Treatment of pneumonia: adherence to a hospital policy
- 1The Hospital Pharmacy, Vejle Hospital, Vejle, Denmark
- 2Department of Safe, Amgros I/S, Copenhagen Ø, Denmark
- Correspondence to Dr Lene Juel Kjeldsen, The Research Unit for Hospital Pharmacy, Amgros I/S, Dampfærgevej 22, Copenhagen Ø 2100, Denmark;
- Received 4 January 2012
- Revised 20 December 2012
- Accepted 3 January 2013
- Published Online First 29 January 2013
Objective To evaluate the adherence of an antibiotics algorithm for the treatment of pneumonia.
Methods A retrospective review of prescriptions was performed from January to April 2010 at three wards of a Danish hospital. Patients were included in the study if they were admitted with pneumonia or for observation for pneumonia. Intravenous benzylpenicillin was the first choice treatment for community acquired pneumonia, while the first choice for nosocomial pneumonia was intravenous ceforuxime. Switching from intravenous treatment to oral treatment was recommended. The algorithm also contained identification of first choice treatment for multiple infection foci.
Results Adherence to the algorithm was 91% regarding choice of antibiotics, 29% for switching from intravenous to oral treatment and 25% for additional infections other than pneumonia.
Conclusions Adherence to the algorithm was high, but there was still room for improvement with regard to switching from intravenous to oral treatment and treatment of infections with additional foci to pneumonia.
In Denmark it is well-known that suboptimal treatment with antibiotics may lead to the development of bacterial resistance, which is likely to compromise the future treatment of bacterial infections due to the potential lack of effect of existing antibiotics. This issue should be viewed in light of the fact that, according to the pharmaceutical industry, no new antibiotics are in the pipeline within the next 10 years, so the rational use of existing antibiotics is essential.1
In Denmark the use of antibiotics is monitored by, among others, DANMAP (established in 1995) while, for example, Statens Serum Institut and the Ministry of Interior and Health also follow the development of resistance. Maybe because of this national focus, the prevalence of antibiotic resistance in Denmark is relatively low.2
Treatment algorithms are tools to streamline as well as ensure the most rational treatment with antibiotics. Hence, Danish medical and surgical societies including the Danish Society of Respiratory Medicine (DSRM)3 and the Institute for Rational Pharmacotherapy (IRF)4 have developed algorithms for the treatment of, for example, pneumonia. However, these treatment algorithms cannot be used internationally since local resistance patterns must be included when developing treatment algorithms.2 According to Andersen et al,2 the first choice for the treatment of pneumonia in hospitals is penicillin G (benzylpenicillin), which correlates with the treatment algorithms developed by the DSRM and IRF.
Vejle and Give Hospital (VGH) has established a working group of specialist physicians and clinical pharmacists to develop policies for antibiotic treatment in order to encourage the rational prescription of antibiotics and, hopefully, to reduce the length of treatment.5 In 2007 the working group at VGH developed a new policy for antibiotic use in pneumonia. Based on this policy, a treatment algorithm was produced as a tool when prescribing antibiotics which correlates with the treatment guidelines made by the DSRM and IRF.6 To support the implementation of the new antibiotics policy, an educational visit was made to the hospital wards by a senior microbiologist and a clinical pharmacist.
The aim of the current study was to evaluate adherence to the new antibiotics policy for the treatment of pneumonia on three medical wards at VGH.
The study was conducted from January to April 2010 and included patients diagnosed with pneumonia or under observation for pneumonia. Patients suspected of having more than one infection focus were not included until all foci were identified, due to the construction of the antibiotics policy. A retrospective review of electronic charts identified patients for inclusion. Patients who had commenced antibiotic treatment before admission were included if the treatment was adjusted to VGH's antibiotics policy on admission. Patients receiving antibiotic treatment not in accordance with the hospital's antibiotics policy were accepted if the treatment was almost complete and effective. Patients who were receiving chemotherapy and/or had neutropenia were excluded as these patients have a suboptimal immune system and require more intense antibiotic treatment.
The policy comprised two steps. In step 1, the algorithm for the treatment of pneumonia was divided into two sections: (1) pneumonia contracted during the hospital admission; and (2) pneumonia contracted outside the hospital. The first choice of antibiotic for the treatment of pneumonia contracted outside the hospital was intravenous benzylpenicillin, while the first choice of antibiotic for the treatment of nosocomial pneumonia was intravenous ceforuxime. In case of penicillin allergy, the use of intravenous cefuroxime was recommended. Medication substitution was recommended when there was a lack of response to the treatment of first choice. The policy was combined with other guidelines when multiple infection foci were identified. For example, if concomitant pneumonia and a urinary tract infection (UTI) were present, a combination of benzylpenicillin and pivampicillin was recommended.
In step 2, switching from intravenous to oral antibiotics was recommended as intravenous formulations are significantly more expensive than oral formulations, and oral treatment may lead to an earlier discharge since the patient may finish the treatment at home.
Adherence to the antibiotics policy was assessed.
Forty-four patients were included in the study and nine patients were excluded, five due to suspicion of more than one infection focus, one who received concomitant chemotherapy, one with neutropenia and two patients continued treatment initiated by their general practitioner/another hospital (figure 1). The average age of the included patients was 72.6 years (range 31–91).
Adherence to the antibiotics policy was found for 40 patients (91%) while three (7%) of the treatments deviated from the policy due to identification of bacteria and advice from the microbiologist. For one patient (2%) the treatment did not follow the antibiotics policy.
