Bad Medicine: Bad data

Lloyd George medical records, an impossible feat of origami, a kilo of paper stuffed into tiny envelopes. There was only one copy and half the time it was misfiled or lost under the piles of paper and Pharma tat of stress balls, plastic joints, and broken clocks that littered our desks. Anyway, it was full of illegible one-word ink entries. But we managed. These were the good old days; confidentiality by default, because no one could actually find information even if they wanted to. Even our hospital referrals were handwritten or dictated affairs, with doctors consciously deciding what to include in the letters.

But today I slowly click legible (if misspelt) entries with one sausage finger, bitterly regretting not taking secretarial studies at school. Young doctors, however, now type paragraphs on every contact, conditioned to record everything in our modern medical legal neurosis. The relevant obscured by the irrelevant.

Electronic records have allowed the production of searchable medical summaries. The summary content varies widely between different GP surgeries, some including every minor event, and others only previous significant illness. This is the new one-click information era where we can find and organise care like never before. The electronic record also means we can access records on multiple computers, remotely on mobile electronic devices or shared between agencies. But this great advance is also a great threat to confidentiality.

Some of this threat is addressed in the Caldicott report,¹ which considers inter-agency data sharing, but there seems no guidance on data sharing between doctors. Yet when it comes to electronic referrals and printing out paper summaries, all these summary data are automatically dumped into them. These unfiltered medical data are often irrelevant to other healthcare professionals but are both highly personal and confidential, and are not in patients’ best interests to share (more importantly, paper summaries have inadvertently ended up in the hands of family members). Patients often don’t realise what was considered a ‘confidential consultation’ remains on their record forever.

Consider termination of pregnancy: this remains a divisive and stigmatising issue for doctors² and society³ but is frequently included on the medical summary list. What might a chance disclosure on a computer summary have on family relationships? Yet termination carries no long-term medical sequelae and would have no impact on potential medical care,⁴ so why should it be included on a medical summary at all?

This concern over confidentiality is also true of other stigmatising issues, like sexually transmitted diseases, sexual abuse, fertility issues, mental illness; conditions that have no impact on most ongoing hospital care. Should this information even be disclosed in communications between doctors? This is especially a concern for doctors, whose confidentiality is frequently undermined by the pernicious medical gossip mill that afflicts our profession.

What is to be done? There is a strong argument that ‘termination of pregnancy’ should not be recorded and should be systematically removed from the current GP clinical summaries. Similar consideration should be given to conditions like chlamydia and genital warts. GPs need to be more reflective on what is recorded in the summary and debate national standards for data recording. When referring to hospitals, only what is relevant to ongoing clinical care should be included, and what is not should be actively edited out.

The electronic record has been the great step forward for health care but remains the greatest threat to medical confidentiality. And medicine without true confidentiality is bad medicine.

• © British Journal of General Practice 2016