This is a controversial issue and the vast majority of doctors and experts recommend giving the treatment shortly after birth.  Nonetheless, there is research to suggest that this routine may not be necessary.

Obviously, this represents a minority point of view both in the hospital and in the pediatric community in general.

I have added my own emphasis to these articles.

Pediatrics 1993 Dec;92(6):755-60
Randomized trial of silver nitrate, erythromycin, and no eye prophylaxis for the prevention of conjunctivitis among newborns not at risk for gonococcal ophthalmitis. Eye Prophylaxis Study Group.
Bell TA, Grayston JT, Krohn MA, Kronmal RA
Department of Biostatistics, University of Washington, Seattle 98195.
OBJECTIVE. To compare the efficacy of commonly used forms of eye prophylaxis for newborns with no prophylaxis in the prevention of nongonococcal conjunctivitis. DESIGN. Randomized doubly masked clinical trial. SETTING. University of Washington Hospital and affiliated clinics, Seattle, between 1985 and 1990. SUBJECTS. The medical records of 8499 women were evaluated for possible participation; 2577 were eligible. Of the 758 enrolled, the infants of 630 were evaluable. INTERVENTION. Comparison of silver nitrate, erythromycin, and no eye prophylaxis given at birth for the prevention of conjunctivitis. MAIN OUTCOME MEASURES. Conjunctivitis during the first 60 days of life and nasolacrimal duct patency in the first 2 days of life. RESULTS. The frequency of impatent tear ducts at the 30- to 48-hour examination did not differ significantly by prophylaxis group. Among the 630 infants randomized and observed, 109 (17%) developed mild conjunctivitis. Sixty-nine (63%) of the cases appeared during the first 2 weeks of life. After 2 months of observation, infants allocated to silver nitrate eye prophylaxis at birth had a 39% lower rate of conjunctivitis (hazard ratio = 0.61, 95% confidence interval = 0.39 to 0.97), and those allocated to erythromycin had a 31% lower rate of conjunctivitis (hazard ratio = 0.69, 95% confidence interval = 0.44 to 1.07), than did those allocated to no prophylaxis. CONCLUSION. Silver nitrate eye prophylaxis caused no sustained deleterious effects and even provided some benefit to infants born to women without Neisseria gonorrhoeae. However, the effect was modest and against microorganisms of low virulence. The results suggest that parentalchoice of a prophylaxis agent including no prophylaxis is reasonable for women receiving prenatal care and who are screened for sexually transmitted diseases during pregnancy.

Pediatr Infect Dis J 1992 Dec;11(12):1026-30

Prophylaxis of ophthalmia neonatorum: comparison of silver nitrate, tetracycline, erythromycin and no prophylaxis.
Chen JY
Department of Pediatrics, Chung Shan Medical and Dental College Hospital, Taichung, Taiwan, Republic of China.
From November, 1989, to October, 1991, 4544 neonates were born at our hospital. Neonatal ocular prophylaxis immediately after birth was used with 1% tetracycline ophthalmic ointment in 1156 neonates, 0.5% erythromycin ophthalmic ointment in 1163 neonates and 1% silver nitrate drops in 1082 neonates. No prophylaxis for neonatal conjunctivitis was given to 1143 neonates. A total of 302 infants (6.7%) developed conjunctivitis during the first 4 weeks of life. Between December, 1991, and January, 1992, 425 neonates were born at our hospital and all were given 0.5% erythromycin ophthalmic ointment twice in the first 24 hours after birth for ocular prophylaxis. Thirty-one (7.3%) infants developed conjunctivitis during the neonatal period. The incidence rates of neonatal chlamydial conjunctivitis in the tetracycline, erythromycin, silver nitrate, no prophylaxis and erythromycin twice groups were 1.3, 1.5, 1.7, 1.6 and 1.4%, respectively. We conclude that neonatal ocular prophylaxis with erythromycin (one or two doses) or tetracycline or silver nitrate does not significantly reduce the incidence of neonatal chlamydial conjunctivitis compared with that in those given no prophylaxis.

5: Am J Epidemiol 1993 Sep 1;138(5):326-32   
The bacterial etiology of conjunctivitis in early infancy. Eye Prophylaxis Study Group.
Krohn MA, Hillier SL, Bell TA, Kronmal RA, Grayston JT
Department of Epidemiology, School of Public Health and Community Medicine, University of Washington, Seattle.
The authors conducted this study to determine the etiologic agents of conjunctivitis in early infancy. From 1985 to 1990, 630 infants enrolled in a randomized, controlled, double-masked study of eye prophylaxis were observed for 60 days after delivery for signs of conjunctivitis. The following isolates were categorized as pathogens: Haemophilus influenzae, Streptococcus pneumoniae, Neisseria cinerea, Klebsiella pneumoniae, and Chlamydia trachomatis. Using conditional logistic regression for analysis of 97 infant pairs, the authors identified isolates categorized as pathogens almost exclusively among cases (odds ratio (OR) = 18.0, 95% confidence interval (CI) 2.3-128). Among the microorganisms which have not usually been regarded as pathogens in the etiology of infant conjunctivitis, Streptococcus mitis was the only microorganism associated with an increased risk of conjunctivitis (OR = 5.3, 95% CI 1.8-15.0).The findings concerning the species of bacteria most often associated with conjunctivitis, as well as the finding that method of delivery is unimportant, suggest that bacteria were transmitted to the infants’ eyes after birth and not from the birth canal.

