Comparison of the Completeness of Prescription Medication Histories: DISCUSSION part 2
Previous work suggests that prescription databases are an accurate means of identifying drugs dispensed to individual patients. One study has demonstrated good agreement between computerized prescription records and written prescription records from a medical clinic. Information in prescription databases is also similar to that obtained from self-administered medication-history questionnaires. However, previous work suggests that medication histories documented in hospital records are often inaccurate. Beers and others reported that 83% of medication histories recorded for persons over the age of 65 years had at least one error when compared with information obtained during a structured interview. The results presented here suggest that the completeness of medication-history information varies with the profession of the person recording the information. The pharmacists’ monitoring form, which was not a part of the health care record at St Paul’s Hospital, included the highest proportion of medications taken before admission, according to the definition used in this study, regardless of whether PharmaNet was reviewed.
The current study has several limitations. First, it was not possible to validate the accuracy of the list of the medications patients took before admission. It is possible that medications listed in some sections of the health care record or in PharmaNet were not currently being used by the patient in question. However, it was assumed that patients and family would be unlikely to overreport the number of medications in their current regimen. A previous study demonstrated that hospital records underreport medications taken before admission much more commonly than list medications that patients said they had not used. tadacip 20 mg
Another limitation was the possible bias in the selection of patients for whom PharmaNet was accessed. Pharmacists may have chosen to use PharmaNet for patients who were more likely to experience drug-related problems. PharmaNet access was available on a limited number of computer terminals in the Pharmacy Department, and profiles contained information about all prescriptions filled within the previous 14-month period. As a consequence, PharmaNet profiles could be complicated and time-consuming to evaluate. Pharmacists were likely to limit the use of PharmaNet to those patients with more complex problems and to those who seemed to be at risk for problems related to medications taken before hospital admission. This assumption seems to be substantiated by the higher number of medications taken before admission for the patients for whom PharmaNet was used.
In most of the charts reviewed for this study, medication-history information was documented by some, but not all, of the professional groups assessed. The interpretation of when documenting medication- history information in the medical record is warranted likely varies among health care professions. This could have resulted in a selection bias that affected the results of this study. For example, if certain professions reserved documentation of medication-history information for patients with more complex medical histories, the documentation of the proportion of medications taken before admission may have been affected. The order in which medication histories were documented could also have been important, since later documentation could have been affected by the details listed by a different health care professional earlier in the chart. A prospective study randomly assigning different health care professionals to take patients’ histories could have avoided such bias. More systematic verification of data extraction and validation (i.e., beyond spot checks) could also have helped minimize the possibility of bias during the data- collection process. Finally, since PharmaNet includes only information about prescription medications, over-the-counter and complementary or alternative medicines were not included in the analysis.
Physicians, nurses, pharmacists, and other heath care professionals frequently describe the importance and difficulty of obtaining accurate medication histories.1,2,22,23 Accurate medication-history information is vital in the prevention of drug-related morbidity and mortality, especially for geriatric patients. These results suggest that pharmacists obtain the most complete medication-history information for geriatric patients, although the difference between that obtained by pharmacists and that obtained by other health care professionals was not statistically significant. The use of a computerized prescription database to obtain medication histories seems to be associated with effective identification of medications taken before admission and resolution of drug-related problems; however, to obtain the most comprehensive information possible, pharmacists may also need to obtain medication-history information directly from the patient. Future research that uses a prospective, randomized design, and a definitive assessment of medications taken before admission as a comparator is required before firm conclusions can be drawn about the relative completeness of medication-history information obtained by various health care professionals, with and without the use of PharmaNet. viagra 50 mg
