Where do we start? Evidence-Based Medicine and antibiotic prescribing.

Fahima was my third patient of the day. I was running almost on time and relatively unstressed.

She bought in two of her children with her, and opened the consultation by saying that she had made the consultation for herself, but now the girls were ill, so she wanted  me to see them instead. This happens quite a lot. I also have patients cancel or fail to attend appointments if they feel too ill to attend or refuse home visits if they don’t feel well enough to make themselves or their homes presentable.

Like many GP practices, we still have ten minute appointments for each patient even though they might have one or several, straightforward or complex complaints. Now there are two, or possibly three patients who may have three related or otherwise, simple or ill-defined problems to work through. Medicine is a practice that continually involves decisions about what to do in unexpected situations like this.

The practice of medicine is cultural, social and philosophical and far more besides being scientific or ‘evidence based’. My decision to attempt to deal with Fahima and her daughters in a single appointment is in part driven by anxiety, ‘what if I miss something serious?’ In part it is pragmatism, ‘I’m pretty certain they’re here with straightforward, related complaints – in fact, they all appear to have colds, I know them and I can hopefully deal with them effectively in the time I have’. I hope that I can use the opportunity to invest some time exploring Fahima’s concerns, that might then reduce their likelihood of attending with the same problem next time. It’s also moral, ‘I’m the kind of doctor that helps patients when they need me, so I’m not going to send them away to book another appointment with another doctor or go to A&E’. These decisions, conscious, but implicit are also evidence-based. There is evidence that continuity of care and exploring parental anxiety improves the safety and effectiveness of care. But I also happen to be the kind of doctor that feels guilty for making subsequent patients wait when this consultation takes longer than ten minutes, and I wonder how harmful this might be.

I know Fahima and her daughters well. I know that her husband used to beat her and they were separated but now he has come back, and though she denies it, I’m not sure he isn’t beating her still. Social services are involved. I know that she has struggled to look after her children, that they are under-nourished and have both been in hospital with chest infections and their asthma and eczema is under-treated despite input from community paediatric nurse specialists. I know that Fahima has panic attacks and her oldest daughter Aisha, who is only 9, has missed school a lot to look after her mother. They have a constantly changing stream of relatives and relative-strangers in their damp, two bedroom flat, many of whom sit up at night, smoking and playing cards with their dad. I’m wondering how ‘evidence-based medicine’ can help me here.

The modern pioneer of evidence-based-medicine (EBM) David Sackett, described EBM in 1996 as the integration of individual clinical expertise with the best available external evidence and the patient’s values and expectations. 

EBM = clinical expertise + external evidence + patient values/expectations

Sackett was well aware of its limitations,

Without clinical expertise, practice risks becoming tyrannised by evidence, for even excellent external evidence may be inapplicable to or inappropriate for an individual patient.

There is a now a critical campaign for a renaissance of ‘real EBM’, headed by Prof. Trish Greenhalgh. According to these critics, the evidence in EBM is misappropriated by vested interests, excessive in volume, of dubious significance, unreliable in the kind of complex patients seen in practice and management-led rather than patient-centred. All of which might be true, and I suspect they are right. After all, doctors are not very good at understanding the statistics on which most evidence is based, and it does seem that evidence is still seen by many health professionals as being authoritarian rather than facilitative. But evidence is only one part of the equation. Much of their criticism focuses on the unreliable external evidence squeezing out clinical expertise and patient values. There may be more fundamental problems.

For one thing, it’s not clear why clinical expertise is on the right side of the equation – as if it is somehow separate from evidence and patient values rather than the ability to integrate them. Why not,

Clinical expertise = external evidence + patient values?

More importantly though, patient expertise and clinician values are excluded despite it being obvious, if under-appreciated that patients are experts and clinician values influence care. Greater attention to clinician values and patient expertise is almost certainly key to understanding variations in care.

A better formulation might be,

Clinical practice is the integration of the best evidence with the combined values and expertise of patients and clinicians.

