Status
Conditions
About
This observational retrospective study aims to learn about the incidence of acute kidney (AKI) injury in newborns in infants exposed to nephrotoxic drugs with a big data approach.
The main question it aims to answer are:
The group of infants exposed to drugs will be defined based on exposure for at least 1-day tone one or more therapies commonly used in the NICU. Once the AKI event has occurred, the observation of the trend of daily creatinine and diuresis values will be continued for the period covered by the study.
Full description
Rationale and background Acute Kidney Injury (AKI) is defined as the sudden impairment of kidney function that results in altered hydro electrolyte balance and renal waste product elimination.
Extensive evidence shows that the onset of AKI in critically ill pediatric patients is associated with an increased risk of death and a longer length of hospitalization. Furthermore, in patients who survive an episode of AKI, there is an increased risk of developing long-term morbidity, particularly Chronic Kidney Disease (CKD).
The neonatal kidney has numerous features on the pathophysiological level that predispose this population more to AKI than at other life ages.
In addition to the peculiarities related to the development and physiology of the kidney, especially if preterm, infants admitted to the NICU are exposed to multiple risk factors that may contribute to the onset of AKI.
A recent multicenter study conducted in neonatal intensive care units in Canada, Australia, India, and the United States reported an overall incidence of AKI of 29.9% and showed that the development of AKI is an independent risk factor for death and prolonged hospitalization. The incidence of AKI by gestational age also showed a typical U-shaped distribution, with the greater occurrence at the two extremes represented by infants with gestational ages >36 weeks (36.7%) and <29 weeks (47.9%), while the lowest value was recorded in the 29-36 weeks range.
The risk of developing acute kidney injury is known to increase with the number of medications used (especially if > 3) and the duration of exposure, especially for the aminoglycoside antibiotic category. In addition, critically ill patients who developed AKI were generally more exposed to nephrotoxic drugs than those who did not develop AKI. In the NICU setting, it is estimated that 87% of VLBW infants are exposed to at least one nephrotoxic drug during their hospital stay, and about a quarter of these infants develop at least one episode of acute renal distress.
The diagnosis of AKI is based on increased serum creatinine (SCr) or decreased urinary output.
To date, the definition of acute kidney injury is based on the former Kidney Disease Improving Global Outcomes (KIDGO) criteria modified for newborns.
However, in light of the premises made, it can be understood how creatinine monitoring allows only passive observation of the phenomenon, noting the rise in creatinine values when renal insult has already occurred by then. This allows only retrospective changes in some aspects of the infant's management, such as optimizing fluid intake, suspending the administration of nephrotoxic drugs, and correcting any electrolyte imbalances, but without being able to prevent the onset of AKI. Ideally, the investigators should be able to understand and predict how different risk factors contribute to the beginning of renal damage in a given patient, thus allowing individualized management.
Given the complexity of patients admitted to the NICU and the number of variables in the field, the problem lends itself well to analysis through AI and general statistical inference methods. Such methods have previously been used successfully for studies on adult patients but, to our knowledge, never on newborn patients.
The investigators hope to apply these models in a prospective cohort study to validate their use and develop a real-time monitoring system of the kidney well-being of our nephrotoxic drug-exposed infants that can guide the clinician in patient management. Indeed, systematic surveillance of at-risk patients can significantly reduce the onset of kidney damage and limit both short-term and long-term consequences, thereby improving neonatal outcomes.
Research question and objectives 2.1 Primary objectives
Methods 3.1 Study design Single-center retrospective observational cohort study at the UOC of Neonatology and Neonatal Intensive Care Unit of Fondazione IRCCS Ca' Granda, Ospedale Maggiore Policlinico, Milan, Italy.
3.1.1 Primary endpoints
• Development of AKI defined according to modified KIDGO criteria for the newborn based on increased serum creatinine values or reduced mean hourly diuresis during NICU admission.
3.1.2 Secondary endpoints
Setting A population of infants admitted to the NICU of the Foundation will be retrospectively considered.
4.1 Study population The study population will consist of patients born between January 2010 and December 2022.
Inclusion criteria
• Patients who meet all of the following criteria will be included in the study:
Exclusion Criteria
Patients who meet even one of the following criteria will be excluded:
Variables Clinical and instrumental data regarding each patient's medical history will be acquired for the NICU admission by the first month of life.
Definition of AKI All available serum creatinine values for each patient during the NICU admission will be recorded, considering that determinations made within the first 48 hours of the infant's life are affected by maternal creatinine values.
The presence of AKI will be determined according to the modified KIDGO criteria for the newborn, as reported in previous work in the neonatal setting.
Exposure to drugs. The group of infants exposed to drugs will be defined based on exposure for at least 1-day tone one or more therapies commonly used in the NICU. Special attention will be given to exposure to antibiotics, antivirals, antifungals, diuretics, anti-inflammatories, inotropes, and vasopressors. In addition to the active ingredient, other significant data such as prescribed dose, administration route, treatment duration, and temporal relationship between treatment and diagnosis of AKI will also be considered. Once the AKI event has occurred, the observation of the trend of daily creatinine and diuresis values will be continued for the period covered by the study.
Source documents Data will be collected from each newborn's Neocare electronic medical record (GPI SpA).
Sample size Approximately 600 newborns are admitted to the neonatal intensive care unit each year. The incidence of renal damage in the critically ill newborn population is reported in the literature to be approximately 30%. In light of these considerations, over the decade of our observation, the estimated number of infants admitted to the NICU is about 6000 patients. Assuming a 30% drop-out related to exclusion criteria (about 1800 subjects), the investigators expect to be able to include approximately 4200 subjects in the study, among whom the investigators should find about 1200 cases of acute renal damage.
Data management Data will be collected anonymously by assigning a code to each patient. They will then be organized and stored on data storage systems with secure access.
Data Analysis Data will be analyzed using statistical inference approaches peculiar to Data Science and Artificial Intelligence (AI), such as Decision Trees, Logistic Regression, and Machine Learning (ML). The data themselves will suggest the choice of methods used.
Primary endpoint: analysis Identifying patterns or items in the administration of so-called off-label drugs in the neonatal patient (premature or not) and the development of renal syndromes related to the administration of the drug itself. Once one or more correlation factors have been identified, they could be entered within the Electronic Medical Record system (and in particular within the Clinical Decision Support System) in order to be able to alert the treating clinical staff of any clinical risk associated with the administration.
Secondary endpoint: analysis Development of a computational Early Warning model for AKI that ensures the interpretability of the prediction.
Enrollment
Sex
Volunteers
Inclusion criteria
Exclusion criteria
Loading...
Central trial contact
Ilaria Amodeo, MD; Giacomo Cavallaro, MD, PhD
Data sourced from clinicaltrials.gov
Clinical trials
Research sites
Resources
Legal