Editor’s Note: The author wishes to acknowledge the contribution of Michael Soriano to this article.

As a result of the revision of Therapeutic Guidelines: Antibiotic (Antibiotic Guidelines) Version 14 and the national move towards implementing antibiotic stewardship programs in hospitals, HPS Pharmacies undertook a project to improve the implementation of the recommendations of the Antibiotic Guidelines for the treatment of community-acquired pneumonia (CAP) in three client hospitals.

The potential benefits of improving concordance with the Antibiotic Guidelines include:

  • Increasing the use of narrow spectrum antibiotics, particularly in the initial empiric treatment of CAP
  • Reserving broad-spectrum antibiotics for more acutely ill patients
  • More appropriate drug choices in the management of CAP and bronchitis
  • Reducing the development of resistant organisms in the hospital setting, and
  • Offering patients treatments which represent the most up-to-date evidence-based therapy, ultimately resulting in the highest standard of medical care and service.

A retrospective audit of 52 patients admitted to three hospitals with a presumptive diagnosis of CAP was performed to determine each hospital’s concordance with the current Antibiotic Guidelines (refer to Table 1).

Table 1. Summary of CAP Audit Results.

Audit Criteria Hospital A Hospital B Hospital C
Concordance to Antibiotic Guidelines 23.5% 31.6% 14.3%
Average length of inpatient treatment 5 days 5.6 days 8 days
Average length of outpatient treatment 5 days 7 days 4.6 days
Average length of stay 8.3 days 11.11 days 10 days


The audit involved inspecting patient case notes for evidence of:

  • Antibiotic choice (guidelines recommend intravenous [IV] benzylpenicillin rather than ceftriaxone)
  • Duration of treatment (maximum five days of IV therapy, or until patient shows signs of improvement), and
  • Length of hospital stay.


Of the patients reviewed, pneumonia severity scoring was not documented in any of the medical notes, and in fact, not all measurements were available to make scoring possible
(especially blood pH and blood gases). This highlights the controversies surrounding the use of pneumonia severity indices.

Hospital A:

Of the 21 patients included in the audit, five (23.5%) had a documented penicillin allergy and were appropriately prescribed ceftriaxone, as advised by the Antibiotic Guidelines. The remaining 16 were not compliant, as ceftriaxone and azithromycin were prescribed in all cases, regardless of the numerous variables such as patient age, comorbidities, prescriber, prior antibiotic use, and pneumonia severity. This result showed an overall concordance with the Antibiotic Guidelines of 23.5% within this hospital.

Initial treatment was extended beyond the recommended seven days in nine cases, with three patients being further discordant by not progressing to oral therapy. Only five patients had successful sputum cultures, and were changed to targeted therapy, before transitioning to oral antibiotics as their condition improved, as recommended by the Antibiotic Guidelines.

The average duration of inpatient antibiotic treatment was five days (2-10 days) and outpatient antibiotic treatment was five days (0-14 days). The total average length of treatment was ten days, and length of stay was 8.3 days.

Hospital B:

Of the 19 patients, only one had a documented pneumonia severity score, which was performed within the emergency department of a public hospital.

Six out of ten patients transferred from public hospitals were empirically prescribed antibiotics in concordance with the Antibiotic Guidelines. Of these six patients; four were maintained on compliant regimes, one was swapped to ceftriaxone and azithromycin, and one patient received targeted therapy following blood cultures. The remaining four patients transferred from public hospitals had initiated on ceftriaxone and azithromycin. One patient was maintained on these drugs for four days and prescribed doxycycline thereafter; one patient was changed to targeted therapy in response to sputum cultures; and two were treated as if they had penicillin hypersensitivity without any accompanying documentation (one patient received moxifloxacin for eight days with no documentation of hypersensitivity to penicillin).

Of those nine patients admitted, two were admitted directly into the intensive care unit and received ciprofloxacin IV, whilst four were prescribed ceftriaxone and azithromycin in the wards. A further three patients were appropriately treated with targeted antibiotic regimes following sputum and blood analysis.

