This site is intended for health professionals only

At the heart of general practice since 1960

Read the latest issue online

CAMHS won't see you now

Your partner failed to tell patient of raised Blood sugar and now she has retinopathy

Three GPs share their approach to a practice problem

Three GPs share their approach to a practice problem

Case history

Mrs Smith has come to see you on the advice of her optician who has noticed retinopathy. She says she is tired all the time; her blood pressure is raised. When you tell her she needs a test for diabetes, she tells you your practice partner checked her blood sugar last year. Looking back you are perturbed to see her random glucose was 12.8mmol/l; your partner had electronically annotated the result 'tell patient to make appointment to discuss'. When you question Mrs Smith, she tells you she'd assumed it must be normal as nobody had contacted her.

Dr Trevor Rees

'You have to be upfront with the patient and tell her last year's result wasn't normal'

Mrs Smith has diabetes until proved otherwise, and in retrospect, it would seem likely that she had diabetes when she saw your partner last year – although a single, random glucose of 12.8mmol/l, if Mrs Smith was asymptomatic, is not diagnostic.

However, she now has retinopathy, has symptoms suggestive of diabetes, and needs to be sorted out in terms of confirming the diagnosis. Assuming that a second, fasting glucose is elevated, then she requires referral to an ophthalmologist for a full assessment of her retinopathy, and her diabetes needs management within primary care. That's the easy part!

Next is the tricky subject of the failure to follow up the elevated blood glucose from last year. You have to be upfront with Mrs Smith and explain that the result last year wasn't normal, and did require further follow-up. As you were not the partner involved, it is difficult to know exactly how that consultation was left, but there might be some clues in the records kept by the partner who instigated the test.

We all have our own ways of dealing with abnormal test results – some GPs make all patients re-attend for results, others are happy, as in this case, to attenuate results with instructions on follow-up; others, still, depend on the patient to contact them for tests results.

If the practice has rigorous record-keeping procedures in place, then it should be possible to see from those records whether attempts were made to contact Mrs Smith either by phone or by letter as these contacts should be logged on her records. If neither is present, indicating the possibility that contact wasn't attempted, then efforts should be made to find out what went wrong.

Discuss the problem with the partner involved, and ask the practice manager to check the audit trail of the electronic result to see where the communication breakdown occurred. Mrs Smith deserves an apology, and is entitled to complain about the 'missed diagnosis' if it appears that no contact was made.

This would make an ideal significant event analysis for the practice for as we all become more paperless and depend on the vagaries of electronic communication, then this type of situation has the potential to be repeated on a regular basis.

Trevor Rees is a partner in a large training practice in Birmingham and a tutor at the University of Birmingham Medical School

Dr Joanne Harris

'It is the practice's responsibility to ensure the patient receives abnormal results'

My first feeling would be one of relief – that at least this had been discovered. For the patient now it is important to make a diagnosis. Although one random sugar of more than 11.1mmol/l can be used to diagnose diabetes, it would be better to order a fasting glucose to make the diagnosis.

I would also like to see her fasting cholesterol and HDL (since most diabetics require a statin to lower cholesterol) in addition to pre-treatment LFTs and U&Es. An HbA1c cannot be used to diagnose diabetes but can be useful at diagnosis to see the status of glucose control over the preceding three months.

Her BP has been noted to be raised but it is not known if she has previous high readings recorded on the computer. I may need her to come back for further readings. I would weigh and measure her and if her BMI is high, as is likely, I would offer appropriate dietary advice.

I would discuss with the patient that her blood sugar had been high at her previous test. Her tiredness and retinopathy suggest she has diabetes but we need to do more tests to confirm this. She may have family members already suffering from diabetes and have some idea of the diagnosis.

There can often be a problem with patients not ringing back for their results as requested. They then assume because they have not heard that everything is all right. There is often an element of denial on the part of the patient that does not chase up results. However, we have been advised by our LMC that it is our responsibility to ensure the patient receives abnormal results.

I would mention this case to the partner who had ordered the test and suggest it would be worth making a significant event report to discuss within the practice.

The optician has found retinopathy, although we do not know if this is diabetic or hypertensive. Since 50 per cent of diabetics have complications at the time of diagnosis anyway it is possible that the findings may also have been present a year ago. As one of the recommendations of the significant event report we could suggest that the person ordering the tests makes clear notes of the intended follow-up at the time of the consultation.

