Categorization, such as that represented by the taxonomy of Linnaeus and periodic table of Mendeleev, can help us make sense of and understand the world around us. The success of systems such as these underscores the lack of accepted categories and the corresponding dearth of insight that we have with regard to preterm birth, despite the fact that it remains one of our major public health problems.

The health and economic burdens of preterm birth are indeed remarkable. It accounts for 33% of blindness among children, nearly 50% of cerebral palsy, and is a leading cause of perinatal death. There is also increased evidence that preterm birth is related to morbid conditions that appear to develop in adulthood, including cardiovascular disease. It has been estimated that the health ramifications of preterm birth cost the health care system >$26 billion annually. All the more sobering is the fact that we have made little progress in reducing its frequency, despite intensive basic science, translational, clinical, and health services research.

The reasons for the lack of progress are doubtlessly multifactorial, although one of the simplest reasons may be the way we conceptualize preterm birth as a result of our present classification system (or lack thereof). Kramer et al describe how, given the multiple causes for preterm birth, the dearth of a more nuanced classification system affects our ability to properly understand its multifactorial nature and properly investigate its myriad causes. They point out, for example, how relatively aimless and difficult the interpretation of a genome-wide association study of adult premature death would be, given that the underlying reasons for this death are so many and so fundamentally different.

This article by Kramer et al is the first of 3 to be presented in this issue of the American Journal of Obstetrics and Gynecology that seek to rectify the present situation with regard to preterm birth classification and conceptualization. These articles are the product of a working group of international experts formed as a result of the Global Alliance to Prevent Prematurity and Stillbirth meeting. They are organized cogently to demonstrate: (1) the inadequacy of historic attempts at classification and the benefits that an adequate classification system would provide, (2) the issues that must be considered and the decisions that must be made in the construction of an adequate classification system, and (3) the approach to construction of a useful classification system.

Kramer et al, in the first of these articles, pointedly describe the many difficulties that preterm birth presents to any group that wishes to make order out of the present chaos. Few will take issue with their description of the complexity of preterm birth and of the difficulty in forming a classification system when we have difficulty uniformly measuring its occurrence or understanding its different phenotypes. I believe few investigators also are unlikely to take issue with Goldenberg et al, who mapped the key questions that should be addressed in any classification system that would be constructed. The following key questions seem obligatory to consider in any system: (1) which gestational age boundaries should be included, (2) whether phenotype or cause should guide classification, (3) whether risk factors should be included, (4) how terminations, stillbirths, and multiple births should be considered, and (5) how “indicated” and “spontaneous” births should be understood.

The answers suggested by Goldenberg et al, and the final article by Villar et al which presents a 5-component classification system for preterm birth, will in contrast, I suspect, provoke controversy and disagreement. Some will take issue with enlarging the gestational age boundary to range between 16 and 39 weeks of gestation. Others will question the decision to include preterm stillbirths as a component of preterm births. And others will wonder whether the 5 components are too few, too many, or even the correct ones.

Yet, these disagreements should not result in our collective paralysis. There is no experiment that can be done to determine whether this is the correct or best system. Any system is bound to provoke disagreement, and it is unlikely that any other group that convenes is likely to have more expertise with regard to preterm birth or a priori be more likely to produce a better concept. The lack of a uniform classification system is one factor that has contributed to our inability to make more substantive progress in ameliorating the damage that is related to preterm birth. The system that is proposed in these articles should serve as a template for and stimulus to decide on just such a system. A new classification system, even if it is not absolutely identical to the one proposed, that is used uniformly by different investigative teams, holds the potential for manifold benefits, including the ability to shape our understanding of preterm birth in unexpected and productive ways.