Selection of opportunities for improvement:
Carbapenem –resistance enterobacteriacease (CRE) incidence was persistently high in Hospital Serdang in the past two years (21 cases). Studies showed inappropriate use of carbapenem leads to emergence of CRE. An audit done in year 2017 showed 84.6% carbapenem prescribed in Hospital Serdang was inappropriate. Thus, this study aim to improve the appropriateness of carbapenem prescribed in Hospital Serdang.

Key measures for improvement:
Percentage of appropriate carbapenem empirically prescribed was used as indicator with standard of ≥80%, which is based on a meta-analysis which reported percentage of appropriate antibiotic prescribed range from 36% to 91%.

Process of gathering information:
A cross-sectional study using universal sampling with calculated sample size of 37 subjects was carried out in 3 high carbapenem usage wards involving two cycles. First cycle was done in January 2018 that all empirical carbapenem prescriptions ordered in period of November 2017 until January 2018 was audited for appropriateness. Data was retrieved from prescriptions, carbapenem order form database and medical records review. Appropriateness of carbapenem use was determined by multidisciplinary team according to pre-specified criteria. Second cycle was done in August 2018 after interventions taken and the result was compared with first cycle.

Analysis and interpretation:
A total of 39 empiric carbapenem prescriptions were audited in first cycle. Result showed only 15.4% carbapenem use was appropriate, which fulfilled all criteria of correct use (indication, dose, route, duration, de-escalation, microbiological sample taken). Forty two percents of the prescriptions were inappropriate indication and only 30.8% of the prescriptions were de-escalated appropriately on day 4. Besides, only 41% of carbapenem prescriptions were reviewed by antimicrobial stewardship (AMS) team.

Strategy for change:
Improving prescribing practice by incorporation of risk-factor guided prescribing checklist, dosage and de-escalation guides into carbapenem order form. In addition, role of AMS is strengthened by enforcing prescriber to discuss with consultant if unable to de-escalate empirical carbapenem on day four. Besides, frequency of carbapenem review was increased by splitting AMS team into smaller teams (ward AMS team).

Effects of change:
Percentage of appropriate carbapenem empirically prescribed increased from 15.4% to 81.1%. Achievable Benefit Not Achieved (ABNA) was improved from 64.6% to -1.1%. CRE incidence was reduced from 21 cases (11 deaths) in year 2017 to 15 cases (7 deaths) in year 2018. Three-monthly usage of meropenem was reduced by 7.3% and imipenem was reduced by 58% compared with pre interventions.

The next step:
We plan to explore more roles of AMS team and couple with infection control measures to further reduce CRE incidence.