Essentialist or consequentialist? Patrick Bossuyt (University of Amsterdam) introduced this concept at the “Evidence Live” conference. Thinking about diagnostic testing he suggested that, rather than focus on the nuts and bolts of a diagnostic test (sensitivity, specificity etc), we might also consider the downstream effects. We know there are benefits to accurate diagnostic testing, but there are also potential harms. In this context, he discussed karyotype testing in recurrent miscarriage and how it might alter a couple’s decision on a further pregnancy and may, for example, dissuade them from considering extending their family. Patrick emphasised the need for further research on the consequences of diagnostic testing but, Richard Saitz (Boston University) reminded us during questions, that the consequences will always be confounded by the prior decision.  

Terry Klassen (Manitoba Institute of Child Health) had studied decisions on diagnostic CT scans in children with head injury. While the decision to include a CT scan may be based on the validity of the test, we seldom think of the consequences—exposure to radiation and possible association with later cancer. The potential harms of diagnosis and subsequent intervention were also highlighted by Ruth Gilbert (University College London) from her work on identifying maltreatment of children. If maltreatment is identified, we first think about removing children from this environment but, where are they placed and is this alternative better? She described huge variation in the levels of out of home care—we don’t know the optimum level.

Two major figures in UK medicine gave interesting insights as they come to the end of their terms of office. Mike Rawlins, with five days remaining as chairman on NICE, may have surprised some with his sympathy for the difficulties faced by the pharmaceutical industry. Developing new agents is becoming more expensive and increasingly complex. Reflecting on moves to encourage transparency in pharma research, he was certain that we should see the results of all clinical trials. He was less certain about the necessity for final study reports to be in the public domain—as these can be enormous documents. And, he had some concerns about the disclosure of individual patient data; they may be reanalysed by opponents or pressure groups using different and perhaps inappropriate methods and who may argue different conclusions. And, he was worried about the privacy and anonymity of participants.

Clare Gerada, chair of council of the RCGP, discussed the forthcoming Health and Social Care Bill for England: She suggested that the bill was based on three types of evidence: Inconclusive evidence, evidence to the contrary, and no evidence at all. In policy making, she reflected, ideology trumps evidence.

Domhnall MacAuley is primary care editor, BMJ.