Making pragmatic sense of data in the tuberculosis laboratory register

SETTING: Tuberculosis (TB) microscopy network in Moldova, Mongolia, Uganda and Zimbabwe. OBJECTIVE: To evaluate how scrutinising the information recorded in the TB Laboratory Register can assist in improving the performance of the microscopy laboratory network and TB case management. METHODS: Review of records for completeness in registration of age, sex, reason for examination, analysis of variability patterns in serial smears and provision of exemplary statistical process charts of scanty positive results over time in the four countries. RESULTS: A total of 128 808 records were analysed. A large proportion of examinees in Uganda (6.9%) and Zimbabwe (3.9%) had no information on sex. The reason for examination was unknown for 7.4%. Among suspects with three smear results with at least one positive result, 56.1% had no variation in the pattern. Statistical process control charts revealed considerable fluctuation in the frequency of scanty positive smears over a calendar year. CONCLUSION: An analysis of information recorded in the TB Laboratory Register not only identifies strengths and weaknesses in the laboratory, it also reveals weaknesses on a much wider scale, in the entire case management system. TB laboratory data can thus demonstrate when, where and what action is indicated, and how performance might be monitored over time. © 2008 The Union.