Implementation of modified q-control chart in monitoring of inspection characteristics with finite quantification sensitivity limits: A case study of bioburden enumeration in capsule shell

Abstract

Application of statistical process control (SPC) methodologies has become increasingly crucial in various business fields to control, monitor and improve processes, in addition to the support of the management decision-making. However, many inspection characteristics have limited quantification limits beyond which results are reported as either “< lower than” or “> higher than” such as testing of microbiological burden in healthcare products after definite dilution level ex. 1:10, 1:50 or 1:100. The present case study demonstrated the application of combined quality score trending chart concept with Laney-modified attribute process-behavior chart for monitoring the bioburden the quality of successive deliveries of empty hard gelatin capsule. Microbiological database was segmented into intervals based on both the specification limits and the minimum sensitivity limit of the Total Viable Aerobic Count (TVAC) and Total Yeast and Mold Count (TYMC). Each segment was assigned a score starting for low bioburden value to the higher threshold. Preliminary investigation of the dataset pattern showed that the record did not follow any presumed distribution for construction of the ordinary control chart. Data are right (positively) skewed with no apparent tendency to follow Poisson, binomial or normal distribution. Laney-Quality chart demonstrated bioburden contents as CFUs ranks with aberrant results spotted.