The results for these patients showed a 31.6% concordance with the Antibiotic Guidelines, and an average length of stay of 11.11 days. Inpatient and outpatient antibiotic treatments were on average 5.6 days (2-29 days) and seven days (5-42 days) respectively. The total average length of treatment was 12 days, however this was skewed by three patients who had complications. Without these patients, the concordance to the Antibiotic Guidelines was 33.3%, and the total lengths of treatment and stay would be marginally reduced.

Hospital C:

Audit criteria was met by seven cases over a six month period. None of the patients in these cases had recorded allergies to penicillin, hence did not justify the use of ceftriaxone instead of penicillin. Three patients had sputum samples tested during their admission.

Three of the seven patients were admitted from public hospitals, two of these had commenced antibiotic regimes including benzylpenicillin IV and azithromycin, which correlated with the Antibiotic Guidelines. One of these two patients was then changed to ceftriaxone and azithromycin. The other patient maintained the original regime for two days, before being swapped to amoxycillin and clavulanic acid. This patient developed a rash after a further two days, and then ceased all antibiotic treatment. The third patient admitted from a public hospital had received ceftriaxone, azithromycin and gentamicin, and continued on ceftriaxone and azithromycin after transfer to Hospital C.

The trend of using ceftriaxone and a macrolide antibiotic for the empirical treatment of presumed CAP was mirrored in all patients directly admitted to Hospital C, along with one patient admitted from a private hospital emergency department.

All three patients who had sputum tests grew mycoplasma, justifying the use of a macrolide in their antibiotic regime, and the only patient to require a quinolone antibiotic did so because of their inability to tolerate a macrolide beyond seven days of therapy.

This hospital showed a 14.3% concordance with the Antibiotic Guidelines, and an average length of stay of ten days. The average length of inpatient and outpatient antibiotic treatments were eight days (1-16 days) and 4.6 days respectively. The total average time of treatment was 12 days.


The inappropriate use of third generation cephalosporins, such as ceftriaxone, has been linked to the development of antibiotic induced diarrhoea and antimicrobial resistant infections, including Methicillin-Resistant Staphylococcus Aureus (MRSA). These complications can extend a patient’s hospital stay, as well as increasing the complexity of treating bacterial infections in the wider population. Local and international opinion leaders, including The South Australian Expert Advisory Group on Antimicrobial Resistance (SAAGAR) have expressed concern over the frequent use of third generation cephalosporins as empiric treatment for CAP and the potential for “collateral damage” with their use.

Hospitals are therefore recommended to develop a guideline for the treatment of CAP which includes a pneumonia severity score and an initial antibiotic management plan which aligns with the recommendations of the Antibiotic Guidelines.

Intravenous amoxycillin or benzylpenicillin must be used for the empiric treatment of moderate to severe CAP, unless the patient has a documented non-anaphylactic allergy to penicillin or has a creatinine clearance of less than 30mL/min. Quinolone antibiotics, such as moxifloxacin, should be reserved for the empiric treatment of CAP in those patients with a documented anaphylactic allergy to penicillin.

The duration of intravenous therapy should be reviewed at every change of the intravenous catheter, and clinical improvement should guide the change to oral therapy as soon as possible.


  1. TQEH and RAH antimicrobial stewardship committee and drug committees. Central Adelaide Local Health Network guideline for the treatment of community acquired pneumonia in immunocompetent patients. Version 7.1. Adelaide: Central Adelaide Local Health Network; 2012.
  2. Community-acquired pneumonia in adults: empirical treatment [Revised 2010 Jun]. In: Antibiotic Expert Group. Therapeutic Guidelines: Antibiotic. Version 14. Melbourne: Therapeutic Guidelines Limited; 2010.

Subscribe Knowledge Centre Updates

Enter your details to receive Knowledge Centre updates