Potential drug interactions : exposure and management in hospital and ambulatory settings

Health care professionals are responsible to ensure safe dispensing and use of drug
regimens involving the use of drug combinations that may interact and cause serious
adverse events. In the last 40 years an enormous amount of data on drug
interactions has been published. But, although potential drug interactions are
probably common only few of them manifest serious adverse events and often only in
predisposed patients. Therefore, health care professionals feel inundated with hints
for potential drug interactions of questionable clinical significance provided by their
drug interactions information sources. Computerised alerts systems enable important
assistance but their performance is not satisfying.
Simply knowing that two drugs may interact does not offer enough information to
health care professional to devise a plan to reduce risk of an adverse outcome. The
risk of most drug interactions can be minimised by an accurate management (e.g. by
dose adjustment, spacing of dosing times and close monitoring of the therapy) and
thus, drug combinations do not have to be avoided. Therefore, drug interaction
information sources should directly provide guidelines about the manageability of a
drug interaction.
The present thesis aimed to focus on four different aspects of the management of
potential drug interactions in hospitalised and ambulatory patients: A) to determine
the influence of patient-related risk factors on the development of an adverse
outcome, B) to assess the prevalence and patient knowledge of potential drug
interactions with over-the-counter (OTC) drugs used for self-medication, C) to assess
preoccupation with potential drug interactions, perception of quality of drug
interaction information sources, information needs, and how their requirements relate
to those expressed by general practitioners, and D) to observe on site the
management of potential drug interactions in daily community pharmacy practice.
Drugs have been recognised as a primary or contributing cause of hyperkalaemia,
especially when administered to patients with underlying risk factors. The objective of
project A was to analyze the influence of known risk factors on the velocity to
develop hyperkalaemia in 551 hospitalised patients. Compared to the drug treatment
at entry, during hospitalisation significantly more patients were treated with drugs
associated with hyperkalaemia such as heparins, angiotensin converting enzyme
inhibitors (ACEIs) or angiotensin receptor blockers (ARBs), potassium supplements,
potassium-sparing diuretics, and/or NSAIDs or COX-2 selective inhibitors. Risk
factors associated with a high velocity to develop hyperkalaemia were in descending
order: use of potassium supplements, severe renal impairment, use of potassiumsparing
diuretics, use of ACEIs or ARBs, and diabetes mellitus. The velocity to
develop hyperkalaemia significantly increased in patients with ≥2 of such risk factors.
Dose-effects could be found for potassium supplements and potassium sparing
diuretics, but not for ACEIs or ARBs. In contrast, use of kaliuretics (loop diuretics or
thiazides) was associated with a decreased velocity to develop hyperkalaemia.
The results of this study have shown that patients with multiple risk factors should be
closely monitored and a rapid change in laboratory values should alert health care
providers to adequate actions.
Project B focussed on selected potential drug interactions of different clinical
relevance between prescription only medicines (POMs) and OTC drugs pharmacy
customers purchased for self-medication. In community pharmacies potential drug
interactions with self-medication arise mainly in two situations: First, if an OTC drug is
purchased by a passer-by customer whose prescribed drug therapy is not known, or
second, if a POM or an OTC drug is requested by a regular customer whose
prescribed drug therapy is usually recorded. Both customer groups were checked for
potential drug interactions. Of 1183 observed passer-by customers, 164 (14.4%)
purchased at least one of selected OTC drugs with risk for potential drug interactions.
Out of them 102 (62.2%) were interviewed: 43 (42.2%) mentioned taking prescribed
drugs, and 3 of them were exposed to potential drug interactions of moderate
severity.
Out of 592 regular customers using at least one selected drug with a risk for potential
drug interactions, 434 (73.3%) could be interviewed. Of them 69 (15.9%) were
exposed to a potential drug interaction between purchased OTC drug for selfmedication
and their POM. Furthermore, 116 (26.7%) regular customers were
exposed to potential drug interactions within their prescribed drugs and in 28 (6.5%)
multiple (≥ 2) potential drug interactions were found. Out of 434 regular customers
203 (46.8%) were aware of potential drug interactions between their POM and OTC
drugs. Of them 96 (47.3%) were informed by their prescribing physician and 52
(25.6%) by their community pharmacist. Awareness of potential drug interaction was
significantly associated with the age of customers and the potential severity of drug
interactions.
