Health Statistics User Group

Subject: RE:Accessing ONS births and deaths data

1.  RE:Accessing ONS births and deaths data

Posted 16-10-2013 10:36
A couple of years ago I carried out a similar exercise, except that I calculated DSRs rather than SMR so the practices could be validly compared with each other. We also had a couple of seriously anomalous practices that warranted further investigation, and the central issue turned out to relate to registration of residents of terminal care and nursing homes.  The PCMD was originally established to enable just this sort of analysis, in the wake ot the Shipman scandal. No-one has said that in the new world PH depts no longer have a duty to monitor GP mortality (which would presumably mean the NHSIC or PHE doing the centrally, for which a case could certainly be argued), and not to disclose the data to practices giving cause for concern woudld be the equivalent of not sharing prosecution evidence with the defence in a criminal trial.  I cannot imagine the intention of the current regulations is to make such investigation illegal.

The data is in fact being used here to produce aggregate statistics. The matter in dispute is the validity of the stats. Anyone with any serious background in analysis knows that producing statistics is not just a matter of crunching formulae, but must involved data cleaning and validation for the results to have any credibility. One could argue the the "public interest test" where "the benefit to the public of processing the information outweighs the public good of ... confidentiality"  (Caldicott 2 p53) provides a legal basis for this process.

An alternative would be to ask the GP Practice involved to provide the PH dept with a list of NHS numbers they believe were their registered patients who died, match these against the PCMD, and only share with the GPP the NHS numbers of those on the PCMD and not the GPP's list. There should be no need to share any other data than NHS number, no need for name and address etc.