Many readers of this blog will be familiar with the stringent protections which the Data Protection Act 1998 (DPA) affords in respect of personal health data (see further the definition of ‘sensitive personal data’ in s. 2 DPA). Thus, for example, if a data controller wishes to avoid contravening the first data protection principle (the fair and lawful processing principle) as and when it is processing health data, it must ensure that: (a) the particular processing is fair and lawful; (b) that it meets one of the conditions provided for in schedule 2 to the DPA and (c) that it meets one of the very narrowly drawn conditions provided for in schedule 3 to the DPA. If the processing is intended to serve the interests of medical research, the data controller will doubtless wish to look in particular at the condition provided for in paragraph 8 of schedule 3. That condition stipulates that the processing must be ‘necessary for medical purposes’ (which includes the purposes of medical research) and be undertaken either be ‘a health processional’ or ‘a person who in the circumstances owes a duty of confidentiality which is equivalent to that which would arise if the person were a health professional’. Of course, the principle which underpins this particular condition is that it is very much in the public interest that, subject to the test of necessity, health data be shared by medical researchers. A recent judgment of the European Court of Human Rights (ECHR) has highlighted the importance of this particular public interest: Gillberg v Sweden (application no. 41723/06).

In Gillberg, two researchers requested access to health data which had been accumulated by Professor Gillberg as part of a long-term project on hypheractivity and attention deficit disorders in children which he was running out of the University of Gothenburg in Sweden. The University refused access on the basis that assurances had been given to the parents of the children and later the children themselves concerning the confidentiality of the data. The researchers challenged the University’s decision relying on Sweden’s long-established and generous rules on access to official documents. The Swedish administrative court upheld the researchers’ claim and ordered that the University disclose the data to them, subject to the imposition of strict conditions on their handling and use of the data. In reaching the conclusion that the data should be disclosed to the researchers, the Swedish court took into account not least the public interest in ensuring the independent and critical evaluation of medical research in the important field of neuropsychiatry. The data was subsequently destroyed by certain of Professor Gillberg’s colleagues. Thereafter, Professor Gillberg was convicted of misuse of office by the Swedish Parliamentary Ombudsman. Having lost his appeals against conviction in the national courts, Professor Gillberg took his case to the ECHR claiming that the conviction breached his Article 8 and 10 rights, particularly in view of the assurances of confidentiality which he had given to the data subjects and their parents. The ECHR dismissed Professor Gillberg’s appeal. It found that, even if the conviction interfered with Professor Gillberg’s Article 8 right to privacy (i.e. his right to privacy in the context of his professional affairs), that interference was justified in the circumstances. It also found that there was no interference with Professor Gillberg’s Article 10 right to freedom of expression as he was convicted not for giving assurances of confidentiality but rather because he misused his office in response to the judgments of the court.

The ECHR’s judgment is interesting not least because it confirms that, at least for the purposes of human rights jurisprudence, the fact that promises of confidentiality have been given to individual patients/research subjects does not create an automatic bar on disclosures which may breach those promises, particularly where the disclosures serve important public interests such as the interests in protecting the integrity and progress of medical research. Query whether the same result would have obtained on an application of the principles embodied in the DPA, particularly in view of the relatively permissive approach to disclosures for the purposes of medical research contained in paragraph 8 of schedule 3.


Sharing patient information in the NHS has proved highly controversial.  We posted about this subject here a while back.  Now there’s a new report from UCL researchers, suggesting that two key recent NHS IT programmes for handling patient information have so far delivered only modest benefits.   A short summary appears here, with links to the executive summary and the full report.  A research paper based on the findings has been published in the BMJ.

The three year UCL project looked at the Summary Care Record (SCR) and at Healthspace, both introduced as part of the NHS National Programme for IT. 

The SCR is an electronic summary of key health data, taken from GP records and other sources, and available to a range of NHS staff.   According to the UCL report, very few people had chosen to opt out; less than 1% of those who had been sent the relevant information.  But SCRs were not yet widely used; even where available, they were only accessed in 21% of clinical encounters.  So far there was little evidence that SCRs improved patient safety or reduced consultation length or hospital admissions.

HealthSpace is a tool that allows patients to update their own health information, plan healthcare appointments, and contact their GP via a secure internet connection.  So far, take up has been very low.  According to the UCL study only one person in 200 who was invited to open a basic account did so, and only one in 1000 opened an advanced account.

The report’s lead author, Professor Greenhalgh, is quoted as saying:  “This reseach shows that the significant benefits anticipated for these programmes have, by and large, yet to be realised – and that they may be acheived only at high cost and enormous effort … It serves to demonstrate the wider dilemma of national databases:  that scaling things up doesn’t necessarily make them more efficient or effective.”