I wrote to all 99 health authority chief executives in December, asking for a copy of their annual report to see what they are doing on patient and public involvement (PPI), and look at the content and consistency of complaints reporting.

By mid-February, 72 chief executives had responded, identifying their lead person on PPI. There were 70 different job titles, only eight mentioning public involvement and only one mentioning patients.

Complaints reporting is so inconsistent it is impossible to make meaningful comparisons.

Five of the 57 annual reports I was sent contained no information about complaints.

Five reported complaints against GPs, dentists and other family health services providers, but failed to say whether they had received any complaints against themselves. The numbers of recorded complaints against HAs ranged from a proud 'none' in two, fewer than 10 in nine others, to a hefty 352 in another. But this turned out to be the one honest enough to report the number of contacts with its complaints officers - not necessarily complaints about the HA.

The main problem in quantifying the number of formal complaints to the NHS - probably the tip of the iceberg, since the system puts so many people off - is the lack of systematic reporting of complaints that end in local resolution. These must be vastly greater than the number taken to the next stage of written complaint to the HA, yet how do we know how many or what they are about?

In my survey of over half the HAs in England, 9,759 written reports were against trusts, 7,666 against GPs, 1,864 against dentists, 93 against pharmacists and 43 against opticians. Even allowing for the absence of reporting formal complaints resolved locally, these figures are meaningless, since only 17 of the 57 HAs gave figures for written complaints against trusts and some lumped together all their family health services complaints.

At the next stage of the complaints procedure - request for an independent review - reporting remained variable, with a few HAs giving no information at all. Others reported in such different ways that comparisons were meaningless. In the 57 reports I examined, the numbers did not stack up at all.

Fifty-one HAs reported 725 requests for independent review; 141 were referred back for local resolution, 119 were refused, 24 were 'investigated', 'established' or 'granted' and 119 independent review panels were held. Seven were withdrawn and 25 recorded as 'no further action needed'. This leaves 290 unaccounted for.

None of these figures reveals anything about what happened to people who failed to get their complaint locally resolved and took it further in a formal way.

What happened to the 141 referred back? How do the 119 who were refused feel? What about the other 290?

We will never know the extent to which NHS complaints procedures work unless the information collected is meaningful and consistent. If the new patient advocacy and liaison services are to fulfil their potential, they will need a system for recording the sort of problems they help solve, alongside any new system of recording formal complaints yet to be announced following the independent review now presumably with ministers. Just as important will be genuine feedback from HAs and others about the changes they have made as a result of patients' complaints, formal or otherwise.

This was singularly lacking from most reports I read.

Since so many complaints about every aspect of the NHS revolve around poor communication, my prize for best practice goes to Birmingham Heartlands and Solihull trust. It developed a 'patient passport' as a framework to provide patients and carers with bite-sized pieces of information at all stages of their care pathway, after consultation with patients and staff demonstrated that poor communication was at the root of almost all complaints.

Patients are given two sorts of information: 'clinical' and 'comfort'. The first, based on evidence, is about diagnostic tests, treatment options, alternatives and risks, care during and after treatment and discharge information. The second is about how to find out where you need to go, what to expect when you arrive at the clinic or ward and who you will meet.

The information is in plain English and provided in advance so patients have time to prepare, reducing fear and vulnerability.

Each subject is covered in a short, self-contained leaflet.

This project's initial phase cost just£6,000, and since it also produced a guide for staff on how to develop such patient information, it ought to be able to recoup the cost immediately by sharing its good practice with other trusts.