Provision of warfarin education to hospital inpatients

Patient education increases compliance and reduces toxicity or lack of effect during warfarin treatment [1–4]. A proportion of patients start warfarin treatment while in hospital. The need for adequate patient education before discharge is essential to ensure treatment efficacy and minimize toxicity. Although studies have shown that drug information provided to hospitalized patients is inadequate there are no specific data regarding warfarin treatment before discharge [5, 6]. This issue is clinically relevant because acute care patients are being discharged earlier than even before and with increasingly complicated medication regimens [7]. The guidelines for use of the National In-patient Medication Chart, introduced in Australia in 2005, state that all hospitalized patients commenced on warfarin should receive counselling and be given an information booklet. The prescribing doctor or pharmacist must fill a warfarin education record (WER), located next to the warfarin ordering section in the medication chart, as a record that these risk mitigation activities have been completed [8]. We audited the provision of patient education and written drug information to those patients commenced on warfarin during hospitalization at Flinders Medical Centre, a public teaching hospital within a regional health service, between June 2006 and December 2006. The following parameters were extracted from the central patient database: patient hospital number, date of admission, and hospital location for the specified time period. In order to determine which of these patients were treated with warfarin investigation results were reviewed to identify which patients had successive coagulation studies performed on at least 2 consecutive days during admission. Medical records for each of these patients were then reviewed to determine whether the patient had been started on warfarin during admission. In order to assess the distribution of a warfarin information booklet concurrent with initiation of warfarin treatment during admission the following parameters were gathered: age, gender, reason for warfarin use, and completion of the WER box in the drug chart. This information was entered into a database for analysis. The primary outcome was WER completion. No ethics committee approval was sought as this was a retrospective audit. Data are presented as means (SD) or frequencies. Comparisons between groups were performed by Student's t-test or Fischer exact test. A binary logistic regression was performed to identify the factors independently associated with WER completion (SPSS for Windows 14.0, Chicago, IL, USA). A P value <0.05 indicated statistical significance. Of the 557 patients treated with warfarin during the study period 43 were commenced on this drug during hospitalization. Their clinical and demographic characteristics are shown in Table 1. WER was filled only in five patients (11.6%). Within the group of patients started on warfarin during hospitalization there were no significant differences between the patients with WER completed and of those with WER not completed (Table 2). Regression analysis did not identify any factor independently associated with WER completion. Table 1 Demographic and clinical characteristics of patients initiating warfarin therapy during admission (n = 43) Table 2 Demographic and clinical characteristics of patients with WER completed and of those with WER not completed The results show that the existing policy on the provision of information to hospitalized patients started on warfarin is poorly followed. The lack of facilitation of educational material was not dependent on ward location, age, gender, or reasons for anticoagulation. One of the main shortcomings could lie in the hospital medication distribution system. The decision to begin a patient on warfarin involves personnel from a variety of disciplines including physicians, nurses, pharmacists, and allied health personnel. Warfarin is often ordered during ward rounds and then later distributed to the patient by nursing staff out of the ward stock. This process can easily bypass a pharmacy intervention altogether, an intervention which is a more traditional avenue for the distribution of educational material along with the drug. The streamlining of an existing process for the distribution of warfarin information, i.e. pharmacy involvement with all patients initiated on warfarin, could improve the quality of information provision. However, some sort of alert system would be necessary to inform pharmacy personnel of a new warfarin patient, and further study would be needed to assess the feasibility of such a system. Another factor that likely contributed to the poor results is the lack of standardization of the storage of warfarin information booklets as (i) the educational materials were difficult to locate and in some cases not present on the wards and (ii) the booklets themselves were in short supply at the hospital pharmacy. Significant improvement in the quality of provision of warfarin education could be achieved by completing these simple steps alone. A potential limitation to the interpretation of our results is that the lack of WER completion in the drug chart did not necessarily imply lack of information provided to the patient. It is theoretically possible that the hospital staff may have forgotten or overlooked this step. In our opinion, however, this is unlikely as the new National In-patient Medication Chart was introduced the year before the audit was conducted and seminar sessions and email alerts were arranged for hospital staff prior to its introduction. In conclusion, the quality of the provision of educational materials to hospitalized patients started on warfarin is generally poor. Strategies including an alert system of new warfarin patients for pharmacists and/or standardized storage and ensured availability of educational material on all wards could improve the results observed in this study. Further investigations into each of these areas, to ensure adherence to existing policies, are warranted to determine their feasibility and cost-effectiveness. Conflicts of interest: None declared.