Patients need to have control over their own information if is to work

The consent process for opting out of should be clear and transparent, and patients should be the first to have access

Surgeon in operating room writing notes

The data being shared by is substantial. It includes mental health data, diagnoses, medications and laboratory results. Photograph: Alamy

The case the government makes for, that allowing medical researchers access to patient data will result in new cures, is a hard one to oppose. If the case were so simple then most people, including me, would welcome it.

The question that the effort brings to light is a simple one – why has the NHS moved so quickly to provide patient medical records to third parties and so slowly to provide data to the patient themselves? For me, this is entirely the wrong way round.

The consent process for opting out of is neither clear nor transparent. The questions page states that the record will not contain information that can identify a patient – and it’s true that NHS number, postcode, full name etc will be removed. However, it’s also true that each patient’s record will be linked to a “meaningless pseudonym that bears no relationship to their ‘real world’ identity”.

In short, this means that if you knew the date and place of a famous footballer’s foot injury, there would only be a few records that showed such an event. Once you have this information, the pseudonym provides the rest of their notes – not so anonymous now. We know this is not a theoretical risk because a Harvard University researcher demonstrated last year that she was able to retrieve 35 patients’ records at $50 per patient from Washington State’s research database.

The information being shared by is substantial. They exclude sexual health data, such as HIV diagnosis or abortions – but they include mental health data. They exclude written notes – but they include actual diagnoses, medications and laboratory results.

In addition, the records will be shared with third parties a patient might not want to have their data. HSCIC public assurance director Dr Mark Davies has already said that, in the eyes of the law, the government can’t distinguish between a request by an insurance company and a medical researcher to access the database. For them, it makes no difference. But it makes a difference to the patient, who could, would and should distinguish. Indeed, at the moment the patient isn’t even allowed to see data in their own record that third parties will have access to.

Government representatives repeat in interviews that they need to open a debate on – so let’s have the debate now while plans have come to a temporary halt. We know from other countries that a proper discussion on government-controlled data leads to different outcomes. In the 90s, there was a huge public outcry about Iceland’s national genealogy database – a plan eventually quashed by the supreme court for violating the constitution.

It doesn’t have to be this way. In the USA, the federal government funded a series of patient-controlled registries for collection of data across sites. And in the UK, the Perinatal Institute has for years had a consent form in the maternity notes for using patient data in research. The midwife explicitly gets consent from the mother, and the mother can see exactly what is in the record, because she is the one who has it.

Further benefits flow from the Perinatal Institute’s patient-controlled approach. First, the notes carefully explain to the mother, on each page, what the medical terms mean – something they solved 10 years ago with a paper-based system and now is even easier to solve with IT. Second, they carefully set expectations about care at each stage of the pregnancy.

These benefits – control, explanation and expectation management – generate instant benefits for the mother. They also mean that when the patient is asked if their records can be used they understand why and what the benefits are.

The government says that can bring huge benefits. If so, it needs to prove what these benefits are and start by giving the patients themselves data about their own bodies.

Source – Guardian