Patient headache questionnaires can improve headache diagnosis and treatment in children.

OBJECTIVE: To examine trends in diagnosis of headache and migraine in a large pediatric neurology cohort, and test whether an electronic health record (EHR)-integrated headache questionnaire can increase specificity of diagnosis and likelihood of prescribing migraine treatment. BACKGROUND: Under-diagnosis of migraine contributes to the burden of disease. As we founded our Pediatric Headache Program in 2013, we recognized that the proportion of patients with headache who were given a diagnosis of migraine was much lower than expected. METHODS: We developed a patient headache questionnaire, initially on paper (2013-2014), then in an electronic database (2014-2016), and finally integrated into our electronic health record (pilot: 2016, full: May 2017). We compared diagnoses and prescribed treatments for new patients who were given a headache diagnosis, looking at trends in the proportion of patients given specific diagnoses (migraine, etc.) versus the non-specific diagnosis, "headache." Next, we conducted a prospective cohort study to test for association between provider use of the form and the presence of a specific diagnosis, then for an association between specific diagnosis and prescription of migraine treatment. RESULTS: Between July 2011 and December 2022 the proportion of new headache patients who were given a diagnosis of migraine increased 9.7% and non-specific headache diagnoses decreased 21.0%. In the EHR cohort (June 2017-December 2022, n = 15,122), use of the provider form increased the rate of specific diagnosis to 87.2% (1839/2109) compared to 75.5% (5708/7560) without a patient questionnaire, nearly doubling the odds of making a specific diagnosis (odds ratio [OR] 1.90, 95% confidence interval [CI]: 1.65-2.19). Compared to those given only a non-specific headache diagnosis who were prescribed a migraine therapy 53.7% (1766/3286) of the time, 75.3% (8914/11836) of those given a specific diagnosis received a migraine therapy, more than doubling the odds of prescription (OR 2.39, 95% CI: 2.20-2.60). CONCLUSIONS: Interventions to improve specificity of diagnosis were effective and led to increased rates of prescription of migraine treatments. These results have been sustained over several years. This headache questionnaire was adapted into the Foundation system of EpicCare, so it is broadly available as a clinical and research tool for institutions that use this EHR software.

Indexed oxygen delivery during pediatric cardiopulmonary bypass is a modifiable risk factor for postoperative acute kidney injury.

BACKGROUND: Acute kidney injury after pediatric cardiac surgery is a common complication with few established modifiable risk factors. We sought to characterize whether indexed oxygen delivery during cardiopulmonary bypass was associated with postoperative acute kidney injury in a large pediatric cohort. METHODS: This was a retrospective analysis of patients under 1 year old undergoing cardiac surgery with cardiopulmonary bypass between January 1, 2013, and January 1, 2020. Receiver operating characteristic curves across values ranging from 260 to 400 mL/min/m2 were used to identify the indexed oxygen delivery most significantly associated with acute kidney injury risk. RESULTS: We included 980 patients with acute kidney injury occurring in 212 (21.2%). After adjusting for covariates associated with acute kidney injury, an indexed oxygen delivery threshold of 340 mL/min/m2 predicted acute kidney injury in STAT 4 and 5 neonates (area under the curve = 0.66, 95% CI = 0.60 - 0.72, sensitivity = 56.1%, specificity = 69.4%). An indexed oxygen delivery threshold of 400 mL/min/m2 predicted acute kidney injury in STAT 1-3 infants (area under the curve = 0.65, 95% CI = 0.58 - 0.72, sensitivity = 52.6%, specificity = 74.6%). CONCLUSION: Indexed oxygen delivery during cardiopulmonary bypass is a modifiable variable independently associated with postoperative acute kidney injury in specific pediatric populations. Strategies aimed at maintaining oxygen delivery greater than 340 mL/min/m2 in complex neonates and greater than 400 mL/min/m2 in infants may reduce the occurrence of postoperative acute kidney injury in the pediatric population.

Sepsis in Complex Patients in the Emergency Department: Time to Recognition and Therapy in Pediatric Patients With High-Risk Conditions.

OBJECTIVES: To compare timeliness of sepsis recognition and initial treatment in patients with and without high-risk comorbid conditions. METHODS: This was a retrospective cohort study of patients presenting to a pediatric emergency department (ED) who triggered a vital sign-based electronic sepsis alert resulting in bedside "huddle" assessment per institutional practice. A positive sepsis alert was defined as age-specific tachycardia or hypotension, concern for infection, and at least 1 of the following: abnormal capillary refill, abnormal mental status, or a high-risk condition. High-risk conditions were derived from the American Academy of Pediatrics sepsis alert tool. Patients with a positive alert underwent bedside huddle resulting in a decision regarding initiation of sepsis protocol. Placement on the protocol and time to initiation of protocol and individual therapies were compared for patients with and without high-risk conditions. RESULTS: During the 1-year study period, there were 1107 sepsis huddle alerts out of 96,427 ED visits. Of these, 713 (65%) had identified high-risk conditions, and 394 (35%) did not. Among patients with sepsis huddles, there was no difference in sepsis protocol initiation for patients with high-risk conditions compared with those without (24.8% vs 22.0%, P = 0.305). Between patients with high-risk conditions and those without, there were no differences in median time from triage to sepsis protocol activation, triage to initial intravenous antibiotic, triage to initial intravenous fluid therapy, or ED length of stay. CONCLUSIONS: Timeliness of care initiation was no different in high-risk patients with sepsis when using an electronic sepsis alert and protocolized sepsis care.

