So are antibiotics really necessary? What happens if you don’t give antibiotics immediately or give the patient a delayed prescription with clear instructions on when to take them? How many of these patients will come back later? How many will be admitted to hospital and how many might die?!
This type of study has been done before for upper respiratory tract infections (URTI) and urinary tract infections (UTI) and the evidence is clear, it is safe for the patient to wait before starting antibiotics. This delayed antibiotic strategy is now increasingly advocated by the National Institute of Health and Care Excellence (NICE). They recommend that if a patient presents with an acute URTI then a throat swab should be taken but that antibiotics should not be given until an acute bacterial cause is identified on culture, in particular the Group A beta-haemolytic streptococcus. For UTI, with lower urinary tract infections (e.g. cystitis), the recommendation is to take an MSU from patients and then treat any specific bacterium identified on culture with the simplest antibiotic to which the bacterium is sensitive e.g. Trimethoprim, Nitrofurantoin or Amoxicillin. “Yes, yes”… I hear you cry “…but that is a UTI or URTI, they are very different to LRTI, I’m not sure I’ll be changing my prescribing habits!”
So what has been the problem with antibiotics and LRTIs?
I suspect there are probably three main reasons why antibiotics have traditionally always been given to patients with symptoms of an LRTI:
- The doctor has been concerned that not prescribing might lead to long term harm or even death if they don’t prescribe
- The doctor may be concerned that the patient will complain or question their clinical judgement if the patient is eventually treated with antibiotics and then improves (irrespective of whether that improvement would have occurred without the antibiotics) and this can lead to doctor/patient conflict
- The patient may be concerned that they might have pneumonia and there is the misconception amongst some patients that pneumonia is always fatal without treatment
The prospective, observational study by Little and colleagues compares the outcome of three initial consultation treatment approaches to LRTI; no antibiotics, delayed antibiotics and immediate antibiotics. The delayed antibiotic strategy involves giving the patient a prescription and clear instructions about when to get this prescription dispensed e.g. if symptoms worsen or they don’t get better within a certain timescale. The measured outcomes were: rates of repeat attendance at the GP practice, admission to hospital or death within 30 days of the initial consultation. In total 28,779 people were studied; this is a big study!
What were the results of this study?
Strategy |
Repeat attendance at GP (%) |
Admission to hospital with LRTI (%) |
Death from LRTI (%) |
No antibiotic (25.5%) |
19.7 |
0.25 |
0.0 |
Delayed antibiotic (13.3%) |
14.1 |
0.29 |
0.0 |
Immediate antibiotic (61.2%) |
25.3 |
0.55 |
0.04 |
Although not statistically significantly different, the repeat consultation rate of only 14.1% suggests to me that patients with a delayed antibiotic prescription might have felt that getting the prescription constituted a “repeat consultation” and they therefore didn’t feel the need to actually see a doctor to do this i.e. the patient wasn’t feeling better so they went and got the prescription rather than making a second appointment to see their GP.
Unfortunately the study did not make clear how many of the delayed prescription patients actually took the prescription. It is entirely possible that every one of these patients left the consultation room and immediately got their prescription even though they were told to wait. This depends on how many patients actually do what their doctor advises them to do!
The higher admission and death due to LRTI in the immediate antibiotic group might suggest a degree of selection bias in that GPs may have been more inclined to give an immediate antibiotic prescription to certain patients they identified as “high risk of complication” although this was not apparent in the statistical analysis of the patient characteristics. This was not a randomised or blinded study. This could potentially have led to more patients at risk of a complication being in the immediate antibiotic group than the other groups.
In addition, this study showed that 60% of patients were prescribed an antibiotic for acute cough. A previous Cochrane review, of six randomised controlled trials, showed that the ONLY effect of antibiotics in this situation is to reduce the duration of cough by half a day from an average duration of 10 days. Therefore these prescriptions were unnecessary in the first place. It should be noted that a third of patients with a cough will continue to cough for more than three weeks irrespective of any antibiotic treatment and it is important that patients are warned that this is the case, which might help reduce repeat consultation rates. Many of these patients with a persistent cough have pertussis… but that’s a different blog!
So what does this study tell us?
The main take-home messages from this study are:
- Hospital admission or death is very rare in acute cough
- Antibiotics do not significantly reduce the risk of hospital admission or death in acute cough
- Not offering antibiotics or delaying antibiotic prescribing is a low risk option for managing acute cough in primary care
Okay, so that might not sound that exciting from a study of nearly 30 thousand patients but it really is! Just think of all of those antibiotic prescriptions that might have been avoided in this study group if no antibiotic or delayed antibiotic prescription options were taken. Then consider that respiratory tract infections are responsible for over 60% of all antibiotic prescribing in the NHS and you can start to see how no antibiotic or delayed antibiotic strategies could prevent a massive selective pressure driving antibiotic resistance as well as save the NHS millions of pounds in unnecessary prescriptions.
So next time you are seeing a patient with an acute cough (or maybe you are the patient with an acute cough) consider whether an antibiotic is required immediately or whether it is better to have no antibiotic or a delayed prescription for an antibiotic if there is no improvement. Why not explain the rationale and evidence to the patient and give it a go?
Reference
Antibiotic prescription strategies and adverse outcome for uncomplicated lower respiratory tract infections: prospective cough complication cohort (3C) study. Little P, Stuart B, Smith S, et al. BMJ 2017;357:j2148