Variations in mortality rates among Canadian NICUs —and anonymous reporting

We agree with the CMAJ editors1 and Antoni Basinski that public health care institutions should be transparent and accountable. However, we are concerned about the editors' decision to not publish future research that does not name hospitals. A principal aim of the Canadian Neonatal Network is to develop evidence-based methods for evaluating outcomes and improving quality of care in the neonatal intensive care unit (NICU).2^,3^,4^,5 Our research created an industry-wide risk-adjustment standard (Score for Neonatal Acute Physiology Version II, or SNAP-II),5^,6^,7^,8 which permits valid comparison of NICU outcomes and is now used by hospitals and researchers in many countries. This research could not have been done without confidentiality provisions. Hospitals are naturally wary about giving a priori permission to name them when it is unclear at the outset whether the research will achieve its goal of creating a validated risk-adjustment instrument. Even the editors acknowledge the importance of honouring prior commitments. Contrary to the editors' concerns, SNAP-II actually enables regulators to ensure hospital transparency and accountability, and the irony of the CMAJ decision is that it directly penalizes and discourages research efforts like ours. Like other authors,9^,10 Basinski himself acknowledges that there is no value in publishing league tables on NICU outcomes.

We are puzzled when the editors maintain that there is no evidence that the public acts inappropriately when this kind of information is released, while asserting that “for expectant parents and their physicians it might be important to know the identities of the institutions.”1 We also take issue with the editors' contention that research confidentiality decreases motivation for quality improvement. As part of our ongoing research program, participating hospitals have partnered with the Canadian Institutes of Health Research to co-fund a $1.7 million study to develop an innovative evidence-based practice identification and change system for improving care in Canadian NICUs.

We would like to assure Basinski that anonymity has not led to a “loss of additional valuable information.” His concern about identifying the causes of outcome variation is relevant, and this is integral to our research. We previously published evidence showing that the success of regionalization programs has meant that patient volume is not related to outcome variation in Canadian NICUs.8^,11 We invite readers to visit our Web site (www.caneonet.org) and to read our 91 peer-reviewed publications to date on NICU outcomes and health-services related issues.

We have many more research findings to share with our health-care colleagues in Canada and elsewhere, and we regret that the CMAJ decision will force us to submit future manuscripts to leading foreign medical journals instead of CMAJ. The editors intended their comments to provoke debate. Now that they have made their point, we question the wisdom of a decision to ban future publication of high-quality research, without regard to science or academics, and in apparent disregard of the principles of peer review.

Shoo K. Lee Associate Professor of Pediatrics University of British Columbia Vancouver, BC Koravangattu Sankaran Professor of Pediatrics University of Saskatchewan Saskatoon, Sask. Li-Yin Chien Assistant Professor of Nursing National Yang Ming University Taiwan Robin Walker Professor of Pediatrics University of Ottawa Ottawa, Ont. Mary Seshia Professor of Pediatrics University of Manitoba Winnipeg, Mani. Arne Ohlsson Professor of Pediatrics University of Toronto Toronto, Ont. for members of the Canadian Neonatal Network