Surviving Sepsis in a Referral Neonatal Intensive Care Unit: Association between Time to Antibiotic Administration and In-Hospital Outcomes.

OBJECTIVE: To determine if time to antibiotic administration is associated with mortality and in-hospital outcomes in a neonatal intensive care unit (NICU) population. STUDY DESIGN: We conducted a prospective evaluation of infants with suspected sepsis between September 2014 and February 2018; sepsis was defined as clinical concern prompting blood culture collection and antibiotic administration. Time to antibiotic administration was calculated from time of sepsis identification, defined as the order time of either blood culture or an antibiotic, to time of first antibiotic administration. We used linear models with generalized estimating equations to determine the association between time to antibiotic administration and mortality, ventilator-free and inotrope-free days, and NICU length of stay in patients with culture-proven sepsis. RESULTS: Among 1946 sepsis evaluations, we identified 128 episodes of culture-proven sepsis in 113 infants. Among them, prolonged time to antibiotic administration was associated with significantly increased risk of mortality at 14 days (OR, 1.47; 95% CI, 1.15-1.87) and 30 days (OR, 1.47; 95% CI, 1.11-1.94) as well as fewer inotrope-free days (incidence rate ratio, 0.91; 95% CI, 0.84-0.98). No significant associations with ventilator-free days or NICU length of stay were demonstrated. CONCLUSIONS: Among infants with sepsis, delayed time to antibiotic administration was an independent risk factor for death and prolonged cardiovascular dysfunction. Further study is needed to define optimal timing of antimicrobial administration in high-risk NICU populations.

Neonatal sepsis registry: Time to antibiotic dataset.

This article describes the process of extracting electronic health record (EHR) data into a format that supports analyses related to the timeliness of antibiotic administration. The de-identified data that accompanies this article were collected from a cohort of infants who were evaluated for possible sepsis in the Neonatal Intensive Care Unit (NICU) at the Children's Hospital of Philadelphia (CHOP). The interpretation of findings from these data are reported in a separate manuscript [1]. For purposes of illustration for interested readers, scripts written in the R programming language related to the creation and use of the dataset have also been provided. Interested researchers are encouraged to contact the research team to discuss opportunities for collaboration.

Machine learning models for early sepsis recognition in the neonatal intensive care unit using readily available electronic health record data.

BACKGROUND: Rapid antibiotic administration is known to improve sepsis outcomes, however early diagnosis remains challenging due to complex presentation. Our objective was to develop a model using readily available electronic health record (EHR) data capable of recognizing infant sepsis at least 4 hours prior to clinical recognition. METHODS AND FINDINGS: We performed a retrospective case control study of infants hospitalized ≥48 hours in the Neonatal Intensive Care Unit (NICU) at the Children's Hospital of Philadelphia between September 2014 and November 2017 who received at least one sepsis evaluation before 12 months of age. We considered two evaluation outcomes as cases: culture positive-positive blood culture for a known pathogen (110 evaluations); and clinically positive-negative cultures but antibiotics administered for ≥120 hours (265 evaluations). Case data was taken from the 44-hour window ending 4 hours prior to evaluation. We randomly sampled 1,100 44-hour windows of control data from all times ≥10 days removed from any evaluation. Model inputs consisted of up to 36 features derived from routine EHR data. Using 10-fold nested cross-validation, 8 machine learning models were trained to classify inputs as sepsis positive or negative. When tasked with discriminating culture positive cases from controls, 6 models achieved a mean area under the receiver operating characteristic (AUC) between 0.80-0.82 with no significant differences between them. Including both culture and clinically positive cases, the same 6 models achieved an AUC between 0.85-0.87, again with no significant differences. CONCLUSIONS: Machine learning models can identify infants with sepsis in the NICU hours prior to clinical recognition. Learning curves indicate model improvement may be achieved with additional training examples. Additional input features may also improve performance. Further research is warranted to assess potential performance improvements and clinical efficacy in a prospective trial.

Utilization of Antipyretics for Nonurgent Fever in a Pediatric Emergency Department.

