Genexpert utilization at a public hospital laboratory in Northern Malawi: a quality improvement study

Abstract

Background Tuberculosis (TB) is one of the major public health challenges in the world. New diagnostic measures have been developed as part of the fight against the pandemic. GeneXpert is a molecular TB diagnostic tool that detects the presence of TB and reports rifampicin resistance in two hours. However, there have been reports of underutilization of this diagnostic method in many settings, Malawi inclusive. Evidence from Mzuzu Central Hospital in Northern Malawi show low utilization of GeneXpert. Quality improvement, using the Model for Improvement (MFI) approach, has been shown to improve practice among healthcare providers and utilization of diagnostic platforms. This study therefore used the MFI to develop and implement change ideas for improvement of GeneXpert utilization at Mzuzu Central Hospital in Northern Malawi. Methods A quasi-experimental study comprising of pre-intervention, intervention and postintervention phases with mixed methods study design was conducted to assess the baseline utilization of GeneXpert, conduct a root-cause analysis, develop and implement change ideas, and evaluate the impact of the change ideas. Record review was done to assess level of utilization before the intervention. The qualitative component was used to identify barriers to GeneXpert utilization and develop change ideas before the intervention, and assess level of improvement after the intervention. Key informants were purposively selected for in-depth interviews. A quality improvement team (QIT) was formed to develop and implement change ideas. Data were analyzed using STATA 14 SE and Nvivo12. Tables, bar graphs and run charts were used to present data. Results At baseline, the mean number of samples run per week was 11.0 (SD: 3.0). The weekly rejection rate was 8.2% (SD: 5.9). The mean weekly turnaround time was 80 hours (SD:15.6). After the intervention, the mean of samples run per week was 27 (145.6% increase), weekly rejection rate was 0.29% (96.5% improvement) and the turnaround time was 16 hours (80% improvement). These improvements were stable and sustained throughout the postintervention study period. Conclusion The implemented small change ideas using the MFI approach improved GeneXpert utilization at Mzuzu Central Hospital. The improvements in all the three indicators were statistically significant. The barriers to utilization were technical (module failures, power interruptions and stock outs), human resource (low doctor to patient ratio), financial (lack of donor support and GeneXpert expenses), motivational (lack of GeneXpert trainings) knowledge and information (lack of interdepartmental communication, incorrect completing of patient forms, insufficient sample production instructions to patients). Thus, it is important that the ministry of health and funding agencies tailor their funding towards increasing GeneXpert knowledge among health personnel and that the continuous professional developments on GeneXpert be encouraged in hospital departments.