Article
Benefits of electronic prescription support in discharge letters
Search Medline for
Authors
Published: | September 10, 2008 |
---|
Outline
Text
Introduction
After discharge from hospital timely and comprehensive information is a prerequisite for a seamless care of a patient. Communication at this step usually occurs in written form with a discharge letter summarizing the most relevant aspects of the latest history and current therapeutic needs. Therefore an essential part of a discharge letter is information on drug therapy [Ref. 1], [Ref. 2], [Ref. 3]. Ambulatory and hospital health care sectors vary in multiple essential characteristics in the health care systems worldwide: the regulatory framework may differ as may economic conditions, available drug formularies and their size. German hospitals are obligated by law to consider price differences inside and outside the hospital and to mention for every drug in a discharge letter cheaper therapeutic alternatives if available (Article 115c of the German Code of Social Law [§115c, SGB V]). Moreover, for every medication in a discharge letter hospitals must also specify its active ingredient to facilitate switching. Hence, to mention only brand names is unlawful.
Keeping this reglementation in mind electronic decision support for drug selection and ordering may maximize effectiveness and safety. Three years after introducing a system for integrating electronic drug recommendations in discharge letters (AiDKlinik) [Ref. 4] usage rate varies among the different departments of Heidelberg University Hospital. We analysed the differences in drug recommendations in letters composed with and without using the electronic system in a hospital providing tertiary care. We also aimed to assess the differences in medication cost that may emerge from the compulsory recommendation of cheaper alternatives.
Methods
We randomly selected a representative sample of 1800 discharge letters from a one-month period. The sample was balanced to consider differences in the number of patients between individual departments of the hospital. Letters were anonymized and converted into PDF format. All drugs mentioned in the discharge letters were reviewed by trained staff and categorized according to the following criteria: unambiguous brand name, unambiguous active ingredient, strength, and dosage schedule. A web-based application for data entry was developed and data was stored in a SQL database (MySQL).
Results
The investigation showed that 1280 out of 1800 (71.1%) discharge letters contained pharmacotherapeutic recommendations and 319/1280 (24.9%) were composed with electronic prescription support (AiDKlinik). Letters written without AiDKlinik contained active ingredients (15.9% versus 97.8%) significantly less often (χ2-test, p<0.001). When the electronic system for drug selection was used, nearly all (98.4%) recommendations were made with the tool. One third of the letters written without AiDKlinik neither denominated an active ingredient (34.1%) nor suggested a cheaper alternative (see Table 1 [Tab. 1]).
The cost difference between the drug recommended by the physician and the alternative listed by the system was on average 4.89 euros per suggested alternative resulting in 17.4 euros per letter.
Discussion
Three years after the implementation of an electronic tool to support pharmacotherapeutic recommendations according to German law one in four discharge letters is written with electronic prescription support. About 8% of the letters without electronic support named only the active ingredient which can cause problems in the frequently occurring subsequent switching process at the interface between primary and tertiary care because no information about the possibility of splitting, allergenic additives or different pharmaceutical forms or routes of application is provided.
In our hospital, which cares for more than 50,000 patients every year, roughly 35,000 discharge letters are written, all of which suggest further drug therapy. Assuming that all 35,000 discharge letters are written with electronic support, the recommended drug therapy could be up to 609,350 euros less expensive than before.
Acknowledgment
We thank all students that took part in the practical on information systems in health care in winter 2007/08 of the medical informatics program (University of Heidelberg/University of Applied Sciences Heilbronn).
References
- 1.
- Cochrane RA, Mandal AR, Ledger-Scott M, Walker R. Changes in drug treatment after discharge from hospital in geriatric patients. BMJ. 1992;305(6859):956-7.
- 2.
- Glintborg B, Andersen SE, Dalhoff K. Insufficient communication about medication use at the interface between hospital and primary care. Qual Saf Health Care. 2007;16(1):34-9.
- 3.
- Wilson S, Ruscoe W, Chapman M, Miller R. General practitioner-hospital communications: a review of discharge summaries. Qual Clin Pract. 2001;21(4):104-8.
- 4.
- Knaup P, Pilz J, Kaltschmidt J, Ludt S, Szecsenyi J, Haefeli WE. Standardized documentation of drug recommendations in discharge letters: A contribution to quality management in cooperative care. Methods Inf Med 2006;45:336-42.