DKA in children with new-onset diabetes is relatively low in Germany (3). One reason may be the availability of specialized pediatric care during the week (11).

During the COVID-19 pandemic, there was a marked increase in the frequency of DKA and severe DKA (13). However, the pandemic did not have a significant impact on the additional outcomes of our study.

In conclusion, our study shows that new-onset diabetes in children is often undiagnosed during weekends and public holidays, and many cases of DKA are missed. In addition, children with newly diagnosed diabetes often do not receive appropriate treatment promptly during these times. These problems existed before the COVID-19 pandemic and do not explain the increased frequency of DKA during the pandemic.

The results of our study suggest that improved knowledge of the clinical signs and symptoms of diabetes in children in primary and acute care, possibly through continuous specialized outpatient pediatric care, would likely improve the detection of children with new-onset diabetes and reduce the rate of DKA. It is essential to provide comprehensive education about the symptoms and clinical presentation of childhood T1D at various levels, including the general public, childcare facilities, and primary care physicians. This could help raise awareness of the symptoms of T1D (14).

Acknowledgments. The authors give special thanks to A. Hungele and R. Ranz (clinical data managers, Institute of Epidemiology and Medical Biometry, Ulm University, Ulm, Germany) for support and the development of the DPV documentation software. The authors thank Dr. Katharina Lipsky (Traunstein, Pediatric Practice Ruhpolding, Ruhpolding, Germany) for her collaboration on the first version of the manuscript. The authors thank all centers participating in the DPV project (https://buster.zibmt.uni-ulm.de/projekte/DPV/). During the course of preparing this work, the authors used DeepL for the purpose of translation assistance. Following the use of this tool/service, the authors formally reviewed the content for its accuracy and edited it as necessary. The authors take full responsibility for all the content of this publication.

Funding. The DPV is supported through the Bundesministerium für Bildung und Forschung within the German Centre for Diabetes Research (grant 82DZD14E03). Further financial support was received by the Diabetes-Stiftung (grant FP-0438-2021), the German Robert Koch Institute, and the German Diabetes Association. The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. C.K., M.S., M.A., and R.W.H. contributed to the concept and design. All authors contributed to acquisition, analysis, or interpretation of data. C.K. and M.S. drafted the manuscript. All authors contributed to the critical revision of the manuscript for important intellectual content. M.A. provided statistical analysis. All authors approved the final manuscript as submitted and agreed to be accountable for all aspects of the work. M.A. and R.W.H. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.