Deciding Whether to Adopt a New Feeding Protocol: Appraising the Evidence
Your NICU is considering a change from scheduled bolus feeding to a more rapid-advance feeding protocol for infants 1,000–1,500 g, intended to shorten time to full enteral feeds. The proposal cites a single-center randomized trial (n=180) showing a 2.1-day reduction in time to full feeds (95% CI 0.4–3.8 days) and no difference in necrotizing enterocolitis (NEC). The trial was conducted at one institution, used a feeding protocol that differs from yours in fortification timing, and excluded infants with a history of umbilical arterial catheter use. Your unit's baseline NEC rate is low and your surgical coverage is limited on weekends. A physician champion wants a decision at the next practice council meeting and argues "the trial is randomized, so it's settled."
1. Appraise the evidence. What can and cannot be concluded, and what is missing?
Read the worked answer
Randomization establishes internal validity for the comparison studied; it does not establish that the result transfers to your setting. What the trial supports: in that single center, with that protocol and those exclusions, a 2.1-day reduction in time to full feeds with no detected NEC difference. What it does not support: a claim that the approach is safe and effective in your unit, for your population, with your fortification timing and weekend coverage.
Key limitations to surface:
- Single center: protocol details, patient mix, and care processes are not generalizable without replication.
- The confidence interval on time to full feeds (0.4–3.8 days) is wide; the effect could be clinically trivial.
- "No difference in NEC" in n=180 is not evidence of no difference - the trial is underpowered for a low-frequency, high-consequence outcome. Absence of a detected difference is not absence of risk.
- Exclusions (UAC history) and differing fortification timing mean your population is not the trial population.
- Your low baseline NEC rate and limited weekend surgical coverage change the risk-benefit calculus: a rare harm is harder to detect and costlier to rescue.
What is missing: local baseline data (time to full feeds, feeding intolerance, NEC), a synthesis beyond one trial, and a plan to detect harm if adopted.
2. The physician champion says "it's randomized, so it's settled." How do you respond, and what decision pathway do you recommend?
Read the worked answer
Respond respectfully and concretely: randomization answers "does it work under studied conditions," not "will it work here, and at what risk." Then redirect to a decision pathway rather than a debate.
Recommended pathway:
- Do not adopt unit-wide on one trial. Consider a structured, time-limited implementation with defined eligibility, exclusions mirroring the trial where appropriate, and a pre-specified monitoring plan.
- Establish local baseline metrics first: time to full feeds, feeding intolerance, NEC, and weekend escalation capacity. You cannot evaluate change without a baseline.
- Pre-specify safety triggers and a stop rule (e.g., a defined cluster of feeding intolerance or NEC triggers review and pause).
- Include the practice council, nutrition, and surgery in the plan; address weekend coverage explicitly.
- Reassess at a defined interval with the same metrics.
Frame it as "how do we adopt this responsibly" rather than "no." That preserves the relationship and the evidence.
3. How would you engage bedside nurses and families in this decision?
Read the worked answer
Bedside nurses hold the tacit knowledge the trial cannot capture: how infants in your unit tolerate advancement, what feeding intolerance looks like early, and where weekend coverage breaks down. Engage them through the practice council and a short structured feedback process (what would make you confident or worried about this change?), and include night and weekend staff, who are most exposed to the risk.
Families: feeding advancement is a decision parents experience as "when will my baby eat normally." Provide plain-language explanation of the change, the uncertainty, and the monitoring plan. Invite parent perspective on communication and consent for the implementation period.
Close the loop: report baseline and follow-up results back to staff and families. Engagement without feedback becomes extraction, and you will lose participation for the next change.