The degree to which the expertise and values of both parties is drawn upon varies considerably according to changing contexts such as how strongly values are held and how much expertise each party has, or is willing to share. Many patients have strong opinions for or against antibiotics, and some have suffered unusually severe throat infections that take significantly longer than average to resolve and know from experience that antibiotics are effective.

Qualitative research can help to reveal hidden values and expectations. In one study from 1976, GPs were found to be far more likely to prescribe antibiotics if they expected patients to have difficulty getting to the practice, if they had an exam or travel commitments the next day or if a sibling was in hospital with pneumonia. I used to look after an opera singer who expected antibiotics within hours of every sore throat, and was very upset when I attempted to challenge this. I was relieved to read in a study published this year that I wasn’t alone in prescribing antibiotics to avoid an unpleasant confrontation with patients. The same study raised the point that GPs tend to over-estimate how often patients want antibiotics and perhaps we expect conflict too often. Almost certainly if we are stressed this is more likely.

Stress and burnout is a serious problem among GPs and is associated with a reduced ability to tolerate uncertainty, for example the point at which an antibiotic prescription may be appropriate. Many GPs will admit in private that they are more likely to prescribe antibiotics at the end of a busy clinic that is running late than at the beginning of a clinic that is running on time. In part this is because prescribing is quicker when delivered with, all you need to do to get better is to take the pills, rather than explaining why antibiotics are not required and discussing how else the patient (or parent) might mange their symptoms. Obviously conversations about what else can be done should happen whether or not a prescription is issued and discussions about why antibiotics are prescribed are just as important as discussions about why they are unnecessary. A point we are well aware of but might skip over in order to keep to time.

One way that doctors try to avoid conflict is to use the clinical examination to build a rapport with patients and the strength of the rapport influences whether they are able to take control of the decision to prescribe. The clinical examination is often used as a time for reflection or an opportunity to uncover a patients’ hopes and fears. In another study from 2002, GPs who were least likely to prescribe were more likely to be older, to spend longer with patients and be more interested in their relationships with patients. Those who prescribed more often described their role with patients in terms of a ‘professional service’ or a business exchange. Interestingly they were also more likely to describe themselves as being ‘firm believers in evidence based medicine’ than their peers who prescribed least. Low cost prescribing doctors had a more relaxed attitude to evidence, being less likely to attend educational meetings and were less concerned with labelling symptoms with a diagnosis.

Continuity of care enables doctors and patients to get to know and trust each other and also increases the likelihood that antibiotics will be prescribed prudently. Continuity and mutual trust can make a brief consultation successful, but lack of continuity can eliminate the effects of knowledge and professional skills. One reason parents expect antibiotics when they, or their children have a cold or a sore throat is that they lack confidence in their ability to cope. A known and trusted doctor is more likely than a stranger to give them the confidence they need.

All of this takes place in a wider social context. I am only able to make a decision that combines evidence, expertise and values if there is a meeting between myself and Fahima and her children. A system of healthcare that enables people to see a GP without charge, when they need, makes this possible. The culture of my GP partnership values timely access and continuity of care and allows doctors the autonomy to see extra-patients where necessary and the time required to discuss values and share expertise. We provide protected time for supervision to help reduce the risk of burnout and give staff the opportunity to discuss difficult cases.

In a broader social context, national culture has a significant influence over decisions to prescribe antibiotics as well. A Dutch GP working over here thinks we’re terribly laissez-faire but we point out that we prescribe less than half as many antibiotics as the French.

Antibiotic prescribing rates by country

Why we prescribe antibiotics matters because of the growing problem of antibiotic resistance, where bacteria previously sensitive to the effects of antibiotics develop resistance as a result of excessive exposure.

