Using classification tree modelling to investigate drug prescription practices at health facilities in rural Tanzania.

BACKGROUND: Drug prescription practices depend on several factors related to the patient, health worker and health facilities. A better understanding of the factors influencing prescription patterns is essential to develop strategies to mitigate the negative consequences associated with poor practices in both the public and private sectors. METHODS: A cross-sectional study was conducted in rural Tanzania among patients attending health facilities, and health workers. Patients, health workers and health facilities-related factors with the potential to influence drug prescription patterns were used to build a model of key predictors. Standard data mining methodology of classification tree analysis was used to define the importance of the different factors on prescription patterns. RESULTS: This analysis included 1,470 patients and 71 health workers practicing in 30 health facilities. Patients were mostly treated in dispensaries. Twenty two variables were used to construct two classification tree models: one for polypharmacy (prescription of ≥3 drugs) on a single clinic visit and one for co-prescription of artemether-lumefantrine (AL) with antibiotics. The most important predictor of polypharmacy was the diagnosis of several illnesses. Polypharmacy was also associated with little or no supervision of the health workers, administration of AL and private facilities. Co-prescription of AL with antibiotics was more frequent in children under five years of age and the other important predictors were transmission season, mode of diagnosis and the location of the health facility. CONCLUSION: Standard data mining methodology is an easy-to-implement analytical approach that can be useful for decision-making. Polypharmacy is mainly due to the diagnosis of multiple illnesses.

Paediatric pharmacovigilance: use of pharmacovigilance data mining algorithms for signal detection in a safety dataset of a paediatric clinical study conducted in seven African countries.

BACKGROUND: Pharmacovigilance programmes monitor and help ensuring the safe use of medicines which is critical to the success of public health programmes. The commonest method used for discovering previously unknown safety risks is spontaneous notifications. In this study we examine the use of data mining algorithms to identify signals from adverse events reported in a phase IIIb/IV clinical trial evaluating the efficacy and safety of several Artemisinin-based combination therapies (ACTs) for treatment of uncomplicated malaria in African children. METHODS: We used paediatric safety data from a multi-site, multi-country clinical study conducted in seven African countries (Burkina Faso, Gabon, Nigeria, Rwanda, Uganda, Zambia, and Mozambique). Each site compared three out of four ACTs, namely amodiaquine-artesunate (ASAQ), dihydroartemisinin-piperaquine (DHAPQ), artemether-lumefantrine (AL) or chlorproguanil/dapsone and artesunate (CD+A). We examine two pharmacovigilance signal detection methods, namely proportional reporting ratio and Bayesian Confidence Propagation Neural Network on the clinical safety dataset. RESULTS: Among the 4,116 children (6-59 months old) enrolled and followed up for 28 days post treatment, a total of 6,238 adverse events were reported resulting into 346 drug-event combinations. Nine signals were generated both by proportional reporting ratio and Bayesian Confidence Propagation Neural Network. A review of the manufacturer package leaflets, an online Multi-Drug Symptom/Interaction Checker (DoubleCheckMD) and further by therapeutic area experts reduced the number of signals to five. The ranking of some drug-adverse reaction pairs on the basis of their signal index differed between the two methods. CONCLUSIONS: Our two data mining methods were equally able to generate suspected signals using the pooled safety data from a phase IIIb/IV clinical trial. This analysis demonstrated the possibility of utilising clinical studies safety data for key pharmacovigilance activities like signal detection and evaluation. This approach can be applied to complement the spontaneous reporting systems which are limited by under reporting.