Delivery volume debated

Dan Dattani makes an important point regarding who scrutinizes the establishment of clinical practice guidelines. We are therefore pleased that the SOGC has joined the CFPC and the Society of Rural Physicians of Canada (SRPC) in developing a new policy statement on the number of births required to maintain competence. Since more than half of family physicians in both rural and urban settings attend fewer than 25 births per year, the previous guideline, if implemented (as it was by some governing authorities), could decimate maternity care in Canada.

T.B. MacLachlan is correct in saying that our results from a well-resourced teaching hospital ought not to be generalized to rural Canada. We made that point strongly ourselves.1 We acknowledged that our study had internal but not necessarily external validity. However, there are settings in rural Canada and elsewhere that have fewer than 25 births per year and good birth outcomes.2^,3^,4 We are now working with colleagues in small-volume settings to continue to study these relations.

We do not agree with MacLachlan's final point. It is not appropriate for the SOGC to be prescribing standards for settings where obstetricians do not practise. The SOGC felt comfortable in rescinding the previous guideline, based on our work and the work of others as well as our joint position paper on rural maternity care.5 This kind of partnership between our 3 organizations is a positive for the women and families of Canada.

Although statistically correct, Lindbloom and LeFevre's critique has focused only on our multivariate tables. We also reported unadjusted outcomes. They revealed 5-minute Apgar scores of less than 7 for low- versus high-volume family physicians (4.0% v. 3.7%) and NICU/SCU admissions of 11.6% versus 11.3%. Regarding procedures, the rates for episiotomy were 22.7% versus 19.1%, for instrumental deliveries 14.4% versus 13.3% and for cesarean sections 17.5% versus 16.3%. We find it difficult to believe that these minimal differences are clinically important, and it is unlikely that more study power would materially change the results in either of our reported formats.

Moreover, low-volume family physicians are a heterogeneous group made up of people with various career backgrounds. This also overshadows the minimal differences. Certainly, policy decisions ought not to be made on the basis of such differences. More important, if policy decisions were made, as they have been, on the unsupported belief that low volume is a problem, the denial of access to maternity care to large numbers of urban and rural women would lead to genuine adverse outcomes.

We do agree that more data on low-volume deliveries would be desirable. Thus we will pool data from urban, rural and remote settings to examine infrequently occurring events. And we are pleased to draw attention to a recent publication based on all births in Alberta, also showing low-volume maternity care to be a non-issue.6

Michael C. Klein Head, Division of Maternity and Newborn Care University of British Columbia Vancouver, BC