14: Pediatr Infect Dis J 1989 Aug;8(8):491-5
Failure of erythromycin ointment for postnatal ocular prophylaxis of chlamydial conjunctivitis.
Black-Payne C, Bocchini JA Jr, Cedotal C
Department of Pediatrics, Louisiana State University School of Medicine, Shreveport 71130.
Chlamydia trachomatis is the most common pathogen associated with conjunctivitis during early infancy in the United States. During a 13-month interval at our medical center 4834 infants were born, 311 of whom (6.4%) had conjunctival specimens tested for chlamydial antigen before the age of 12 weeks. In 44 (14% of all tested infants, 0.9% of live births) chlamydial antigen was present. Because the rate of asymptomatic maternal chlamydial endocervical colonization is estimated to be 26% at our institution (previous prospective study), we calculated a minimal failure rate for erythromycin ocular prophylaxis of from 7 to 19.5%. A subsequent case-control study revealed that mothers of infants with chlamydial conjunctivitis were more likely to be primiparous (P = 0.03) and experience longer duration of rupture of membranes before delivery (P = 0.046). We conclude that a substantial percentage of infants exposed to Chlamydia develop chlamydial conjunctivitis despite receiving erythromycin ocular prophylaxis.

In the first 24 to 72 hours after birth babies tend to lose about 3-10% of their birth weight and then regain that weight over the next 2 to 3 weeks. If a mother receives lots of IV fluids during labor, the baby could be born “heavier” because of the increased water. The somewhat higher weight could be measured if a baby were weighed right before it peed for the first time. The difference of this extra fluid retention might only be a few ounces, but some parents are told to be concerned when, at their baby’s two week checkup, the baby is a few ounces under birth weight.

Another common problem at early checkups is a baby that is not gaining what the practitioner considers to be “normal weight gain.” There is not general agreement on normal weight gain and the range in texts are from 4 to 8 ounces a week. Some babies are genetically destined to be a lot smaller or larger than others. As I mentioned in the first paragraph: Easy concept, isn’t it?

If you have been told that weight gain is not acceptable, look hard at this list of questions:

• Is your baby eager to nurse?
• Is your baby peeing and pooping well?
• Is your baby’s urine either clear or very pale yellow?
• Are your baby’s eyes bright and alert?
• Is your baby’s skin a healthy color and texture?
• Is your baby moving its arms and legs vigorously?
• Are baby’s nails growing?
• Is your baby meeting developmental milestones?
• Is your baby’s overall disposition happy and playful?
• Yes, your baby sleeps a lot, but when your baby is awake does he have periods of being very alert?

If you have answered yes to the above questions, you may want to progress on to two important questions which the “charts” seem to ignore.

• How tall is mom?
• How tall is dad?

If someone were to ask you what weight a 33 year old man should be, you would laugh. The range of possibilities varies according to height, bone structure, ethnicity and many other factors. Yet babies are expected to fit onto charts distributed throughout the country with no regard to genetics, feeding choice or almost anything else.

There can be nursing problems that can cause slow weight gain; an inadequate “latch-on” is probably the only common breastfeeding problem in the first weeks. This is an easily remedied problem with the right help. In the best of circumstances, breastfeeding should be assessed within the first day or two after birth by a skilled lactation expert. Good hospitals have these LC’s and IBCLC’s on staff and, if not, please line up a consultation within the first 12 hours of life. Your pediatrician can help you with this. If not, call La Leche League and ask them whom they recommend in your area. This is a crucial step in becoming a parent and must not be skipped.

If there are nursing problems, the first answer should never be supplementation but must be to find the best advice and help available. Find quality help in person if possible and online if needed. There is nothing better than having an experienced breastfeeding expert watch you and your baby and give you the help and encouragement and support you need and deserve.
Too many mothers and babies lose the breastfeeding experience and the lifesaving and illness preventing benefits because we doctors are trained to look harder at the scale than we are at the baby.

A few notable examples:

• Baby, birth weight: 9 lbs. 12 oz.
  Weight 36 hours after delivery: 9 lbs. 2 oz.

I have seen mothers encouraged to supplement because “they have no milk, the baby is hungry and losing weight.” The baby looks good and is nursing every 1 to 3 hours and mom’s nipples are not getting sore. There is no need to do anything but nurse often, switch breasts every 5 minutes or so and wait another day or two for the milk to come in. A thirsty baby nurses strongly and is in no danger. A baby given water or formula might not nurse so strongly and mom’s confidence (and milk supply) will suffer for it. This mom only needs the support of an expert who can be sure that she knows how to latch her baby on to the breast.