This retrospective cohort study aimed to describe antipyretic use among healthy patients in a pediatric emergency department (ED) with nonurgent fever defined as: triage level 4 or 5, chief complaint fever or temperature 38°C to 39°C, and otherwise normal vital signs, and determine if antipyretic administration is associated with increased ED length of stay (LOS). We compared continuous variables using Kruskal-Wallis and Wilcoxon rank sum testing. We adjusted confounding variables using logistic regression modeling. A total of 22 169 patients were included. Of these, 52% received antipyretic: acetaminophen (38%), ibuprofen (19%), or both antipyretics (5%). ED LOS (median hours) varied by number of antipyretic types given (none, 2.2; ibuprofen, 2.7; acetaminophen, 2.7; and both 3.4, P < .001) and number of doses (0 doses, 2.2, 1 dose, 2.7; 2 doses, 3.4, P < .001). Patients who received antipyretic were more likely to have ED LOS greater than 2 hours (adjusted odds ratio 1.99, 95% CI 1.88-2.11) compared with those with no antipyretic, controlling for age, imaging studies, laboratory studies, antibiotic administration, and disposition.

Do changes in socio-demographic characteristics impact up-to-date immunization status between 3 and 24 months of age? A prospective study among an inner-city birth cohort in the United States.

INTRODUCTION: Low-income child populations remain under-vaccinated. Our objective was to determine differences in the relative importance of maternal health literacy and socio-demographic characteristics that often change during early childhood on up-to-date (UTD) immunization status among a low-income population. METHODS: We performed secondary data analysis of a longitudinal prospective cohort study of 744 Medicaid-eligible mother-infant dyads recruited at the time of the infant's birth from an inner-city hospital in the United States and surveyed every 6 months for 24 months. Our primary outcome was infant UTD status at 24 months abstracted from a citywide registry. We assessed maternal health literacy with the Test of Functional Health Literacy in Adults (short version). We collected socio-demographic information via surveys at birth and every 6 months. We compared predictors of UTD status at 3, 7, and 24 months. RESULTS: The cohort consisted of primarily African-American (81.5%) mothers with adequate health literacy (73.9%). Immunizations were UTD among 56.7% of infants at 24 months of age. Maternal health literacy was not a significant predictor of UTD immunization status. Instead, adjusted results showed that significant predictors of not-UTD status at 24 months were lack of a consistent health care location or "medical home" (OR 0.17, 95%CI 0.18-0.37), inadequate prenatal care (OR 0.48, 95%CI 0.25-0.95), and prior not-UTD status (OR 0.31, 95%CI 0.20-0.47). Notably, all upper confidence limits are less than 1.0 for these variables. Health care location type (e.g., hospital-affiliate, community-based, none) was a significant predictor of vaccine status at age 3 months, 7 months, and 24 months. CONCLUSIONS: Investing in efforts to support early establishment of a medical home to obtain comprehensive coordinated preventive care, including providing recommended vaccines on schedule, is a prudent strategy to improve vaccination status at the population level.

High-resolution mapping and analysis of copy number variations in the human genome: a data resource for clinical and research applications.

We present a database of copy number variations (CNVs) detected in 2026 disease-free individuals, using high-density, SNP-based oligonucleotide microarrays. This large cohort, comprised mainly of Caucasians (65.2%) and African-Americans (34.2%), was analyzed for CNVs in a single study using a uniform array platform and computational process. We have catalogued and characterized 54,462 individual CNVs, 77.8% of which were identified in multiple unrelated individuals. These nonunique CNVs mapped to 3272 distinct regions of genomic variation spanning 5.9% of the genome; 51.5% of these were previously unreported, and >85% are rare. Our annotation and analysis confirmed and extended previously reported correlations between CNVs and several genomic features such as repetitive DNA elements, segmental duplications, and genes. We demonstrate the utility of this data set in distinguishing CNVs with pathologic significance from normal variants. Together, this analysis and annotation provides a useful resource to assist with the assessment of CNVs in the contexts of human variation, disease susceptibility, and clinical molecular diagnostics.

Identifying factors predicting immunization delay for children followed in an urban primary care network using an electronic health record.

OBJECTIVE: An opportunity exists to use increasingly prevalent electronic health records to efficiently gather immunization, clinical, and demographic data to assess and subsequently reduce barriers to immunization. The objective of this study was to use data entered at the point of care within an electronic health record to identify factors that predispose children in an inner-city population to immunization delay. METHODS: Retrospective cohort data from an electronic health record were used to evaluate the association between demographic, clinical, and immunization variables on immunization delay at 24 months. Patients 2 to 5 years old as of May 31, 2003, with an office visit between May 31, 2002, and May 31, 2003, were selected (N = 5464). Univariate and multivariable models were developed to predict vaccination delay at 24 months per the Advisory Committee on Immunization Practices guidelines. RESULTS: Overall up-to-date immunization rates at 3, 7, 13, and 24 months were 75%, 45%, 82%, and 71%. Multivariable models using electronic health record data showed that early immunization status was the strongest predictor of immunization delay at 24 months. Multivariate analysis revealed that children who were inadequately immunized at 3 months of age were more than 4.5 times as likely to be immunization delayed at 24 months. In this analysis, patient and caregiver factors associated with immunization delay included insurance status and nonparent caregiver. Children who were premature were less likely to be delayed. CONCLUSIONS: Using an electronic health record with information entered at the point of care, we found that early immunization status is a strong predictor of immunization delay for young children that can be identified as early as 3 months of age. Electronic health records may prove useful to clinicians and health systems in identifying children at high risk for immunization delay.