The primary purpose of peer review is twofold: to provide constructive feedback to authors and to assess whether a manuscript meets the standards required for publication. These goals are closely intertwined. A good reviewer should aim to be both critical and supportive — offering comments that are polite, insightful, and ultimately helpful in improving the work.

In neonatology, many of the methodological concerns encountered during peer review are not unique to the field. Common issues include multiple testing without appropriate correction, unclear or potentially overadjusted regression models, underpowered studies, and an excessive reliance on statistical significance. These challenges mirror broader problems seen across medical research.

However, neonatal research also faces a particularly important challenge: defining what constitutes a clinically meaningful effect. Studies in neonatology frequently report multiple outcomes, such as bronchopulmonary dysplasia (BPD), necrotizing enterocolitis (NEC), intraventricular hemorrhage, or mortality. Yet discussions often focus heavily on statistical significance while paying less attention to the magnitude and practical importance of the observed effect.

For clinicians, the key question is not simply whether an intervention reduces risk, but whether the reduction is large enough to meaningfully influence patient care. What degree of reduction in BPD or NEC should be considered clinically important? What level of benefit justifies changes in practice, especially when interventions may carry costs or unintended harms?

The issue becomes even more challenging when dealing with rare but serious neonatal outcomes. In sufficiently large datasets, statistically significant relative risk reductions may be identified even when the absolute benefit is modest. Determining whether such findings are truly meaningful in clinical practice is often difficult — and this uncertainty is frequently encountered during peer review.

Over the years, important progress has been made in standardizing outcome reporting in neonatal research. The next step may be the development of clearer frameworks for interpreting clinical meaningfulness alongside statistical findings. Such discussions would help reviewers, authors, and clinicians place research findings into a more practical clinical context and could ultimately improve the quality and interpretability of neonatal literature.

Ilari, in Kuopio
May 16th, 2026

The greatest joy in research comes from teaching and guiding undergraduate and PhD students in their projects. I have supervised 7 PhD theses and 20 MD theses, and I currently have more than 10 PhD students and over 20 MD thesis students working on various projects. While mainly working with undergraduates, it is important to keep in mind, that they are not yet clinical professionals of the topic.

The core philosophy in our group is that no one works alone. The process begins together, with an emphasis on learning by doing collaboratively, and gradually transitions toward a more independent working style. Traditionally, in many research groups, young researchers are expected to work independently first and receive feedback afterward. However, I believe that teaching and collaboration should come first, followed by increasing independence.

The typical pathway for MD or BM thesis students is as follows:

• We begin with an initial meeting where the topic is discussed and the methodology of the thesis is defined. This may involve either a review or an original research project.

• The goal of the first meeting is to establish an initial plan and ensure the student becomes familiar with the study protocol and the Covidence platform, where the study selection process is conducted.

• After this, we minimize formal meetings and instead rely on rapid communication via email, scheduling meetings only when necessary.

• The project progresses step by step, focusing on one task at a time. There is no need for students to manage multiple tasks simultaneously. The sequence typically includes study screening, full-text review, data extraction, and drafting the introduction, all with appropriate guidance.

For PhD students, the first publication follows a similar structured and guided process. By the second publication, greater independence is expected, and by the third (final) publication, the student is encouraged to lead as independently as possible as the first author.

For me, the key to successful supervision is being available and communicating clearly.

Ilari, in Kuopio
28.4.2026

Meta-epidemiology refers to studies in which, instead of diseases, we aim to study previous studies, reviews, and guidelines — particularly the methods they have used and how those methods have potentially influenced their outcomes and conclusions.

My initial interest in meta-epidemiology arose from the following situation. I was reviewing a Cochrane review and questioned why the certainty of evidence for a mortality outcome had been downgraded due to lack of blinding, given that the intervention in question is nearly impossible to blind. This led to a deep dive into classic meta-epidemiological studies that have analyzed the influence of different risk-of-bias domains on treatment effects. Following this, I wanted to further explore whether this issue was also present in neonatal reviews. At that point, I reached out to Professor Daniele De Luca and proposed a collaboration. Professor De Luca kindly agreed, and we proceeded to publish our review on blinding assessments in neonatal ventilation reviews, in which we found that blinding was mostly handled incorrectly.

As neonates are a vulnerable population, we should first aim to understand what treatments are currently available and what the evidence landscape looks like for these. The next step is to assess the quality of the existing evidence and whether methodological errors may have led to misclassifications of intervention effectiveness — in either a positive or negative direction. Finally, new trials should aim to be clinically relevant and to demonstrate meaningful effects. As the answers to these prior questions remain largely unsatisfactory, meta-epidemiology is clearly and urgently needed.

This site is designed to illustrate and highlight my own scientific journey and reflections that are not worthy of formal publication in the neonatal field.

Ilari in Kuopio
8.4.2026