This poses a particular problem. There is no accepted definition of recurrence. A recurrence may hold different value after different treatments. For example, a recurrence after PNF or CCH might be readily treated again with a simple needle technique or by surgery that has not been rendered difficult by the initial treatment. So even if the recurrence rate after CCH and especially PNF is very high, if that recurrence is easily, effectively, safely and relatively cheaply treated, then the value of recurrence as an outcome measure is diminished.

In contrast, recurrence after previous LF is a complex and more risk-prone procedure due to a scarred field compounding the recurrent DD. So a lower recurrence rate after primary surgery may be offset by the complexity of treating that recurrence. Liberal use of dermofasciectomy as the first operation of choice for DD, whilst being a little more complex, has a much lower recurrence rate so overall may be a more cost-effective procedure and, by minimising the chance of a more hazardous second operation, be an overall safer option.

An RCT requires funding. The cost is unlikely to be supported by a pharmaceutical company (which might be concerned that the drug would fair unfavourably, particularly if the study is designed to minimise the issue of heterogeneity by including only those cords suitable for either PNF or CCH). A state-funded trial would, in the absence of a robust protocol with the issues discussed in this chapter, be unlikely to receive funding in competition for the same funds towards a more generalisable study.

A randomised trial should, whenever possible, be blind to the procedure and assessment outcome from the perspective of both the patient and the investigator. This would present challenges when comparing PNF, CCH and LF. The original RCTs that lead to licencing of CCH used a placebo but there is no further need for a placebo study. A sham operation or sham PNF is not possible. The outcome measures could however be blinded from the investigator (but not patient) if, for example, a glove covers the scars which would reveal if surgery had been performed.

Although recurrence can discriminate between CCH, PNF and LF, the measurement of recurrence needs prolonged follow-up – probably 5 years. Such a long follow-up poses challenges with funding, loss to follow-up and currency of the study. A study would probably take two years to set up and fund, three years to recruit, five years to follow-up and two years to analyse and publish, so twelve years from inception. Healthcare systems, patients, surgeons and indeed the pharmaceutical company which has underwritten drug development, marketing and distribution costs might reasonably expect to use CCH in a considered manner without waiting for this duration.