Thus, the results of this study support efforts to increase awareness of potential drug
interactions with OTC drugs. Although community pharmacies are adequately
equipped with computerised drug interaction surveillance systems this is often not
applied to self-medication. Vigilance for potential interactions of all drugs, including
those sold over the counter, should be increased.
Project C aimed to analyze the current drug interaction management in Swiss
community pharmacies with a particular focus on electronic systems and to compare
the results with those gathered among German general practitioners in a recent
survey. A postal questionnaire was randomly sent to 500 community pharmacies of
the German part of Switzerland. The response rate was 57.4%. Only 24.7%
pharmacists reported to be confronted less than daily with potential drug interactions.
Use of computer software to identify potential drug interactions was widespread in
community pharmacies (90.2%) and the software was the primary source of
information (81.2 ± 29.6%). The quality of the interaction software was judged
sensitive (identifying all dangerous interactions) by 80.5 ± 21.5% but specific
(identifying only relevant interactions) by only 38.3 ± 32%. Pharmacists declared a
low override rate (14%) of drug interaction alerts although unjustified alerts were
reported by 60.6 ± 33.1%. In contrast to general practitioners pharmacists opted less
often for information on the mechanism of the interaction and more frequently for
details for dose adjustment. Both groups complained about deficient information on
non-interacting alternatives. The information needs of community pharmacists
differed considerably from general practitioners.
Substantial improvement of drug interaction software systems is thus required at
least in two important aspects: the suppression of inappropriate alerts and the
tailoring to the needs of the user.
Drug interaction alert systems are commonly used in community pharmacies. They
intend to ensure safe medication dispensing and use. But, pharmacists are inundated
with alerts and override is possible. In project D on-site practice of community
pharmacies was observed and the nature and management of drug interaction alerts
were analysed. During two days 15 researchers assessed in 15 different pharmacies
data of 600 regular customers with multiple drug therapy (≥ 2 drugs) and interviewed
the responsible pharmacists about the management actions in consequence of drug
interaction alerts. The median frequency of drug interaction alerts increased from 0.5
to 40 to 76 depending on the settings of the 15 community pharmacies’ computer
systems to flag only severe (N=4), severe and moderate (N=6) or severe, moderate
and minor (N=5) potential drug interactions. Because of these settings out of 787
potential drug interactions detected on new or repeated prescriptions 277 (35.2%)
were technically overridden by computer systems. Only 256 (32.5%) of 787 potential
drug interactions emerged from a new prescription. The drug interaction alert
systems produced 656 alerts of which 146 (22.3%) were invalid because of multiple
alerts for the same interaction or alerts for combinations of which one drug was no
longer taken. Of the 510 remaining relevant drug interaction alerts 289 (56.7%) were
overridden by community pharmacists without any evaluation. The attendance of the
patients by the pharmacists themselves was associated with a lower override of
alerts. The sum of technical and pharmacist’s override results in a rate of 71.9%. Of
the remaining 211 potential drug interactions 87 (41.2%) were analysed trough
consultation of literature, a physician or the patient himself and of them 55 (63.2%)
resulted in an intervention (close monitoring, adjustment of dose or ingestion time,
stop of therapy, or alternative therapy). Determinants associated with the analysis of
drug interaction alerts were the potential high severity (severe or moderate) and the
alert flagged for the first time.
As long as no sophisticated solutions are available it is important to avoid override of
clinically relevant potential drug interactions. All of the 10 potential drug interactions
classified as severe were detected and adequately managed. Therefore,
classification of potential drug interactions is a very strong determinant for detection.
Two conclusions are drawn from this study: Firstly, a focus on first-time alerts
generated by new prescriptions and the elimination of invalid alerts would result in a
substantial improvement in the specificity of drug interaction alert systems, and
secondly, the claim to reduce their sensitivity by filtering drug interaction of moderate
or minor severity might be reduced.