Doxycycline in UK guidelines for hospital-acquired pneumonia: where is the evidence base?

Sir, Hospital-acquired pneumonia (HAP) is defined as an acute lower respiratory tract infection arising >48 h after hospital admission. It is a significant cause of morbidity and the most common healthcare-associated infection contributing to death.1 Predominant bacterial agents of HAP include Staphylococcus aureus, Streptococcus pneumoniae, Haemophilus influenzae, Enterobacteriaceae, Pseudomonas aeruginosa and Acinetobacter spp. Many of the Gram-negative agents show increasing multidrug resistance.2,3 It is generally agreed that empirical treatment for HAP should reflect: (i) this common aetiology; (ii) local resistance rates; and (iii) risk factors for a resistant pathogen in the individual patient, such as prior treatment with an antimicrobial agent and comorbidities.4 These general principles are formalized in national treatment guidelines produced in the UK, USA, Europe and Canada.2,5–7 None of these guidelines advocates doxycycline for the treatment of HAP. In a recent study to cross-relate susceptibility data from the BSAC Resistance Surveillance Programme (http://www.bsacsurv.org) with HAP guidelines produced by NHS Trusts in the UK, we therefore were surprised to find doxycycline widely advocated. Across 260 NHS Trusts, 46 guidelines for antibiotic treatment of HAP were accessible via online searches, using the criteria ‘NHS hospital antibiotic prescribing guidelines’, ‘NHS trust antibiotic prescribing guidelines’, ‘hospital prescribing guidelines for HAP’, ‘prescribing guidelines for hospital-acquired pneumonia UK’, ‘prescribing guidelines for hospital-onset pneumonia UK’, ‘NHS hospital antimicrobial prescribing guidelines’, ‘NHS antibiotic guidelines’, ‘NHS guidelines hospital-acquired pneumonia’ and ‘NHS guidelines hospital-onset pneumonia’. Trusts’ antimicrobial guidelines were summarized by their first- and second-line treatment recommendations for mild and severe HAP, along with any specific recommendations for patients with risk factors for MRSA infection (Table 1). Some guidelines were further subdivided into early-onset HAP, occurring <5 days after hospital admission, and late-onset HAP, occurring ≥5 days after hospital admission.8 Antimicrobial agents (single agents and combinations of agents) recommended for the treatment of mild or severe HAP in 46 NHS Trust hospital prescribing guidelines Where disease severity was not specified, recommendations were included in ‘mild’ HAP. Antimicrobial agents (single agents and combinations of agents) recommended for the treatment of mild or severe HAP in 46 NHS Trust hospital prescribing guidelines Where disease severity was not specified, recommendations were included in ‘mild’ HAP. Considerable heterogeneity was seen among the Trusts’ recommendations for mild and severe HAP (Table 1). Strikingly, however, doxycycline was the most commonly recommended first-line single agent for empirical treatment of mild HAP, indicated in 15/46 guidelines. It was also indicated in 7/46 guidelines as a second-line single agent treatment. Five further guidelines recommended doxycycline in combination with other antimicrobials, variously including amoxicillin, trimethoprim and rifampicin plus gentamicin. The online guidelines identified dated from 2009 to 2018, but all bar one of those recommending doxycycline dated from 2017 onwards. Doxycycline is a bacteriostatic antibiotic that is broadly active against the bacterial causes of community-acquired pneumonia (CAP), including S. pneumoniae, H. influenzae, Moraxella catarrhalis and ‘atypical’ pathogens such as Mycoplasma pneumoniae, Chlamydia pneumoniae and Legionella species.9 Among HAP pathogens, doxycycline is widely active against S. aureus but its activity is marginal at best against Enterobacteriaceae and Acinetobacter, with no published EUCAST breakpoints and frequent high-level resistance; P. aeruginosa is inherently resistant. Some literature suggests that early-onset HAP is often caused by community-type pathogens2 and clinical studies support the use of doxycycline in CAP patients requiring hospitalization.10 However, other studies suggest that early-onset HAP is increasingly caused by nosocomial pathogens, including Gram-negative bacilli,2 where doxycycline has no controlled clinical trial data demonstrating efficacy. Five NHS Trusts advocate doxycycline even for late-onset HAP. Such cases seem especially unlikely to be due to community-acquired pathogens. Six further doxycycline-recommending guidelines did not distinguish between treatment for early- and late-onset HAP. Fourteen of the 15 NHS guidelines recommended doxycycline specifically for ‘mild’ infection and, as an oral agent, it seems unlikely to be used in the severely ill; one guideline however advocated it as an oral stepdown after prior cefuroxime in severe HAP. Whilst some literature suggests that severe HAP is more likely to be caused by Gram-negative bacilli such as Pseudomonas,7 there is no evidence that mild HAP is predominantly caused by doxycycline-susceptible genera. We believe that doxycycline has entered hospitals’ guidelines based on a desire to avoid antimicrobial agents that select for Clostridium difficile and to reduce dependency on β-lactam/β-lactamase inhibitor combinations. However, there is no evidence-base for the use of doxycycline in HAP and no support in national or international guidelines. Its spectrum of activity is inappropriate for a substantially Gram-negative aetiology. Poor oral absorption may lead to variability in clinical response.11 In discussion, it has been suggested to us that HAP is over-diagnosed and that it is preferable that patients who do not have a significant infection received doxycycline rather than a broader-spectrum agent. However, this seems a weak argument given that some of these patients will genuinely have HAP and deserve to receive a drug with suitable coverage; those not needing antimicrobial treatment would better be distinguished using biomarker tests such as procalcitonin.12 We would urge Trusts advocating doxycycline in HAP to review their advice or to initiate formal efficacy trials. We acknowledge members of the BSAC Resistance Surveillance Standing Committee for constructive comments regarding this letter. D. M. L.: Advisory Boards or ad-hoc consultancy for Accelerate, Achaogen, Adenium, Allecra, AstraZeneca, Basilea, BioVersys, Centauri, Integra Holdings, Meiji, Melinta, Nordic, Pfizer, Roche, Shionogi, T.A.Z., Tetraphase, VenatoRx, Wockhardt, Zambon and Zealand, paid lectures for Astellas, AstraZeneca, bioMérieux, Beckmann Coulter, Cardiome, Cepheid, Merck, Pfizer and Nordic, research funding from Melinta, Merck, Paratek, Roche and VenatoRx, and relevant shareholdings in Dechra, GSK, Merck, PerkinElmer and Pfizer amounting to <10% of portfolio value. All other authors: none to declare. This study was supported by BSAC, which also supervised the internship by A. H. R. as part of a PhD with the Norwich Research Park Biosciences Doctoral Training Partnership funded by the Biotechnology and Biological Sciences Research Council (BBSRC).