Quality must be caused, not controlled 1 We are generally more familiar with the ISO (International Organization for Standardization) acronym as consumers than as doctors and even more so than as sonologists. The ISO has been an international standard-setting body for over 60 years. Its ability to set standards that often become law, either through treaties or through national standards, makes it more powerful than most non-governmental organizations 2 . Although the ISO dictates that no fewer than eight steps, including more than 20 standardized procedures using specifically designed instruments, are taken for testing condoms 3 , only 10 procedures in medicine have been granted ISO certification. The closest the ISO has got to ultrasound in medicine is via the quality assessment of workflow and data management in health informatics and digital imaging and communication (DICOM) 4 .Quality assessment studies in prenatal sonography have been based largely on detection rates of fetal abnormalities 5 . Although this is an important issue, congenital anomalies are encountered in only 2% of all pregnancies, and such a quality control process cannot realistically be applied to every operator due to the varying frequency of their exposure to anomalies. Antenatal ultrasound has been subjected to several guidelines from national and international scientific bodies and professional organizations, most of which target specific and specialized clinical objectives in which reproducibility is rarely assessed or achieved. However, this has not prevented an increase in malpractice lawsuits for missed fetal anomalies, now the most common type of litigation involving ultrasound 6 . Poor performance at scanning may well contribute to our failure to objectively establish the effectiveness of routine ultrasound in decreasing the perinatal death rate 7 . Quality assessment in prenatal ultrasound should therefore aim to go beyond record-keeping of ultrasound examination reports and images and clinical outcomes. It should address both technology-and operator-related issues, because these two components of the procedure cannot be dissociated.
The definition of quality is conformance to requirements 1Monitoring the performance over time of an ultrasound machine and its auxiliaries is beyond the capability and interest of most practitioners. Physicists have produced specific equipment to assess the quality of ultrasound machines and have obtained informative and reproducible results. However, sonologists have shown no more than a passing interest, and buying a new machine every 5 years together with an expensive maintenance contract is the contribution of most of them to technical quality assessment. However, there is an obvious conflict of interest and the virtuous cycle should therefore include internal quality assessment by the manufacturer as well as external quality assessment by an independent body. Quality assessment should incorporate performance tests that are most likely to detect common system problems. These include assessment ...