• Same baby, two week checkup: 9 lbs. 6 oz

Forgetting that this represents a 4 oz. weight gain from the 36 hour weight, some docs might recommend supplementation. Again, watch breastfeeding and if everything is going well, don’t worry. A dry, jaundiced baby with darker yellow urine is a different case and needs more help with nursing. This baby still should not get formula. Make sure mom is drinking enough water, nursing often without a set schedule (every 1 to 3 hours) and make very sure that she gets help latching her baby on, especially if she has sore nipples.

• Same baby, six month checkup: 15 lbs.

Lactation consultation had been successful in the early weeks thanks to mom having found a supportive, smart doctor and being determined to succeed at feeding her baby the best. This big baby (9 lbs. 12 oz. at birth, remember?) had weighed 13 pounds at her four month visit and now weighs 15 pounds. The doctor is paying attention and sees that Mom is 5′ 3″ and Dad is 5′ 9″ and slender. He looks at the charts second and the baby first and isn’t concerned about the baby dropping from a very high percentile at birth to a lower one and then to a lower one still.

I think I’ll conclude this scenario with this happy ending.

In summary, babies who are nursing, peeing clear urine and wetting diapers well in the first weeks of life are almost always all right. I cannot recall seeing a baby for whom slow weight gain in the first 2 to 6 weeks was the only sign of a problem.

Older babies, 2 to 12 months of age, grow at varying rates. Weight gain should not be used as a major criterion of good health. Developmental milestones and interaction with parents and others are more important. Do not be persuaded to supplement a baby who is doing well. Get help with breastfeeding and use other things besides weight to guide you.

In extreme situations, hyperbili can stain the brain and damage it or make babies sleepy, lethargic and prone to under eating and dehydration. Severe infections often have jaundice as their first sign. These are rare problems but it is dangerous to ignore the possibilities.

In the vast, vast majority of healthy full term babies, mild jaundice is normal and harmless. Breastfeeding babies get more jaundiced than formula fed babies and there are some researchers who think this represents an advantage because bilirubin is a “bacteriostatic” chemical. That is, it inhibits the growth of infectious bacteria.

Jaundice itself may not be a problem but increased jaundice may indicate poor breastfeeding. This is a problem and this is one of the reasons that the “knee jerk” reaction to jaundice is so dangerous. Instead of just attempting to change the baby’s color from slightly yellow (or quite yellow) we should be looking much harder at the reason for the increased bilirubin. Lactation consultation is usually much more valuable than bilirubin tests and bili lights.

PLEASE REMEMBER, this information is strictly applicable to healthy full term babies; preemies and sick kids must be cared for differently and this is beyond the scope of this brief discussion.

Jaundice may continue and even increase through the first week or two and if a baby is doing well, milk is in, nipples are not sore, jaundice is rarely a problem. If a baby continues to lose weight and jaundice is increasing, lactation help is needed and closer medical observation is crucial.

It is almost never correct to interrupt breastfeeding nor to supplement babies with mild to moderate jaundice. The disease entity called “Breast Milk Jaundice” is rare and possibly involves a chemical in breastmilk which inhibits the metabolism of bilirubin. This jaundice is longer and more yellow and almost always harmless. Some experts advocate breastfeeding interruption for a day to help make the diagnosis in a baby with jaundice beyond week two or three. Many experts do not recommend this and instead will wait if the baby’s clinical appearance is reassuring.

Jaundice is a normal condition of babies in the first week of life. Healthy full term babies who are nursing well, urinating and looking good do not need intervention. Supplementing with water makes this worse and supplementing with formula is not necessary and interferes with breastfeeding success and good health.

More information:

• Breastfeeding and Jaundice
• LLLI: Learning About Jaundice

Bilirubin has the ability to function as an antioxidant in the brain, scavenging free radicals and protecting the brain against oxidative damage.

“When women breastfeed, the babies have higher levels of bilirubin and are healthier. Babies with higher bilirubin levels are more disease-resistant,” said Dr. Sylvain Dore of Johns Hopkins School of Medicine, Baltimore, Maryland. “Bilirubin also protects against retinopathy in premature babies.”

Dr. Dore has done research on the neuroprotective effect of bilirubin in the hippocampus. His studies have indicated that low concentrations of bilirubin decreased oxygen-radical mediated injury, suggesting that bilirubin could act as an antioxidant.

Dore further experimented on cultured neurons showing that bilirubin protects against oxidative stress. The enzyme hemeoxygenase is responsible for making bilirubin. In these experiments researchers prevented bilirubin synthesis by eliminating the gene for hemeoxygenase and found, as a result, twice the level of stroke damage in mice.

There is also some belief amongst medical professionals that bilirubin is a bacteriostatic compound which acts to slow or eliminate bacterial growth and therefore give an advantage to babies with higher levels of jaundice. This theory would contribute to the fewer infections in breastfed babies, whose bili counts often descend at a slower rate.

These studies shed new light on the way in which bilirubin in a newborn should be viewed. There is no reason to overreact to bili counts ranging up to low 20′s as long as the mother is frequently nursing (every 60 – 90 minutes during the mother’s waking hours and no more than two stretches of four hours maximum at night), baby is wetting and the counts have begun to slow in ascension or have begun their descent.

Mother must be diligent at following a pattern of frequent nursing until jaundice is gone, even if it includes having to wake the baby, because jaundice tends to make a baby sleepy.