If we think we (or the French) are to blame, then we need to look also at what is happening in developing countries. I have worked in Afghanistan, Nepal and India with doctors and nurses whose careers have taken them all over the world. A very common situation is as follows. A poor person goes to see a doctor or other healthcare professional with a cough. The ‘professional’ assesses their ability to pay for treatment and gives them a scrap of paper with a list including antibiotics, pain-killers, vitamins, antihistamines and possibly more and sends them to their brother’s pharmacy where they buy whatever they can afford. Often they won’t complete the course of antibiotics, so they give (or sell) them to someone else. In many, perhaps most poor countries it is possible to buy antibiotics direct from pharmacies, and during the avian flu panic a few years ago there was a roaring trade in antiviral drugs being sold from online pharmacies in rich and poor countries. The widespread use of antibiotics and antiviral in intensive livestock farming is also alarming. In the case of TB (tuberculosis) the problem of resistance is significant because antibiotic treatment has to continue for several months and is frequently interrupted by conflict, natural disaster and so on. Medications are frequently sold or exchanged for other medications, food, etc. Unsurprisingly totally drug resistant TB is now, officially a thing.

I don’t wish to ignore the  enormous efforts to circumvent these problems with directly observed TB treatment programs and attempts to provide Universal Healthcare Coverage in poor countries, but the damage is being done all the time.

The future for prudent antibiotic prescribing at home isn’t much brighter. A political agenda that is aggressively encouraging patients to consume and professionals to compete and be judged according to superficial measures of patient satisfaction, combined with a loss of continuity of care and the premature retirement of older GPs is worrying to say the least. External evidence – about when antibiotics should be prescribed-  is occupying too much attention in our efforts to make medicine more ‘evidence-based’.

Back in my own surgery I am faced with Fahima. Her experience of healthcare in Turkey was that colds were quite frequently treated with antibiotics and other medicines and it has taken me about two years to build the trust necessary to convince her that she and her children can manage without them. Nevertheless she is very anxious, both children have been sick recently and I had seen one of them with symptoms of a cold about a week before they were admitted to hospital with a chest infection and her confidence in my diagnostic skills has taken a blow. In fact my confidence in myself took a knock after that. She spends most of the allotted appointment time complaining about the overcrowding at home and how much it is contributing to her children’s ill health so we have almost no time to discuss the use of antibiotics. Most of the conversation takes place as I examine them all one by one, on the couch. I’m examining their behaviour as much as I am their lungs. When the examinations comes to an end, I am prepared to talk about antibiotics and self-care, even though I am running late, but Fahima is gathering her bags and getting ready to leave,

“They’re OK? Yes?” she asks, with her hand on the door.

“Yes, they’ll be fine” I say, taken by surprise.

“Thanks doctor, thanks for listening”


Time of day affects decision to prescribe antibiotics: JAMA Dec 2014 http://archinte.jamanetwork.com/article.aspx?articleid=1910546

16/10/2014 Prescribing of antibiotics by GPs in England fell by 4.5% between 2012 & 2103. https://www.gov.uk/government/publications/english-surveillance-programme-antimicrobial-utilisation-and-resistance-espaur-report

05/08/2014: Public Health England study into antibiotic prescribing in General Practice 

6 responses to “Where do we start? Evidence-Based Medicine and antibiotic prescribing.

  1. Very helpful discussion. But not assisted by a mathematically illiterate graphic. DDDs per 100,000 inhabitants per what? And why does it yield a percentage? 12% of a DDD per 100,000 inhabitants per year would be homeopathic.

  2. Please amend to per 1000 inhabitants. Comment still holds.

  3. Pingback: NICE *didn’t* say Punish GPs over antibiotic use – who did? | living in a microbial world

  4. Thank you for a very interesting piece highlighting the problems of how our primary care system is designed. From my point of view having studied a BSc in Health Sciences: Herbal Medicine, it is a shame that primary health care practitioners in the UK have little training in antibiotic alternatives for minor infections: so many plants and essential oils are anti-microbial and fulfill patient expectation as well as being clinically effective.

Leave a Reply to abetternhs Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s