Sunday, 24 June 2018

Decision Fatigue and What to Do About It - When to Use Antibiotics for URTI, AOM and Tonsillitis in Children

Recently I was speaking to a GP colleague about the ways to protect oneself from decision fatigue.  Decision fatigue is a serious issue for anyone in a high volume, high turnover medical job.  He had some great insights into the problem and the solutions.

What are the effects of decision fatigue?  In the short term, your decision making ability gradually declines.  In the long term there is a risk of burnout.  From your patient’s point of view, your fatigue could mean that because you have already made too many decisions, you will not make the right decision when it really matters.  It is possible that this could lead to harm to a  patient.  Decision fatigue affects our ability to show compassion or provide patient centred care.  Subconsciously we protect ourselves from too many decisions by caring less and being more directive.

My GP friend’s solution to all of this was elegantly simple: make fewer decisions.  His rationale was this: there is only so much that we can give and we need to choose when to use our decision making energy.  If decision making is a finite resource then to use it indiscriminately is could even be seen as irresponsible.

So, how do you choose what to stop deciding?  Well, I would start with a commonly occurring dilemma that creates a great deal of uncertainty.  How about antibiotics for sore throats and ears in children?

You will notice I don’t talk about tonsillitis, URTI or otitis media.  These terms all imply an aetiology.  That is a presumption that is completely misleading.  Tonsillitis may be viral and red throat without exudate may be streptococcal.  The truth is that we don’t have a reliable way of discriminating between viral and bacterial aetiology when we examine throats and ears.  So we can't know who to give antibiotics to.  Rather than exhausting ourselves trying to get it right, perhaps we should just stop, but is that safe and justifiable? I am not the first person to ask that question. (1)

The decision that we are all faced with, to antibiotic or not-antibiotic, has to have a valid goal.  So the next question has to be, “What is the benefit in giving antibiotics?”

Do we give antibiotics to prevent complications?  In the UK this is not the case.  The evidence is very much against a need to give antibiotics as a way of preventing complications of URTI.  Antibiotic prescribing rates are falling and yet there is no crisis caused by increased numbers of invasive infection or the sequelae of streptococcal infection.(2)  Logically, if there was a quantifiable risk of complications related to reduced antibiotic prescribing, we would all have to justify each decision not to prescribe.  As previously mentioned, there is no reliable discriminator, so shouldn’t we be hearing from the public health authorities that we need to be more proactive in our antibiotic prescribing.  That’s not the message we are getting at all.  Why?  Because prescribing antibiotics for sore throats and sore ears in children (in a country with a low prevalence of complications such as rheumatic fever) is not part of a strategy for prevention of secondary infection, invasive infection, sepsis or any other complication.(3)

Should we be giving antibiotics to control symptoms?  Let’s look at that as a reason to prescribe antibiotics.  What are the facts?
  • The odds of antibiotics helping the symptoms of any one child are low.  The actual number varies by age, study and whether we are talking about ear or throat symptoms but they are all in the same region.  The odds of benefit are in the region of 10-20%.  
  • Decision tools such as Centor and FeverPain are designed to improve the odds that antibiotics will help symptoms but there are  major problems with these aids.  Firstly, they are not validated in the younger children who account most of the presentations of sore ears and sore throats.  Secondly, these tools imply a binary outcome.  If you score above a certain number, antibiotics will help right?  Wrong.  A high score means slightly less awful odds that antibiotics will help.  Again, that is only validated if your patient is an older child. (4,5)
  • Rapid antigen testing has been validated as a way of reducing antibiotic prescribing but has not been shown to have a high sensitivity from the point of view of directing treatment to where it is effective.  These two things are very different. (6)
  • There is a significant harm done by antibiotics in children.  Depending on the antibiotic and the study, the odds of making a child unwell (vomiting, abdominal pain, diarrhoea) with an antibiotic is 5-10%.  
So where have we heard 10% before.  Wasn’t it something to do with odds of benefit?  What would a statistician say if they looked at the odds of benefit and the odds of harm and saw that they overlapped.  In all truthfulness I couldn’t stay awake for the full answer but the gist was that there’s not a lot of point in such a treatment being used as a way to manage symptoms.
Finally, here are two things that make a nonsense of the whole question.
  1. Children often refuse the antibiotics we give them.  Phenoxymethyl penicillin in particular is disgusting and children tend to be quite discerning in their medicine preferences.  Often the outcome of a difficult decision over whether to give antibiotics is later made meaningless as the child decides for all involved that the antibiotics are not going to happen.  The parent, remembering that it was a choice rather than a must-do usually gives up the fight.
  2. The issue of antibiotics for tonsillitis and otitis media fails an important test: Snelson's Safeguarding Test.  It goes like this:  A parent brings a 2 year old to you with a fever and a cough.  You see exudate on the tonsils and are about to prescribe penicillin.  The parent says that they prefer not to treat their child with antibiotics.  You have confidently ruled out sepsis, meningitis and pneumonia.  What are you going to do? Get a court order to force the parent to give the antibiotics?  Refer the child to social services?  I don't think so.
So if the parents and the child are allowed to refuse antibiotics for sore throats and ears, how important can they be?  We wouldn't allow these barriers to get in the way if the child's life was at risk or even if the child was going to suffer as a result of non-treatment.  This way of looking at it is a good way of identifying the children who should be having antibiotics:
  • Children with severe symptoms despite maximal analgesia
  • Children with complications of URTI (such as infected lymph nodes)
  • Scarlet fever (typical rash and oral inflammation alongside pharyngitis/tonsillitis and febrile illness) implies a more pathological strain of steptococcal infection
  • Children with prolonged symptoms e.g. no signs of improvement after five days of illness
So next time you see a child with URTI, ask yourself, could I insist that this child should have antibiotics?  If not, save yourself a decision.  You know it makes sense.  All we have to do is convince the parents that this is the right thing to do.  (more on that very soon)

Edward Snelson
Disclaimer: I was replaced by a robot three years ago.

  1. Morton P. Should we treat strep throat with antibiotics? Canadian Family Physician. 2007;53(8):1299.
  2. Kvaerner KJ, Bentdal Y, Karevold G., Acute mastoiditis in Norway: no evidence for an increase, Int J Pediatr Otorhinolaryngol. 2007 Oct;71(10):1579-83. Epub 2007 Aug 20.
  3. NICE, Sore Throat (acute): Antimicrobial Prescribing, NG84, January 2018
  4. Little Paul, Hobbs F D Richard, Moore Michael, Mant David, Williamson Ian, McNulty Cliodna et al. Clinical score and rapid antigen detection test to guide antibiotic use for sore throats: randomised controlled trial of PRISM (primary care streptococcal management) BMJ 2013; 347 :f5806
  5. Roggen I, van Berlaer G, Gordts F, et al Centor criteria in children in a paediatric emergency department: for what it is worth BMJ Open 2013;3:e002712. doi: 10.1136/bmjopen-2013-002712
  6. Little Paul, Hobbs F D Richard, Moore Michael, Mant David, Williamson Ian, McNulty Cliodna et al. Clinical score and rapid antigen detection test to guide antibiotic use for sore throats: randomised controlled trial of PRISM (primary care streptococcal management) BMJ 2013; 347 :f5806