Four patients (9%) died as a result of age and general infirmity; the deaths were not associated with the antibiotic treatment. Thus, 40 patients were alive at the end of the survey, 35 of whom (88%) had their antibiotic treatment changed from intravenous to oral treatment. Of these 35 patients, the treatment was adhered to in 10 (29%) and, in the remaining 25 (71%), substitution of intravenous antibiotics with oral antibiotics resulted in treatment with antibiotics with a broader spectrum than that recommended by the policy.
The medication reviews showed that adherence to the antibiotics policy decreased when more than one infection focus was evident. This was observed for patients who were admitted with pneumonia as well as those with UTI. For this infection combination, treatment with benzylpenicillin and pivampicillin was recommended. Only two (25%) of the patients received the recommended treatment, while the remaining six (75%) were treated with cefuroxime.
The study showed that the three medical wards at VGH had a high adherence rate to the antibiotics policy for the treatment of pneumonia with regard to choice of antibiotics (91%, n=44). However, room for improvement still existed with regard to switching from intravenous to oral antibiotic treatment (29%, n=35) and for the treatment of concomitant infections with pneumonia (25%, n=8).
Treatment of community acquired pneumonia has been studied worldwide, and areas such as initial empirical therapy, switching from intravenous to oral antibiotics and prevention of community acquired pneumonia have been particularly identified as having room for improvement.7 The current study showed a high level of adherence to treatment guidelines regarding the choice of antibiotics, but a switch from intravenous to oral treatment was insufficiently performed.
A Swiss study at a hospital showed a fair but lower adherence rate to treatment with antibiotics for pneumonia (86%).8 The high rate of adherence to the antibiotics policy in the current study may be a result of the comprehensive educational visits to the wards. This is supported by Deuster et al who showed that multidisciplinary teaching may increase adherence when implementing antibiotics policies.9 However, no baseline data were collected to evaluate this. The high adherence rate may therefore be due to the focus on prescribing of antibiotics over several years in Denmark and not solely the result of the comprehensive educational visit.
However, the adherence rate was low for switching from intravenous to oral treatment and for the treatment of infections in addition to pneumonia. Instead of adhering to the policy, broad-spectrum antibiotics were frequently prescribed when switching from intravenous to oral treatment, even though the change in administration form was solely due to an increased length of the initiated intravenous treatment and a change in choice of antibiotics was therefore not warranted. The aforementioned Swiss study reported a delay in the switch from intravenous to oral antibiotics for the treatment of pneumonia in 38% of patients resulting in increased costs, but the rationality of the switch was not discussed.8 An American study found that only 65% of eligible patients had their pneumonia treatment switched from intravenous to oral antibiotics.10 Hence, these studies (including the current study) all concur with the international statement regarding room for improvement in switching from intravenous to oral antibiotic treatment.7 The reason for not switching from intravenous to tablet formulation according to the policy may be uncertainty about the efficacy of the oral formulation compared with the intravenous formulation when the intravenous formulation was substituted with a broader spectrum oral formulation. With regard to a delay in switching, this may be explained by uncertainty of the effect or perhaps lack of attention to the benefits of switching.
Infections localised to two foci were primarily treated with one broad-spectrum antibiotic instead of two narrow-spectrum antibiotics, as recommended by the policy. It is possible that the prescribing physician chose the broad-spectrum antibiotic to make sure that even the most resistant bacteria were eliminated,11 but these treatment choices may lead to unnecessary use of broad-spectrum antibiotics. There is evidence that the majority of physicians view the suboptimal use of antibiotics as a cause of resistance development, but they still prescribe broad-spectrum antibiotics to be sure of eliminating resistant bacteria.11
Another possible reason for using one broad-spectrum antibiotic rather than two narrow-spectrum antibiotics is non-compliance with multiple medications. However, the risk of developing resistance by using one broad-spectrum antibiotic instead of two narrow-spectrum antibiotics exceeds the risk of developing resistance to narrow-spectrum antibiotics due to non-compliance.
In addition, narrow-spectrum antibiotics are often cheaper than broad-spectrum antibiotics, which is another argument in favour of using narrow-spectrum antibiotics when recommended by the antibiotics policy.
Patients who did not have their antibiotic treatment adjusted to VGH's antibiotics policy on admission were excluded. It is likely that the adherence rate would have been lower if these patients had been included in the study. However, treatment according to the policy can only be commenced when the foci have been identified.
The strengths of the study include the high quality of the electronic data used to assess the adherence to the policy. However, the retrospective design of the study may be considered a weakness since the evaluation was purely descriptive and did not include an intervention when suboptimal adherence to the policy was detected.
In conclusion, the study showed a high adherence rate to the antibiotics policy for the treatment of pneumonia with regard to choice of antibiotics but there is still room for improvement with regard to switching from intravenous to oral antibiotic treatment and for the treatment of concomitant infections to pneumonia. To improve the adherence rate of these aspects of treatment, a focused intervention could be implemented—for example, by systematic reviews of antibiotic prescriptions by clinical pharmacists. Audit and feedback of antibiotic prescriptions may also be an effective intervention to improve adherence to the policy.
High adherence to an antibiotics policy for the treatment of pneumonia in a hospital setting was found in this study.
This suggests that, when implemented properly, antibiotics policies may be excellent tools to prevent development of bacterial resistance.
Contributors HKB performed the data collection while supervised by LJK. HKB provided the first draft of the manuscript. Both authors approved the final manuscript.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.