A Trial to Assess Optimal Postoperative Feeding Regiments Following Pyloromyotomy

1. SUMMARY AND SPECIFIC AIMS B. RESEARCH DESIGN AND METHODS C. REFERENCES

SUMMARY AND SPECIFIC AIMS

Infantile hypertrophic pyloric stenosis is a condition well known to pediatric surgeons, and has been deemed "the most common cause for surgery in the first 6 months of life". This condition affects a large group of children annually with an incidence of 1.8 per 1000 births. Postoperative length of hospital stay is a financial concern and remains a potential target for reduction in hospital costs. Ultimately, these costs are directly affected by the ability to effectively advance postoperative enteral nutrition. Some experts calculated this savings to be $392.00 per patient while others estimate an impressive $1,290 saved per patient.

There are several choices to consider when determining the timing of initiation of enteral feeds following a pyloromyotomy. Some practice patterns have initiated feedings as soon as the infant awakens from anesthesia. Some authors have suggested a period of withholding feedings for several hours postoperatively, while others have recommended a significantly longer period of starvation (18 hours) before initiating feedings. The ongoing debate arises over whether a physician chooses a standardized, incremental feeding regimen versus an ad libitum feeding schedule which allows the infant to decide when and how much to eat. Neither has been studied effectively in a randomized controlled setting. A recent review of the literature would suggest that a period of 4 hours of NPO, followed by ad lib feedings may be the best postoperative regimen. This however, has not been studied in a randomized, controlled trial.

The investigators therefore hypothesize that 1) infants provided with an ad lib diet following pyloromotomy will progress more rapidly to full volume feeds than infants provided an incremental advancement in diet, and 2) Preoperative blood chemistry abnormalities will predict increased emesis postoperatively.

Therefore, the specific aims of this study are to determine whether:

1. Infants who are provided ad lib feeds following pyloromyotomy will progress more quickly to goal feeds (100ml/kg/day) and be able to be discharged more rapidly.
2. Preoperative abnormalities in potassium, chloride, and bicarbonate will predict increased postoperative emesis
3. RESESRCH DESIGN AND METHODS

B1. Experimental Design.

Specific Aim 1. Hypothesis: Infants who are provided ad lib feeds following pyloromyotomy will progress more quickly to goal feeds (100ml/kg/day) and be able to be discharged more rapidly.

Rationale: If infants are able to reach goal feeds more quickly without experiencing significant emesis, then they may be able to be discharged more quickly, thereby decreasing hospital costs associated with longer hospital stays

Experimental protocol:

Infants will be examined and assessed for a diagnosis of pyloric stenosis. Infants will be diagnosed based on standard of care measures (clinical, ultrasound or UGI findings) and will have preoperative blood chemistries drawn as per standard of care. At the time surgical consent is obtained, the attending or fellow physician will educate the parents about the study and obtain consent if they wish to participate.

The unit of randomization will be the child. Randomization will be done using numbered, sealed, security envelopes available to investigators in the pediatric surgery office. Half of envelopes will assign the child to the "Standard/Incremental" feeding regimen, and the other half will assign the child to the "Ad-lib" regimen. The envelopes will be arranged in a box in random order. The investigator will select and open the envelope sequentially after each enrollment and informed consent of parents of the children.

Standard Incremental

1. NPO for 4 hours following surgery
2. Initiate Pedialyte at 15ml PO Q2 hours x 2 feeds
3. Then switch to formula or breastmilk 30ml PO Q2 hours x 2 feeds
4. Then increase to
5. 45ml PO Q3 hours x 2 feeds, or
6. to goal (100ml/kg/day) if goal volume is less than 45 ml
7. Then advance to goal feeds (100ml/kg/day) PO Q3 hours for 2 feeds
8. "Time to goal feeds" will be recorded when the infant has tolerated two consecutive goal feeds without vomiting and is getting ready to have a third feed
9. Infants will remain on IV fluids until tolerating goal feeds so as to maintain glycemic control and hydration

• If vomiting ensues at any stage, repeat same volume of feeding until tolerates 2 total feedings at that volume and then advance
• If infant vomits three of the same volume feeds in a row, skip the next feed and restart at same volume thereafter. If vomiting still persists, hold the next two feeds and then restart feeds at same volume thereafter
• Reglan may not be used during the first 72 hours postoperatively

Ad-lib

1. NPO for 4 hours following surgery
2. Ad-lib formula or breastfeeds until tolerating goal feeds (100ml/kg/day) x 2 total feeds. Feedings will be limited to 12.5ml/kg per feed so as to avoid overdistention of the stomach
3. "Time to goal feeds" will be recorded when the infant has tolerated two consecutive goal feeds without vomiting and is getting ready to have a third feed
4. Infants will remain on IV fluids until tolerating goal feeds so as to maintain glycemic control and hydration

• If infant vomits three feeds in a row, skip the next feed and restart ad lib feeds thereafter.
• Reglan may not be used during the first 72 hours postoperatively

Infants will need to tolerate two goal feeds without vomiting prior to being discharged. After tolerating two feeds, infants will need to stay in the hospital until it is time for the third feed. The initiation of the third feed will be noted as the "time to goal feeds."

Additional variables to be recorded include age, sex, birth weight, gestational age at birth, preoperative weight, preoperative days of vomiting, preoperative resuscitation volume, dimensions of pylorus on ultrasound/UGI, preoperative blood chemistries, laparoscopic or open procedure, and parental satisfaction of the feeding regimen. Additionally, a phone call will be made within 7-14 days following discharge to assess the infants clinical status as well as continued parental satisfaction. Parents of infants who do not consent for randomization will be asked to allow for perioperative data collection only so that we can get a baseline of our entire pyloric stenosis population. The infants whose data will be collected but who are not randomized will not have their data included in the randomization analysis

Specific Aim 2. Hypothesis:. Preoperative abnormalities in potassium, chloride, and bicarbonate will predict increased postoperative emesis.

Rationale: A single previous study has suggested that the degree of preoperative dehydration, particularly associated with blood chemistry abnormalities in potassium, chloride, and bicarbonate, may predict which infants will have more vomiting postoperatively. If able to be identified, these infants may benefit from a tailored feeding regiment postoperatively.

Experimental protocol: Post hoc analysis will be undertaken to confirm if preoperative blood chemistry abnormalities predict postoperative emesis. Also, we will be able to determine if an incremental or adlib feeding regiment is best suited for this population of infants as well.