Implementation and evaluation of tailored intervention strategies to influence antibiotic prescribing for community-acquired pneumonia

Adherence to guidelines for the management of community-acquired pneumonia (CAP) has been shown to improve patients' clinical outcomes. However, several studies have indicated that the chosen antibiotic regimen is frequently not consistent with guideline recommendations. This might lead to suboptimal treatment, either by exposing patients to a greater risk of treatment failure or by unnecessary use of broad-spectrum antibiotics, which contributes to the emergence of antibiotic-resistant pathogens or consequent development of Clostridium difficile-associated diarrhoea. It has been demonstrated that active implementation of CAP guidelines can significantly improve adherence to recommendations, which consequently, might improve patients' clinical outcomes. The present research developed, implemented and evaluated tailored intervention strategies to improve physicians' concordance with CAP guidelines. A number of inter-related studies were conducted as a part of this research project. Firstly, baseline data was collected to measure the level of physicians' adherence to national CAP guidelines in two Tasmanian hospitals, the Royal Hobart Hospital (RHH) and North Western Regional Hospital (NWRH). It was evident in that study that adherence to CAP management guidelines was poor at both study sites (16.1% and 7.5% for RHH and NWRH, respectively). This was followed by a study to identify and quantify potential barriers to the adherence to CAP guideline recommendations. A questionnaire was distributed to RHH doctors in non-surgical areas of practice. Of the study population, 43.1% doctors responded to the survey; of those who responded, 46.4% thought the influence of senior doctors on their juniors could be a factor affecting adherence to the guidelines. Other barriers noted were a lack of guideline awareness (39.3%), the requirement to calculate the severity of CAP (35.7%), and the existence of other guidelines that conflict with Therapeutic Guidelines: antibiotic, version 14 (TG14; 28.6%). A qualitative study was then designed to determine factors that influence doctors working within the emergency department (ED) to prescribe ceftriaxone outside the TG14 recommendations. Eight face-to-face interviews were performed with ED doctors. Five main themes emerged as influencing decisions regarding the selection of ceftriaxone for patients with CAP: (i) clinical intuition compared to a structured evaluation of severity, (ii) clinical uncertainty, (iii) prior clinical experience, (iv) source of guidance and (v) prescribing etiquette. A questionnaire survey was then sent to infectious disease pharmacists nationally in order to identify the strategies that have been used and perceived as successful for the management of CAP in their institutions. Of the study population, 41 pharmacists (27.3%) responded to the questionnaire. Of these, 90.2% pharmacists reported their hospitals having an antimicrobial stewardship (AMS) program. Multifaceted strategies to enhance antibiotic prescribing in ED for CAP, were mentioned as being in place in all responses. However, the largest number of the respondents (34.1%) considered use of CAP clinical pathways to be the most effective strategy. Intervention strategies were subsequently developed and implemented based on the findings from the above studies. Two interventions were implemented over two time periods: one with general strategies across medical units and a second focused on the ED. During the general intervention period, local CAP guidelines (based on TG14) were released. The guidelines were developed and approved by the hospital's medical and emergency departments. The release of the CAP guidelines was accompanied by a multifaceted educational package to increase awareness of the guidelines. Medical and ED teams were targeted in the educational package, which included group sessions, wall posters and laminated lanyard cards summarising the local guidelines. During the second time period, two further strategies were introduced (a CAP clinical management pathway and monthly auditing with feedback) and targeted specifically at ED staff. We evaluated the impact of the interventions on guideline adherence rates and clinical outcomes (mortality rates and hospital length of stay, LOS). To evaluate the impact of the intervention, two hospital sites were selected, one (RHH) acted as an intervention site and the other (NWRH) as a control site where no intervention was made. The study found the intervention had an overall impact on guideline adherence rates at the intervention site, and it reduced overall mortality rates and LOS for patients with non-severe CAP. Compared to the baseline data, the adherence rate increased significantly at the RHH during the intervention period (16.1% vs 50%; p < 0.05). However, no significant improvement was indicated in the control site (7.5% vs 19.1%; p > 0.05). The in-patient mortality was significantly lower in the intervention group when compared to the non-intervention groups (all baseline data plus the data from the NWRH during the intervention period) (3.4% vs 7.3%; p < 0.05). Sub-group analysis revealed patients with non-severe CAP in the intervention group had an average LOS 0.8 days shorter than the non-intervention groups (p < 0.05). Results from the previous study indicated a positive impact of the intervention in the overall adherence to CAP recommendations. However, two main strategies were conducted in two consecutive times during the intervention periods, a general intervention and an ED-focused intervention. Therefore, a time-series analysis was conducted to determine the impact of strategies over time at the intervention site. The rates of adherence to the CAP guidelines during the pre-intervention (5 months) and general intervention periods (5 months) were 28.1% and 31.2%, respectively. The difference was not statistically significant. During the ED-focused intervention period (7 months), the level of adherence with guidelines was significantly higher at 61.5% (p < 0.05). Finally, we evaluated the use of ceftriaxone in all indications in two time periods, before and after the initiation of the intervention. The aim of this study was to determine if our intervention in CAP management could affect the use of ceftriaxone in other indications. Concordance to the TG14 for all indications, with the exception of respiratory tract infection (RTI), was similar between the two study periods. For the RTI, concordant use of ceftriaxone significantly increased from 50% during the first period to 64.5% during the second study period (p < 0.05). Among community-acquired lower respiratory tract infections, our findings indicated a significant decrease in the unnecessary use of ceftriaxone for patients with mild CAP and acute exacerbation of chronic obstructive pulmonary disease in the intervention group (both 19% vs 3.2%; p < 0.05). However, there were no significant changes in the appropriate prescribing of ceftriaxone for other indications. In conclusion, this research project identified, with in-depth analysis, potential factors that lead to the prescription of discordant antibiotic regimens for empirical management of CAP. It was subsequently demonstrated that a tailored multifaceted intervention significantly improved adherence to CAP guidelines, which consequently reduced the inappropriate prescribing of ceftriaxone for this indication. This was associated with a decrease in mortality rate and length of hospital stay among patients in the intervention group.