Also, as well as annotating the computer, we could write a note to one of the administration staff asking them to telephone the patient to make an appointment to discuss their results.

Hopefully these measures would prevent similar situations happening in the future.

Joanne Harris is a part-time partner at a practice in Ealing, west London, and also teaches medical students from Imperial College, London

Dr Richard Stokell

'Assuming the patient was meant to call in for results, she bears some responsibility too'

I would tell Mrs Smith her test last year did actually show that she was probably developing diabetes then. I would apologise to her for not informing her of the test and suggest I need to look into how that has happened.

I would then want to repeat her fasting blood glucose to confirm the results and organise fasting lipids, biochemical profile and so on in preparation for booking her into a new diabetic appointment with our diabetic nurse. It is probably too soon to start treating her elevated blood pressure as we have only one isolated reading.

She needs advice straightaway about diet and she probably needs an ophthalmic referral at this stage, depending upon the details of the optician's report.

I would discuss what has happened with the partner involved and would expect this to go on to a significant event analysis. This would give us a chance to look at our system for making sure patients are told about their abnormal results.

No practice system is infallible and it is likely that the partner involved has marked the result and filed it away without passing it on for action by the receptionist in this case. Providing the patient was advised to telephone to check the result, she also bears some responsibility for not having been informed of the abnormal result.

Whether the patient formally complains or not, it is likely that a further discussion will take place with her about this. Being able to demonstrate that we have had a full discussion about the problem and have improved the practice system to reduce the risk of recurrence is likely to go a long way towards appeasing her.

Richard Stokell is a GP in Birkenhead, Merseyside

Learning checklist

Ordering investigations – where does it go wrong?

• Patient is advised to make an appointment to have blood tests but does not do so (because of fear, misunderstanding, difference of opinion, forgetfulness, lack of available appointments).

• Patient is told that a referral for tests (say ECG,or X-ray) will be made. Doctor forgets or referral goes astray en route; patient assumes doctor changed mind and does not follow up.

• Result fails to come back – doctor fails to notice, patient assumes no news is good news and does not follow up.

• Abnormal result comes back electronically or by post and is inadvertently filed without action, is seen by doctor who assumes patient has been told to phone in, or doctor leaves message that does not reach patient. Patient assumes no news is good news and does not follow up.

• Practice protocol allows automatic filing of normal results by computer or staff who may not recognise when a normal result may be highly significant.

• Normal result comes back to another GP who is unaware of management plan/need to follow up (for example, a normal folate level may only be part of anaemia investigations)

• Hospital or other clinician orders investigation and passes result to GP; GP assumes requesting clinician will have arranged action.

What safeguards do practices need?

• The doctor responsible for requesting an investigation is responsible for ensuring that the patient is informed of any abnormal results. Undertaking to inform every patient of every result is logistically and financially impractical for most practices, although the national cervical cytology and mammography screening programmes do it.

Practices should:

• Routinely advise patients which tests they have had and how to contact the surgery for all their results (telephone, in person or by making an appointment). This advice should be recorded in the notes.

• Ensure they have a valid and current contact number for the patient; this should be written on the request form so that, for example, the patient can be contacted promptly (through the out-of-hours service if necessary) about seriously abnormal results such as a high serum potassium.

• Keep a record of all blood tests, X-ray requests and referrals sent by the practice so that missing results can be identified and pursued (keep these on the computer to allow for easy searching).

• Have a failsafe protocol for processing incoming results and informing patients whose results are abnormal, or require action even if normal.

• Safety-net referrals by advising patients to contact the surgery if they have not heard anything within a specified time.

• Audit these processes.


• If a mistake has potentially serious consequences it may be better to consult your defence society before taking action.

• Where a mistake has been made, patients are entitled to a full explanation, an apology and rectification where possible.

• Patients should be also informed about the NHS complaints system.• Where there has been a mistake or a near-miss, the practice should conduct a critical event analysis, which may flag up procedural, training, performance or attitudinal issues.

Melanie Wynne-Jones is a GP in Marple, Cheshire

Rate this article 

Click to rate

  • 1 star out of 5
  • 2 stars out of 5
  • 3 stars out of 5
  • 4 stars out of 5
  • 5 stars out of 5

0 out of 5 stars

Have your say