Problem

Prior to the start of this study, the uptake of 48-hour antimicrobial reviews on the surgical wards at Darent Valley Hospital (DVH) was considered to be at an unacceptably low level, thus demonstrating a lack of antimicrobial stewardship. The aim of this quality improvement project (QIP) was to improve the proportion of 48-hour antimicrobial reviews completed and documented on two surgical wards at DVH with a goal of 100% compliance.

DVH is located in Dartford, a commuter town for London. The majority of patients staying on the two surgical wards included in this study were prescribed at least one antimicrobial and had a short length of stay, an average of 5.1 days, meaning that there was a high turnover of patients throughout the duration of our study.1 The two surgical wards were classed as general surgery wards, although the majority of patients were lower gastrointestinal cases. The therapeutic guidelines used on the two surgical wards for antimicrobial prescriptions were in line with hospital and trust-wide policy. DVH employs a paper-based system for recording patient notes and patient drug charts with no electronic system available at present.

The issue of poor compliance with 48-hour antimicrobial review completion and documentation was raised by the hospital’s pharmacy department who identified this issue in their bimonthly audit of antimicrobial stewardship (Hospital Antimicrobial Prudent Prescribing Indicator) which identifies departments and standards requiring improvement. The pharmacy department viewed the current level of 48-hour antimicrobial review uptake as detrimental to patient care with the surgical wards consistently the lowest performing, so we attempted to tackle this problem by using the interventions described in this study.

Background

The misuse of antibiotics leads to the emergence of antimicrobial resistance, an important public health and patient safety issue, while concurrently reducing the efficacy of antimicrobial treatments against infection.2 Infections caused by resistant organisms are associated with poorer clinical outcome and a prolonged hospital stay when compared with infections caused by susceptible organisms.3 4

Antimicrobial stewardship enforces measures to reduce the adverse effects of antimicrobial use by reducing inappropriate antibiotic prescriptions.3 This encompasses a reduction in unnecessary antibiotic use, unsuitable use of broad-spectrum antibiotics and inappropriate dosage or duration of antibiotic use.5

From November 2011, the ‘Smart Start-Then Focus’ antimicrobial stewardship package encouraged documentation on the drug chart and in medical notes of the route, indication, does and duration of any antibiotic treatment. The rationale behind this is that antibiotics are often continued unnecessarily because clinicians caring for a patient do not have information specifying why the antibiotic was initially commenced, and the intended duration of antibiotic course.3

The ‘Smart Start-Then Focus’ initiative also suggested reviewing the clinical diagnosis and the need for continuing antibiotics by 48 hours. At this stage, a clear, active, antimicrobial stewardship decision should be made and documented (stop the antibiotic, switch from intravenous to oral, change antibiotic, continue and review again at 72 hours, change to outpatient parenteral antibiotic therapy).3 The fulfilment of this particular task is what we are trying to improve in our QIP.

A recent QIP aimed to improve the completion of 48-hour antimicrobial reviews in drug charts on a respiratory ward.6 This project was deemed to be successful, as demonstrated by an increase in completion of 48-hour antimicrobial reviews in drug charts, and a reduction in the number of days that patients were prescribed intravenous antibiotics for. Of the seven PDSA (plan, do, study, act) cycles carried out, the most efficacious were: educational talk by consultant, inclusion of 48-hour review status on daily handover and active ward pharmacy involvement. Interventions with less impact were visual prompts and magnets on the patient whiteboard.6 7

A recent audit demonstrated that consistent review of antibiotic prescriptions at 48 hours would improve adherence with the Department of Health’s ‘Smart Start-Then Focus’ guidance, and would improve patient care by allowing for a switch to oral therapy or completion of antibiotic course to take place earlier.8 This audit also estimated that with appropriate observance of 48-hour antimicrobial reviews, 1800 hospital bed days could be saved locally.8 In the context of the National Health Service (NHS) as a whole, the number of hospital bed days saved would be significantly greater, with an associated increase in patient safety and reduction in costs.

Measurement

Initial data collection formed the baseline measurement. This involved collecting data from two surgical wards at two points in time on consecutive weeks. Data were collected on the antimicrobials prescribed and whether they had been reviewed at 48 hours in the drug chart, whether the completion of a review had been documented in the patient notes and whether any changes had been made to the antimicrobial prescription as a result of the 48-hour review. In order to assess the success of our PDSA cycles, we would continue to collect data in these three domains 1 week after the implementation of an intervention.

The proportion of 48-hour antimicrobial reviews documented in patient notes was our primary process measure. The patient notes are where all information regarding a patient for the duration of their hospital stay accumulates, and is used as a reference point by all branches of the multidisciplinary team for any queries regarding a patient. It was deemed to be the most important place to document a 48-hour antimicrobial review as it would be easily accessible to all staff involved in the patient’s care, and in contrast to the patient’s drug chart, would be less likely to be changed during the patient’s stay in hospital, and provides more space for the prescriber to provide information on the indication and length of antimicrobial course which, following guidance from the ‘Smart Start-Then Focus’, would be useful for other clinicians in future management decisions.

