OPEN Research Support

PhD student
Mads Andersen
Paediatrics and Adolescent Medicine, Aarhus University Hospital

Projekt styring
Projekt status    Open
Data indsamlingsdatoer
Start 15.10.2021  
Slut 31.12.2026  

Validation of ICD10 codes for early-onset neonatal infections: A Danish Study

Short summary

As the diagnoses of early-onset neonatal infection is uncertain and no consensus guidelines exist. We aim to study the positive predictive value of the diagnoses based on well-defined criteria. This may enable future cohort studies of this population with investigations of long-term outcomes and determinants hereoff.


Bacterial infections in newborns

With approximately 3 million cases annually worldwide - are a leading cause of mortality and morbidity during the neonatal period.

Early-onset infection occurs during the first week of life and is most often caused by bacteria such as Streptococcus Agalactiae or Escherichia coli acquired from the mother prior to or during delivery. Early-onset infections most often occur in term newborns; however, with a higher risk presenting in preterm newborns.

The diagnosis of early-onset infection in term newborns is challenging and no consensus definition currently exists. Without validated and clinically available molecular testing, culture of bacterial pathogens is thought to be the diagnostic "golden standard". However, recent studies reported that 6 to 16 times more culture-negative newborns are treated with antibiotics compared with culture-positive newborns. This may be explained by low blood concentrations of bacteria and obtainment of small volumes in critically ill newborns, use of maternal antibiotics around the time of delivery, transient bacteremia, or infections contained to the respiratory- and/or central nervous system. However, an over-diagnosis of newborns with early-onset infections may also be evident. When the bacterial cultures are negative and valid molecular tests are not performed, the diagnosis of early-onset infection will be based on the newborn's clinical presentation and biochemical markers. Yet, the clinical signs are unspecific and biomarkers may have poor sensitivity and positive predictive value.

The Danish National Patient Register (DNPR) contains information on all hospital contacts in Denmark including time of admission, physician assigned diagnoses, and specific procedures. The registry is frequently used in Danish epidemiological research and enables nationwide population-based observational studies. Through the central personal registration (CPR) number assigned to all Danish citizens at birth, it further allows for linkage to medical records and several other national databases. However, as the diagnosis of early-onset infection is undefined and may be based on clinical judgement - it would be valuable to investigate the validity of the ICD10 codes for newborn infections registered in the DNPR. This may be a valuable tool for future epidemiological research of newborn infection with investigations of long-term outcomes and determinants hereof. Therefore, we aim to study the validity of the ICD10 codes for early-onset infection used in term newborns by estimating the positive predictive value of the diagnoses based on well-defined microbiological, clinical, and biomarker criteria.

Description of the cohort

All singleton, term newborns (36 to 42 weeks of gestation) born in Denmark from 2010 to 2018 who received an ICD10 code indicative of bacterial infection within the first week of life will be identified in the Danish National Patient Register.

Data and biological material

From above population, 500 newborns will be randomly selected born in either Aalborg University Hospital, Aarhus University Hospital, Odense University Hospital, Zealand University Hospital, or Rigshospitalet. The CPR number will be used to enable linkage to the subjects' electronic medical records. Data from the medical records will be extracted using a standardized form including clinical signs, biomarkers of infection, and type and duration of antibiotic treatment. Two investigators will independently examine the medical records followed by discussion of any discrepancies. The medical records from the first month of life will be collected on request from the above-mentioned departments. From the Danish Medical Birth Register, information will be collected regarding the mother, the newborn, and the delivery.

Collaborating researchers and departments

Department of Clinical Microbiology, Sygehus Lillebælt

  • Stine Yde Nielsen