It does raise the issue of the intra- and inter-observer reliability, however. Tests conducted by Öhberg et al (42) showed excellent intra- observer reliability, which increases with the degree of experience of the radiologist. Sengkerij et al (45) reported excellent inter-observer reliability for power Doppler for detecting neovascularisation. There are several reasons why not all Achilles tendinopathy tendons show neovascularisation in these studies. Richards et al (44) theorised that neovascularisation might go undetected in tendons that are less than 6.5 mm thick. As the other studies included painful tendon thickening among their diagnostic criteria for Achilles tendinopathy, but failed to state the actual tendon thicknesses, it is not possible to test this theory. Furthermore, among the studies

that showed 100% neovascularisation, Öhberg et al (28) scanned patients before asking them to participate in the study, and Alfredson et al (29) verified the diagnosis using ultrasound. In both cases, it is unclear just what was found on ultrasound – if the presence

BE USED AS A CARDINAL DIAGNOSTIC SIGN. UNFORTUNATELY MANY OTHER STUDIES HAVE FOUND THIS NOT TO BE THE CASE

of neovascularisation was part of the diagnosis (and subsequently part of the the inclusion criteria) then both studies were subject to selection bias.

Assessing progress of therapy To determine clinical improvement, most studies use the VISA–A questionnaire or visual analogue scales and patient satisfaction questionnaires as outcome measures. In contrast, Öhberg et al (42) were unique in using the degree of neovascularisation as part of their ultrasound assessment of pathological features of Achilles tendinopathy. They graded neovascularisation according to three levels: n 0 for no vessels, n (+) for 1–2 vessels n + for several irregular vessels.

They concluded that eccentric exercises were an effective therapy for Achilles tendinopathy and that a normal tendon appearance on ultrasound (together with the absence of neovascularisation) is associated with clinical improvement. Such conclusions appear reliable because the sample size was adequate and because the methodology, ultrasound equipment and techniques were identical to the other studies. However, future research is needed to add support to their findings. There is no consensus on

how to quantify the degree of neovascularisation. Boesen et al (46) expressed neovascularisation as a pixel percentage, and Peers et al (43) expressed it as surface area of the signal on the power Doppler

WANT TO ADVERTISE IN

IF PROVED TO BE CORRECT, THE ABSOLUTE PRESENCE OF NEOVASCULARISATION COULD

Call Tor Davies on

+44 (0)845 652 1906 or email tor@sportex.net

28 sportEX medicine 2010;45(Jul):23-30