The proportion of antimicrobial prescriptions changed at 48 hours was our secondary process measure as it would indicate whether changes in the primary process measure were leading to a positive change in practice, for example, switching from intravenous to oral antibiotic. As was highlighted by the ‘Start Smart-Then Focus’ initiative, an antimicrobial stewardship decision at 48 hours is vital to improving standards of patient care.6

Data were collected from all patients who were staying on the ward on the day on which data collection was taking place. Baseline data were collected on two dates in order to ensure an adequate sample size from which the baseline measurement could be made. Thereafter, data were collected 1 week after the implementation of each respective PDSA cycle. The four dates for data collection after each PDSA cycle are as follows: 26 November 2018, 5 December 2018, 13 February 2019 and 20 February 2019.

For the duration of the QIP, data were collected by a team of five medical students, with four medical students being involved with each round of data collection. The students were divided into two pairs, with each pair collecting the data for a ward; this took approximately 90 min. Prior to the start of the project, the medical students attended lectures at their university on how to successfully carry out a QIP in a clinical environment. The medical students then received further training and support from the QIP lead at DVH, as well as from the surgical consultant and pharmacists with whom they were working with closely for the duration of the project.

Design

This QIP implemented four PDSA cycles over a period of 19 weeks. The PDSA cycles were carried out on two surgical wards, led by medical students in conjunction with antimicrobial pharmacists and the microbiology team.

A process mapping exercise identified key areas amenable to demonstrate improvement. Additionally, the team shadowed ward rounds to establish how 48-hour antimicrobial reviews were completed in practice and to provide an opportunity to discuss the topic with clinical staff. The primary process measure, the proportion of antimicrobial reviews completed at 48 hours in patients’ clinical notes, was developed to quantify any improvement. Other measures were recorded to review improvements in adherence to national antimicrobial stewardship guidelines and as a surrogate measure of improvement in patient outcome.9

The PDSA cycles focused on raising awareness of antimicrobial resistance and stewardship, educating both junior and senior healthcare staff, and designing stickers and posters to serve as sustainable forms of reminders on the wards.

Strategy

Lessons and limitations

The aim of this QIP was to improve the proportion of 48-hour antimicrobial reviews completed and documented on two surgical wards with a goal of 100% compliance. As our data were collected over a 19-week period, it was essential for our interventions to be sustainable over this period to see positive results at the end of our QIP, and beyond the completion of the project to ensure that any changes made could be maintained in the long term to improve the quality of patient care in the future. We found issues with the sustainability of our initial PDSA cycles and made attempts to correct this in PDSA cycle 3 and PDSA cycle 4 with interventions that did not require constant input. This was particularly notable as half of our interventions took place after the Christmas break, with the new year introducing a different cohort of staff as both nursing and medical staff joined the surgical wards as part of a new training rotation. However, new medical staff were exposed to PDSA cycles 1 and 2 (trust-wide email and education sessions with junior doctors). Furthermore, senior nursing staff and senior medical staff were retained which suggests their important role in maintaining antimicrobial review documentation, and provides an explanation as to why there was still an improvement in antimicrobial review documentation. The poor results experienced in PDSA cycle 3 highlighted the importance of implementing more sustainable interventions that are not affected by the turnaround of staff, or targeting interventions on staff who would remain on the surgical ward for a prolonged period of time such as the ward sister or surgical consultants.

Difficulty in engaging the ward staff was the most significant barrier to the success of the project. In our discussions with the ward staff, we learnt that our project was often considered surplus to the work carried out by an already overburdened workforce. This was problematic as staff collaboration was essential for the successful implementation of some interventions, for example, PDSA cycle 3 (stickers). This could have been overcome by involving surgical consultants at an earlier stage; with the support of senior staff, it was possible to bridge the gap with ward staff and implement more successful interventions. It is likely we were undermined due to our position as medical students, and if we held a more senior role within the trust (eg, junior doctor), or were seen to be working closely with senior hospital staff, there would have been a greater degree of engagement in our project.

The secondary process measure, the proportion of changes made to antimicrobial prescriptions, changed a little during our project and was below the baseline measurement by the end of the project. This is disappointing as it was hypothesised that an increase in 48-hour antimicrobial reviews being completed would result in more antimicrobial prescriptions being changed at 48 hours. On reflection, this may be explained by placing too much of an emphasis during PDSA cycles on improving compliance with 48-hour antimicrobial reviews without providing a thorough explanation as to why it was important and what changes we hoped to bring about, thus making the 48-hour antimicrobial review appear as a tick box exercise. There were efforts made to address this, such as in PDSA cycle 2 where there was discussion with the FY1 and FY2 doctors on the benefits of completing the 48-hour antimicrobial review, but again, its effect was limited. Another reason may be that before this QIP, antimicrobials were already being changed appropriately at 48 hours but not documented. A future project could try to establish how many antimicrobials could change to a more appropriate one and if this can be improved. This should take into account both guidelines and clinical indication, which may be hard to measure retrospectively. Our project aimed to measure guideline adherence with our secondary process measure; however, this was difficult to assess as we only collected data on one aspect of the guidelines. A potential solution to antimicrobials being changed without a documented review, which was beyond the scope of this project, is to implement a hard time-limiting prescription of antimicrobial, requiring a review at 48 hours, and if this is not completed, the prescription will be stopped.10

No significant issues were encountered with data collection. Before data collection began, the nursing team on the wards were made aware of our presence and we were never limited in our access to the ward or any documentation. In hospitals with electronic patient records and prescribing systems, data collection could be significantly streamlined and lead to an increase in the number of patients and wards reviewed. Furthermore, employing an e-prescribing system would aid compliance with conducting 48-hour antimicrobial reviews as this can incorporate features to remind clinical staff that a review requires completion.