I commend the efforts of my New Jersey colleagues undertaking quantitative risk assessment for shoulder dystocia, and applying it prospectively.

It is likely, however, that the commonly used but somewhat restrictive definition of shoulder dystocia chosen by the authors (use of maneuvers following a failure of downward traction) is, to a great extent, responsible for the decline in shoulder dystocia incidence each year, and when compared to the group without enhanced assessment and counseling. It also appears that documentation of shoulder dystocia was based on retrospective chart review. Their definition of shoulder dystocia has been criticized for its subjectivity, resulting in underreporting of shoulder dystocia incidence. Illustrating this point is the study by Spong et al, using the same definition, which prospectively evaluated unselected deliveries by uninvolved observers, and reported a shoulder dystocia incidence of 11% with only half being documented. This observed incidence is much higher than that reported by many retrospective studies and supports the impression that the true incidence of shoulder dystocia is unknown due to underreporting.

Of note, the shoulder dystocia incidence and cesarean birth rate of the other group is not stated except that it was unchanged. Also unstated is how shoulder dystocia was defined by the other group. In my opinion, the importance of the definition used by both investigators and practitioners cannot be overstated, since different definitions of shoulder dystocia can influence its reported incidence, and that of brachial plexus injury (BPI). A uniform definition is crucial to any comparative study.

In addition, a recent review examined the relationship of various definitions to shoulder dystocia incidence and causation of BPI, concluding that we can and must try to diagnose shoulder dystocia before potentially injurious traction is applied. Otherwise BPI will continue to occur without shoulder dystocia being recognized, or before it is recognized even if appropriately executed maneuvers are employed.

Standard obstetric texts describe normal delivery of the shoulders as spontaneous (without traction) in most cases. Although using a less restrictive definition, ie, failure of the shoulders to deliver spontaneously, by the next contraction, or within 1 minute of the head may increase reported shoulder dystocia incidence, they are simple objective definitions amenable to prospective study that could enable us to learn the true incidence and optimal management of shoulder dystocia.