Interpretation of systematic review evidence: a tale of two nations

Article type
Authors
Fitzgerald A1
1New Zealand Guidelines Group, New Zealand
Abstract
Background: In 2011 the New Zealand Guidelines Group were commissioned to complete systematic reviews on the diagnostic accuracy of positron emission tomography (PET) for diagnosing lymphoma, and the impact PET has on decision making and treatment planning. Cancer Care Ontario (CCO) based in Canada was planning to conduct the same systematic review in early 2012. The two groups agreed that NZGG would complete the review and provide the results to CCO.

Objectives: The objectives were to compare recommendation and decision outcomes for the use of PET scanning in patients with lymphoma between the two countries, using the same systematic review data.

Methods: The review questions, inclusion and exclusion criteria and other details were agreed prior to commencement of the systematic review. While both organisations shared the same systematic review data, the method of recommendation development differed between Canada and New Zealand. Canada commissioned a guideline committee to oversee the development of recommendations arising from the systematic review, while New Zealand used oncology experts to interpret the evidence and make recommendations.

Results: The recommendations developed by the two organisations varied in terms of their content. Commissioning a guideline group resulted in a more thorough discourse regarding how decisions were made. Recommendations resulting from diagnostic accuracy data were more closely aligned than recommendations resulting from data on treatment decisions and data on patient oriented outcomes.

Conclusions: It is often difficult to co-ordinate collaboration on systematic reviews, particularly internationally, so this work represented a good opportunity to investigate how interpretation of a single systematic review can vary. Some of the differences in recommendations are likely to be a product of the different methods of recommendation development between the two organisations, while others differences are likely to be more complex and relate to populations, healthcare contexts and values.