In response to my commentary on payment by results in smoking cessation services , Steven Wyatt's letter  describes a system set up in a region of England, the West Midlands which, it is hoped, has in place the safeguards needed to avoid the problems to which I referred. It is very useful to have this case from which others can learn. There are three points made in the letter that I would like to take up.
First, the letter argues that paying only for success above a threshold (to pay for quitters who would not have quit anyway) would drive up transaction costs, but I do not think that is the case. The same system for payment would obtain as currently, but only counting numbers above the threshold. The West Midlands system, that only specifies a minimum threshold, incentivizes providers to just beat the threshold rather than to do their best for every smoker. Moreover, penalizing providers after they have failed to achieve some target is highly problematic administratively, and does not help the smokers who were let down by the poor performance.
Secondly, the issue of accreditation can often consist of a very bureaucratic paper exercise which does not necessarily address the important issues. In practice the process can be very streamlined, involving a brief self-assessment questionnaire and submission of key supporting documentation. Unfortunately, commissioning organizations are often not incentivized to keep things focused and simple. It would be useful to have a repository of such procedures being used by different commissioners, along with comments from providers who have engaged with them to work towards maximum efficiency in this.
The third point is possibly the most important. The big question that did not appear to be addressed directly in the letter is: what independent checks are carried out on claims of successes? Those checks need to be undertaken by an agency that is independent of both commissioner and provider because both have an interest is showing that everything is fine. It is imperative not to rely upon data provided by providers, however detailed it may be. A procedure is being piloted by the National Health Service (NHS) Centre for Smoking Cessation and Training (http://www.ncsct.co.uk) that is simple and inexpensive: one takes a sample of 50 claimed successes and contacts the clients to ask a few simple questions. If doubts arise over more than 10% of these one expands the sample by a further 50. If 10% still show a discrepancy then an investigation is triggered. There need be no ‘data protection’ or confidentiality issues if a few simple procedures are put into place at the start.