In reply: We agree with Murphy and Redman that pre-eclampsia remains an important disorder and a major cause of maternal and perinatal mortality. However, we disagree with their interpretation of our data.

Murphy and Redman allege that, by eliminating routine urinalysis, a quarter of cases of pre-eclampsia may remain undetected for up to four weeks. As we pointed out in our article,1 three of the six women who developed dipstick proteinuria before they developed pre-eclampsia were already considered "at risk" for pre-eclampsia — two because of multiple pregnancies and one with a history of prior pre-eclampsia. These women would, in our practice, continue to have routine urine tests during their pregnancies. The argument is then whether it is justifiable to undertake repeated urinalysis in almost 1000 women to detect three who have dipstick proteinuria, but no other warning signs before the onset of their hypertension. In practice, we would have had to increase antenatal clinic visits for 338 women with dipstick proteinuria (most of whom will have had false-positive results2) to detect these three women with proteinuria before they developed pre-eclampsia. It is already our normal practice for women to have antenatal visits every second week in their third trimester. Therefore, it is just as likely that their blood pressure changes would have been detected by our routine surveillance as by the knowledge that they had dipstick proteinuria.

We did acknowledge clearly in our discussion that our study had a potential for type II error and that the best approach is a randomised controlled trial of outcomes between those who do and do not have continued urinalyses during pregnancy.

None of this belittles the importance of pre-eclampsia, nor the need for us to separate women considered at "low risk" from those considered "at risk" for pre-eclampsia on the basis of well recognised risk factors. The latter group should never be considered among those in whom routine urinalysis can be omitted.