Introduction

Healthcare associated infections are a significant clinical problem and their prevention is promoted by action plans initiated by the Scottish Government. Antimicrobial prophylaxis can prevent one major group of healthcare associated infections, surgical site infections (SSI). At the same time antimicrobial prophylaxis holds the risk of antibiotic associated diarrhoea caused by Clostridium difficile infections and increases the risk of resistant strain growth such as Methicillin Resistant Staphylococcus aureus (MRSA). Therefore, all surgical procedures that require antimicrobial prophylaxis should be covered with antimicrobial regimens with sufficient antibiotic spectrum but avoiding broad spectrum antibiotics. Recent guidelines promote the prudent use of antimicrobials for surgical prophylaxis. The present research project set up an investigation to compare current prescribing practice with these surgical antimicrobial prophylaxis guidelines.

Methods

Data on surgical antimicrobial prescribing practice was collected at six sites across NHS Greater Glasgow and Clyde. Data was entered into a preformatted SPSS™ file, double-checked, aggregated and analysed. The adherence to 30 quality criteria was calculated. These criteria were developed and validated in previous MSc projects and were based on national and local surgical prophylaxis guidelines. They comprised items to evaluate documentation practice and patient care. Furthermore, potential differences in the adherence to local guidelines during a six months time period, across the six sites and across the four types of surgery, namely general surgery, orthopaedics, urology and ear, nose, throat surgery, were assessed. Fisher’s exact test was used to determine statistical signifinance of findings. It was analysed which antimicrobial regimens were most frequently used in specific surgical procedures.

Results

Data of 820 patients were analysed in the present project. The adherence to the 30 quality criteria varied by criterion and ranged from 24.9% to 99.8% in criteria applicable to more than 50 patients.
The adherence to quality criteria evaluating the documentation of patient’s allergy status and details on the antimicrobial agents used for prophylaxis was high. The adherence was intermediate and low in terms of transcribing prescribed antimicrobials to the ‘once only’ section of the kardex (30%) and recording of the time of incision (36%), the rationale for extra doses (49%) and the nature of allergic reaction (60%).
In order to evaluate patient care it was focussed on criteria measuring the adherence to the duration, the indication of prophylaxis and the choice of antimicrobial agents. A high proportion of patients (92%) were not prescribed any prophylactic doses post-operatively and the majority of patients received antimicrobials for less than 24 hours.
In only 89% of patients with an indication for antimicrobial prophylaxis, antimicrobial agents were prescribed. Consequently, 11% of patients were under increased risk of suffering from avoidable SSI. There were 78 patients without indication for antimicrobial prophylaxis. However, 27% of these patients received antimicrobial prophylaxis even though it was not recommended.
The choice of agents for prophylaxis was adherent in 52% of the patients. Cephalosporins were found to be still popular antibiotics in surgical prophylaxis despite the risks associated with using broad spectrum antibiotics. In gastrointestinal surgery, the combination of metronidazole and cefuroxime and the three agents amoxicillin, gentamicin and metronidazole were the non-adherent regimens most frequently used. The latter is not recommended for surgical prophylaxis but for the treatment of intra-abdominal sepsis.
By comparing the four types of surgery it emerged that the adherence to local guidelines in general surgery appeared to be more challenging than in orthopaedic or urologic surgery. The performance in antimicrobial prophylaxis practice varied among the six settings. In hospital H4 procedures that required prophylaxis were covered with prophylactic antimicrobials less frequently than in the other five hospitals (67% vs. 91%, p<0.001). The adherence to the choice of agents was low in hospitals H2 and H6. 
There was no significant change found in the adherence to most of the criteria when comparing the present findings with results of the baseline study which was enrolled six months before.

Conclusion

With the help of 30 explicit criteria this audit detected areas of low and intermediate adherence to antimicrobial prophylaxis guidelines that can be targets of future intervention to implement the guidelines. Regarding the documentation, interventions should focus particularly on the time of incision, the nature of allergic reaction and the rationales for guideline deviation. The documentation of the incision time will be essential in future audits if the timing of prophylaxis is evaluated. The nature of allergic reaction is important for the prescriber to evaluate the risks of his or her prescribing decision.
In order to improve the quality of care and to prevent avoidable SSI as well as risks of antimicrobial over-treatment, interventions should focus on the indication of prophylaxis and the choice of agents rather than the duration of prophylaxis. Interventions could start in areas where adherence was found to be particularly low, namely in general surgery and hospitals H4 (indication) and H2 and H6 (choice of agents).
Future research could analyse rationales for deviation from best practice and reasons for varying implementation of the guidelines across different settings.
In conclusion, interventions are needed to raise the awareness of local antimicrobial prescribing guidelines and encourage the adherence in order to improve surgical prophylaxis practice. That might result in a decreased number of SSI in Scottish hospitals from which both the patients and the healthcare system would benefit. Preferably, antimicrobials pharmacist in co-operation with specialist surgeons should conduct these interventions.