1. Interesting facts/excerpts from a medscape article on neonatal hematology:
-At birth, normal hemoglobin is 19, hct is 60, and retics are 4-7% (dropping to 0 by 4 days of age perhaps due to the drop in epo 2/2 increased oxygen tension in blood)
-MCHC > 35 implies spherical RBC, since more hgb/cell=fatter cell.
-"Immediately following birth, the neonate's H/H levels increase about 10% over the first 6 to 8 hours of life. This increase represents equilibration from a placental transfusion (from mother to neonate) occurring at the time of birth. When infants are born prematurely, the umbilical cord is cut more quickly... so this phenomenon does not occur."
-"Neonates... erythrocyte levels differ depending on the site of sampling. Capillary (heelstick) samples range from 5% to 25% higher than simultaneous samples taken from veins, arteries, or umbilical vessels. The neonate's generally larger RBCs tend to sludge in the smallest blood vessels, so capillary blood samples contain higher erythrocyte counts. If the infant's heel is warmed prior to lancing, blood vessels dilate, and the resulting capillary hematocrit is closer to venous or arterial values. The site of blood sampling must always be taken into consideration when interpreting the neonate's RBC parameters"
2. Pearls about sodium: 
-increased glucose, protein, lipid, or urea can falsely deflate your Na levels-- "pseduohyponatremia". In the case of glucose, high glucose levels pull water out of cells (since glucose equilibrates across cell membranes slowly in the absence of insulin), diluting out the serum Na. For every 100 mg/dL over normal glucose, add 1.6 mEq to your sodium. in the case of protein or lipid, sodium concentration is only measured in the fraction of free water (normal: 93% of serum), and lipids/proteins take up space and reduce the % of free water per unit of serum. In patients with severe hyperlipidemia or hyperproteinemia (i.e. multiple myeloma), the % of free water in serum may go below 80, artificially reducing the measured sodium.
-If someone is seizing from hyponatremia, they won't stop until you correct the Na-- AEDs will not help you. If they're <120, correct to 120 with 3%NaCl. If you don't have 3%NaCl, you can use sodium bicarb, which is by volume 8%NaCl. Give 1cc/kg boluses until they stop seizing.
-Acute hyponatremia: correct 0.5-1 mEq/hour. Chronic: 0.5mEq/hr
3. Pearls about fluids: 
-Best way to detect changes in fluids in infants and small children is by weight, since it can be hard to find edema on exam.
-NS: 154 Na/Cl, LR: 130 Na, 110 Cl, 4 K, 28 bicarb.
-No difference in outcomes [death, hosp LOS, ICU LOS, vent time, dialysis time] between colloidal and crystalloid solutions in the ICU. SAFE trial: RCT, 28 days, n=7000 NEJM)
-GI>>>IV for fluid resuscitation. If they can drink, make them drink.
4. Resuscitate with isotonic solutions (rather than hypotonic):
-stress => increased ADH/increased sensitivity to ADH, diluting their blood. You don't want to dilute more by giving hypotonic solutions.
-it takes less hyponatremia to cause cerebral edema in a kid vs adult-- their skull/brain ratio is smaller, they have less room to expand. While adults may be able to tolerate sodium into the low 110, with kids its more like 120s.
5. Neonatal fever: how to figure out if a neonate with a fever is harboring a serious bacterial infection. The boston (n>500), philadelphia (n>700), and rochester criteria (n>900) were all based on large retrospective data analyses trying to determine who was at high vs low-risk. '
6. Rochester Criteria for detecting neonatal sepsis: negative predictive value: 98.9, sensitivity 92%, PPV=12%, n=931.
-infants age<60 days, T>38.0
-History: term infant, no illnesses, went home with mom at birth, no perinatal antibiotics
-Exam: well appearing, no ear/bone/tissue infections.
-Labs: WBC 5000-15000, absolute bands <1500, UA wbc <10/hpf, stool smear wbc<5/hpf  [NO LP needed]
-Management: if all above criteria are met, they are "low risk" and can go home with no antibiotics, f/u in a few days.
-Outcomes: no adverse outcomes.
7. Philadelphia Criteria for detecting neonatal sepsis: NPV 99.7, sensitivity 98%, PPV 14%, n=747
-age 29-60 days, T>38.2
-Exam: well appearing, unremarkable exam
-Labs: WBC<15, band-neutrophil ratio <0.2, UA<10 wbc/hpf, urine gram stain neg, CSF wbc<8, CXR no infiltrate, stool smear (if they had diarrhea): no blood, few wbc.
-Management: if all criteria are met, home with no antibiotics
8. Boston Criteria for detecting neonatal sepsis: n=504
-age 28-89, T>38.0
-History: no immunizations/antibiotics within last 48 hours, no dehydration
-Exam: well appearing, no ear/tissue/bone infections
-Labs: WBC<20, CSF wbc<10, UA<10, CXR: clear
-Management: home, WITH empiric antibiotics (in the study: 50mg/kg IM ceftriaxone)
-Outcomes: this group was the most liberal, and as would be expected, a relatively larger false-negative rate. Overall, 5% of the cohort deemed "low risk" eventually were found to be harboring serious bacterial infections, however there were no adverse outcomes.
9. Management of SVT in children:
-If they are well perfused (good cap refill, good mental status): for the older kids, you can do carotid rub; younger kids, cold packs on their neck. If that fails, adenosine (start with 0.1mg/kg, then do 0.2, 0.3, etc. Once you get to 0.4-0.5, consider cardioversion)
-If they do not look well perfused (poor mental status, etc): synchronized cardioversion if getting the defib is faster than getting a line. Otherwise, try adenosine
10. Pearls about steroids: 
-They impair bone healing: i.e. if someone has asthma and just had orthopedic surgery, consider holding the steroids unless they really need it
-Max dose in kids is 60mg/day, so if they're 60kg, they will get 0.5mg/kg BID instead of 1 or 2.