Readiness for antimicrobial resistance (AMR) surveillance in Pakistan; a model for laboratory strengthening

Limited capacity of laboratories for antimicrobial susceptibility testing (AST) presents a critical diagnostic bottleneck in resource limited countries. This paper aims to identify such gaps and to explore whether laboratory networks could contribute towards improving AST in low resource settings.

A self-assessment tool to assess antimicrobial susceptibility testing capacity was administered as a pre-workshop activity to participants from 30 microbiology laboratories in 3 cities in Pakistan. Data from public and private laboratories was analyzed and capacity of each scored in percentage terms. Laboratories from Karachi were invited to join a support network. A cohort of five laboratories that consented were provided additional training and updates sessions over a period of 15 months. Impact of training activities in these laboratories was evaluated using a point scoring (0-11) tool.

Results of self-assessment component identified a number of areas that required strengthening (scores of ≤60%). These included; readiness for AMR surveillance; 38 and 46%, quality assurance; 49 and 55%, and detection of specific organisms; 56 and 60% for public and private laboratories respectively. No significant difference was detected in AST capacity between public and private laboratories [ANOVA; p > 0.05]. Scoring tool used to assess impact of training within the longitudinal cohort showed an increase from a baseline of 1-5.5 (August 2015) to improved post training scores of 7-11 (October 2016) for the 5 laboratories included. Moreover, statistical analysis using paired t-Test Analysis, assuming unequal variance, indicated that the increase in scored noted represents a statistically significant improvement in the components evaluated [p < 0.05].

Strengthening of laboratory capacity for AMR surveillance is important. Our data shows that close mentoring and support can help enhance capacity for antimicrobial sensitivity testing in resource limited settings. Our study further presents a model wherein laboratory networks can be successfully established and used towards improving diagnostic capacity in such settings.