Text analysis is a method originating mainly from empirical social sciences to assay the content of written communication with respect to a special problem. In a simple quantitative approach the frequency of relevant key words is assessed. Qualitative analyses include a valuation of the content. Prognostic analyses try to estimate the impact of content to the reader. Text analyses provide a basis for the comparison of written communication and for the test of hypotheses. Summary of product characteristics (SPC) of drugs are prepared according to a rigid structure as stipulated by regulatory authorities. Because of this structure and the limited size, SPCs are convenient targets for text analysis. Text analysis of SPCs may be important in the control of drug therapy because SPCs define the pattern of clinical application of a drug. A catalogue was developed for the text analysis of SPCs with respect to the content related to therapeutic drug monitoring (TDM), i.e. assay of plasma levels. It was avoided to use a comprehensive checklist of key words to facilitate the development and application of the catalogue. Instead, the paragraphs dose, adverse drug reactions, drug interactions, overdose, pregnancy/breast feeding, and pharmacokinetics (items 1 to 6) are assessed for explicit (score 2 per paragraph) or implicit (score 1 per paragraph) references to TDM. Examples of key words and typical phrases defining explicit and implicit content are provided. Moreover, the entire SPC is screened for ranges of plasma concentrations at defined doses as well as toxic and therapeutic ranges of plasma concentrations (items 7 to 9: counted with 3, 3 and 4 points, respectively). Finally, in item 10, the SPC is screened for an explicit and general recommendation to use TDM in the clinical application of the drug, which is counted with 5 points. Item 10 is counted with 3 points in case of a recommendation of TDM in only one special clinical situation. Thus, a summary score of the SPC-Content related to TDM (SPCCTDM) can be calculated by addition of the ten items with a maximum score of 27. The SPCCTDM is regarded as a measure of the impact that gets a reader on the value of TDM in the clinical application of a drug. Items 1 to 9 provide a quantitative estimate of the content of TDM in an SPC. A valuation pro TDM in an SPC is derived mainly from item 10. Spearman's rho=0.945 was found in a preliminary study of the reliability of 2 raters for 8 SPCs with a range of SPCCTDM of 0 to 22. The application of SPCCTDM to 48 German SPCs of psychopharmaceuticals resulted in considerable differences with respect to medico-scientific evidence [1].
Literature: [1] Ulrich S, Hiemke C, Laux G et al. Value and actuality of the prescription information for therapeutic drug monitoring of psychopharmaceuticals: A comparison with the medico-scientific evidence. Pharmacopsychiat 2007; 40(3): 121-127.