Faculty Bibliography

BACKGROUND:Because of the strong link between childhood obesity and adulthood obesity comorbidities, and the difficulty in decreasing body mass index (BMI) later in life, effective strategies are needed to address this condition in early childhood. The ability to predict obesity before age five could be a useful tool, allowing prevention strategies to focus on high risk children. The few existing prediction models for obesity in childhood have primarily employed data from longitudinal cohort studies, relying on difficult to collect data that are not readily available to all practitioners. Instead, we utilized real-world unaugmented electronic health record (EHR) data from the first two years of life to predict obesity status at age five, an approach not yet taken in pediatric obesity research. METHODS AND FINDINGS/RESULTS:We trained a variety of machine learning algorithms to perform both binary classification and regression. Following previous studies demonstrating different obesity determinants for boys and girls, we similarly developed separate models for both groups. In each of the separate models for boys and girls we found that weight for length z-score, BMI between 19 and 24 months, and the last BMI measure recorded before age two were the most important features for prediction. The best performing models were able to predict obesity with an Area Under the Receiver Operator Characteristic Curve (AUC) of 81.7% for girls and 76.1% for boys. CONCLUSIONS:We were able to predict obesity at age five using EHR data with an AUC comparable to cohort-based studies, reducing the need for investment in additional data collection. Our results suggest that machine learning approaches for predicting future childhood obesity using EHR data could improve the ability of clinicians and researchers to drive future policy, intervention design, and the decision-making process in a clinical setting.

Background: The risk-benefit ratio of ACE-I/ARB therapy after an AKI episode is unclear.
Method(s): We studied 1570 patients recently discharged from hospital and enrolled in a multi-center prospective cohort study (ASSESS-AKI). Follow-up began 3 months after index hospitalization and continued through November 2018. Half of the participants had AKI during the index hospitalization. ACE-I/ARB use and covariates were ascertained 3 months after discharge from the index hospitalization. We used multivariable Cox regression adjusting for demographics, cardiovascular disease, diabetes mellitus, heart failure (HF), blood pressure, urine protein to creatinine ratio, and eGFR to examine the association between ACE-I/ARB use and subsequent death, AKI (>=50% difference between peak and nadir inpatient serum creatinine), renal progression (ESRD or halving of eGFR), and adjudicated HF events.
Result(s): Among study participants who did not have AKI during index hospitalization (N=806), mean age was 65 years, mean eGFR 74 ml/min/1.73m2, and 45% self-reported use of ACE-I/ARB 3 months after hospitalization. Among study participants who did have AKI during index hospitalization (N=764), mean age was 64 years, mean eGFR 65 ml/min/1.73m2, and 50% self-reported use of ACE-I/ARB 3 months after hospitalization. Mean follow-up time was 3.6 years. ACE-I/ARB therapy 3 months after an AKI hospitalization was associated with a lower risk of another hospitalized AKI event and a lower risk of death (Table).
Conclusion(s): Use of ACE-I/ARB in survivors of hospitalized AKI was not associated with increased risk of subsequent AKI but was associated with lower risk of death

Learning Objectives: Radiologist interpretations contribute to anchoring bias. Differentiating problems with similar presentations requires thorough exams and detailed patient histories. Correct management often requires a patient-centered approach. Case Information: A 48 year old woman developed a limp from pain in the ball of the right foot while training for her fifth half marathon. A radiologist diagnosed Freiberg's Disease Stage 2, a rare avascular necrosis of the metatarsal head, based on an x-ray ordered by her podiatrist. X-ray findings include flattening of the metatarsal head, which is a normal variant in 10% of people. The podiatrist prescribed a fracture boot. The patient consulted an orthopedist who changed management to a metatarsal pad for her insole after reviewing the x-ray without an alternative diagnosis. Doubting her diagnosis, the patient consulted a physiatrist who confirmed Freiberg's Disease. After 6 weeks of pain, her orthopedist ordered an MRI that ruled out Freiberg's and showed a partial plantar plate tear with significant localized bursitis. She was taught to tape her toe, but pain persisted with the metatarsal pad. A new podiatrist noticed that the metatarsal pad was creating gait problems and he altered her running shoe insole instead. After two weeks, she was running again.
Discussion(s): Metatarsalgia has many causes, yet three doctors anchored their diagnosis on an incorrect radiology report. Listening to the nuances of the patient's story and performing an extensive exam may have expedited the correct diagnosis. Many doctors use metatarsal pads, but this management may cause harm in some patients. (Figure Presented)