What is known and objective: The validation of a method for recording pharmaceutical interventions measures the instrument's ability to provide consistent values when the same analysis is performed several times. Our aim was to validate the inter-rater reliability of the method used to record pharmaceutical interventions in our hospital.
Methods: We recorded interventions in a database, entering variables related to the patient, treatment and impact of the recommendation. We also recorded the type, cause and clinical significance of the negative outcome associated with use of the medicinal product (NOM). Twenty interventions performed during a 3-year study period (2007-2009) were randomly tested for consistency to analyse the kappa (κ) coefficient statistic of the recommendations as coded by nine senior and junior clinical pharmacists.
Results and discussion: There were 87·8% global consistency for NOM cause, 66·1% for intervention impact and 95·0% for NOM type. Agreement was substantial for 'intervention reasons', with a κ value of 0·74 (95%CI 0·61-0·87), fair for 'intervention impact', with a κ value of 0·24 (95%CI 0·15-0·32) and excellent for 'NOM type', with a κ value of 0·87 (95%CI 0·71-1·00), respectively. Our results are globally good, especially with regard to the analysis of intervention reasons and NOM type, which matches other authors' findings. Furthermore, our validation method is suitable for recording and considering the impact of pharmaceutical interventions.
What is new and conclusion: We describe a systematic method for clinical pharmacists to record their activities and assess their value. This methodology should help in the development of clinical pharmacy in Spain and should be translatable to other settings.
© 2011 Blackwell Publishing Ltd.